IEEE
You are not logged in, please sign in to edit > Log in / create account  

Oral-History:Jack Balde

From GHN

(Difference between revisions)
Jump to: navigation, search
m (Text replace - "[[Category:Power, energy & industry application|" to "[[Category:Power, energy & industry applications|")
(15 intermediate revisions by 5 users not shown)
Line 1: Line 1:
== About Jack Balde<br> ==
+
== About Jack Balde ==
  
Balde received his Bachelors of Electrical Engineering from RPI. He then worked at Western Electric on airborne radar, underwater sonar, and navigational systems autopilots, hearing aids, and other military subjects up to ca. 1958. From 1956 on he taught at Western Electric’s in-house graduate engineering program, where he taught radar systems designing, engineering statistics, human factors engineering, and other systems design. He then moved to Western Electric’s Engineering Research Center, where he worked on tantalum thin film circuits. He worked with the IEEE to help create standards for package for semi-conductors. He mentions his association with the Electronic Components and Technology Conference. He spends extensive time discussing the structure of various conferences, how they aid or resist technological dynamism, and the way association of conferences unifies fields, the creation of new conferences creates new fields.
+
<p>[[John W. Balde|Balde]] received his Bachelors of Electrical Engineering from RPI. He then worked at Western Electric on airborne [[Radar|radar]], underwater sonar, and navigational systems autopilots, hearing aids, and other military subjects up to ca. 1958. From 1956 on he taught at Western Electric’s in-house graduate engineering program, where he taught radar systems designing, engineering statistics, human factors engineering, and other systems design. He then moved to Western Electric’s Engineering Research Center, where he worked on tantalum [[Thin Film Deposition Technologies|thin film]] circuits. He worked with the [[IEEE History|IEEE]] to help create standards for package for [[Semiconductors|semiconductors]]. He mentions his association with the Electronic Components and Technology Conference. He spends extensive time discussing the structure of various conferences, how they aid or resist technological dynamism, and the way association of conferences unifies fields, the creation of new conferences creates new fields. </p>
  
== About the Interview<br> ==
+
== About the Interview ==
  
JACK BALDE: An Interview Conducted by Robert Colburn, IEEE History Center, 2 June 1999
+
<p>JACK BALDE: An Interview Conducted by Robert Colburn, IEEE History Center, 2 June 1999 </p>
  
<br>Interview # 348 for the IEEE History Center, The Institute of Electrical and Electronics Engineers, Inc., and Rutgers, The State University of New Jersey
+
<p>Interview # 348 for the IEEE History Center, The Institute of Electrical and Electronics Engineers, Inc. </p>
  
<br>
+
== Copyright Statement  ==
  
<br>
+
<p>This manuscript is being made available for research purposes only. All literary rights in the manuscript, including the right to publish, are reserved to the IEEE History Center. No part of the manuscript may be quoted for publication without the written permission of the Director of IEEE History Center. </p>
  
== Copyright Statement<br> ==
+
<p>Request for permission to quote for publication should be addressed to the IEEE History Center Oral History Program, 39 Union Street, New Brunswick, NJ 08901-8538 USA. It should include identification of the specific passages to be quoted, anticipated use of the passages, and identification of the user. </p>
  
This manuscript is being made available for research purposes only. All literary rights in the manuscript, including the right to publish, are reserved to the IEEE History Center. No part of the manuscript may be quoted for publication without the written permission of the Director of IEEE History Center.<br><br>
+
<p>It is recommended that this oral history be cited as follows: </p>
  
<br>
+
<p>Jack Balde, an oral history conducted in 1999 by Robert Colburn, IEEE History Center, New Brunswick, NJ, USA. </p>
  
Request for permission to quote for publication should be addressed to the IEEE History Center Oral History Program, Rutgers - the State University, 39 Union Street, New Brunswick, NJ 08901-8538 USA. It should include identification of the specific passages to be quoted, anticipated use of the passages, and identification of the user. <br><br>
+
== Interview  ==
  
<br>
+
<p>INTERVIEW: Jack Balde </p>
  
It is recommended that this oral history be cited as follows:<br>Jack Balde, an oral history conducted in 1999 by Robert Colburn, IEEE History Center, Rutgers University, New Brunswick, NJ, USA.
+
<p>INTERVIEWER: Robert Colburn </p>
  
<br>
+
<p>DATE: 2 June 1999 </p>
  
== Interview<br> ==
+
<p>PLACE: San Diego, California </p>
  
INTERVIEW: Jack Balde<br>INTERVIEWER: Robert Colburn<br>DATE: 2 June 1999<br>PLACE: San Diego, California<br>
+
=== Education and Western Electric career overview  ===
  
=== Education and Western Electric career overview<br> ===
+
<p>'''Colburn:''' </p>
  
'''Colburn:'''<br>If I could start with your early academic training, schooling and how you got into the field?<br>  
+
<p>If I could start with your early academic training, schooling and how you got into the field? </p>
  
<br>  
+
<p>'''Balde:''' </p>
  
'''Balde:'''<br>My background is a Bachelors of Electrical Engineering from RPI. I had worked at Bell Laboratories for a couple of summer jobs, and I went to Western Electric when I came back out of school working on radar and navigational systems autopilot, hearing aids, up until about 1958. I worked on mostly airborne radar and navigational systems, but also some Navy systems as well, including underwater sonar, and a variety of military things. <br>  
+
<p>My background is a Bachelors of Electrical Engineering from RPI. I had worked at Bell Laboratories for a couple of summer jobs, and I went to Western Electric when I came back out of school working on radar and navigational systems autopilot, hearing aids, up until about 1958. I worked on mostly airborne [[Radar|radar]] and navigational systems, but also some Navy systems as well, including underwater sonar, and a variety of military things. </p>
  
<br>  
+
<p>In about 1956 Tim Shea, vice president of Western Electric Co decided that Western Electric was going to have a graduate engineering training kind of activity very much like M.J. Kelly had done for [[Bell Labs|Bell Laboratories]], and I was drafted into being on the academic staff for that. I taught courses in radar systems designing and engineering statistics. (It was obvious that someone who thought statistics was a branch of calculus and didn’t have the faintest idea of how to use it had taught the average college graduate.) And I taught human factors engineering and radar systems engineering and basically system design kind of things, and wrote a couple of textbooks. </p>
  
In about 1956 Tim Shea, vice president of Western Electric Co decided that Western Electric was going to have a graduate engineering training kind of activity very much like M.J. Kelly had done for Bell Laboratories, and I was drafted into being on the academic staff for that. I taught courses in radar systems designing and engineering statistics. (It was obvious that someone who thought statistics was a branch of calculus and didn’t have the faintest idea of how to use it had taught the average college graduate.) And I taught human factors engineering and radar systems engineering and basically system design kind of things, and wrote a couple of textbooks. <br>
+
=== Western Electric Engineering Research Center  ===
  
=== Western Electric Engineering Research Center<br> ===
+
<p>'''Balde:''' </p>
  
'''Balde:'''<br>
+
<p>Then I got tagged to go to the Engineering Research Center, which was something a Vice President Tim Shea set up basically to do manufacturing research. [[Bell Labs|Bell Laboratories]] did not serve the combined Western Electric/[[Bell Labs|Bell Laboratories]] team well in anything but design, and many of the designs were not manufactured at the lowest cost, by a long shot. So the idea was to do manufacturing research to improve things. </p>
  
Then I got tagged to go to the Engineering Research Center, which was something a Vice President Tim Shea set up basically to do manufacturing research. Bell Laboratories did not serve the combined Western Electric/Bell Laboratories team well in anything but design, and many of the designs were not manufactured at the lowest cost, by a long shot. So the idea was to do manufacturing research to improve things.  
+
<p>I worked in the area of interconnections, and the most important interconnection that we needed to deal with was of Tantalum thin film circuits, which was very much needed for telephone equipment in order to produce the frequency determining filters for touch tone dialing and things like that. That’s enough background to get you to where I first began going to the ECTC. </p>
  
<br>  
+
<p>'''Colburn:''' </p>
  
<br>  
+
<p>Yes. </p>
  
I worked in the area of interconnections, and the most important interconnection that we needed to deal with was of Tantalum thin film circuits, which was very much needed for telephone equipment in order to produce the frequency determining filters for touch tone dialing and things like that. That’s enough background to get you to where I first began going to the ECTC.<br>  
+
<p>'''Balde:''' </p>
  
<br>  
+
<p>The goal for Tantalum thin films was to make both resistors and capacitors, but not so much inductors. The man who led the activity at Bell Laboratories was Dave McLean. Others working on the team were Bob Barry, John Fisher and others at Allentown, and others including myself at the Princeton Engineering Research Center. I was working at the Engineering Research Center, and of my sixteen-eighteen patents about eleven-twelve of them were in the Tantalum film area, and we did some really pioneering work in Tantalum thin films. </p>
  
'''Colburn:'''<br>Yes.<br>
+
=== Electronic Components and Technology Conference ===
 
+
<br>
+
 
+
'''Balde:'''<br>The goal for Tantalum thin films was to make both resistors and capacitors, but not so much inductors. The man who led the activity at Bell Laboratories was Dave McLean. Others working on the team were Bob Barry, John Fisher and others at Allentown, and others including myself at the Princeton Engineering Research Center. I was working at the Engineering Research Center, and of my sixteen-eighteen patents about eleven-twelve of them were in the Tantalum film area, and we did some really pioneering work in Tantalum thin films. <br>
+
 
+
<br>
+
 
+
=== Electronic Components and Technology Conference<br> ===
+
  
 
==== Early involvement, IEEE standards task force  ====
 
==== Early involvement, IEEE standards task force  ====
  
'''Balde:'''<br>
+
<p>'''Balde:''' </p>
  
I started going to the ECTC when it was at the Marriott Twin Bridges. I guess I went as early as 1951 or 1952, and I went pretty consistently, although I missed a couple of years in the late fifties, and then went consistently from the middle sixties on. In the early stages ECTC was only a components conference. By components I mean resistors, capacitors, and maybe inductors and not much else. <br>  
+
<flashmp3>348 - balde - clip 1.mp3</flashmp3>
  
<br>  
+
<p>I started going to the ECTC when it was at the Marriott Twin Bridges. I guess I went as early as 1951 or 1952, and I went pretty consistently, although I missed a couple of years in the late fifties, and then went consistently from the middle sixties on. In the early stages ECTC was only a components conference. By components I mean resistors, capacitors, and maybe inductors and not much else. </p>
  
By about the mid-‘70s we had had a need to move packages, on printed circuit boards; from the dual inline package, which needs a mass of through-holes, to a surface mount package, the square chip packages were called chip carriers, because other people had used the term chip carrier. <br>  
+
<p>By about the mid-‘70s we had had a need to move packages, on printed circuit boards; from the dual inline package, which needs a mass of through-holes, to a surface mount package, the square chip packages were called chip carriers, because other people had used the term chip carrier. </p>
  
<br>  
+
<p>In combination with Dan Amy of Sperry Univac, we proposed to bring some kind of order out of the chaos using an IEEE task force. At that stage there were over 3,500 different chip carriers, most of them within 1/8th to 1/10th of an inch of another per side. The leads were spaced differently, the sizes were spaced differently, and so on and so forth. So we proposed bringing a standardized footprint for all packages of ceramic or plastic. We then organized a task force of some twenty-seven companies, most of them competitors, who agreed to cooperate on the grounds that they had to do something about the chaos. We used the old rule saying that no two sizes should be separated any less than by the square root of two, to give a decent family of sizes. We hammered out what the sizes were and what the lead spacings of fifty millimeters were and the commonality between ceramic with edge clips or ceramic without edge clips, or the molded plastic packages. Molded plastic cases with open lids were proposed by [[Dimitry Grabbe|Dimitry Grabbe]] of AMP and were filled with silicone gel. </p>
  
In combination with Dan Amy of Sperry Univac, we proposed to bring some kind of order out of the chaos using an IEEE task force. At that stage there were over 3,500 different chip carriers, most of them within 1/8th to 1/10th of an inch of another per side. The leads were spaced differently, the sizes were spaced differently, and so on and so forth. So we proposed bringing a standardized footprint for all packages of ceramic or plastic. We then organized a task force of some twenty-seven companies, most of them competitors, who agreed to cooperate on the grounds that they had to do something about the chaos. We used the old rule saying that no two sizes should be separated any less than by the square root of two, to give a decent family of sizes. We hammered out what the sizes were and what the lead spacings of fifty millimeters were and the commonality between ceramic with edge clips or ceramic without edge clips, or the molded plastic packages. Molded plastic cases with open lids were proposed by Dimitry Grabbe of AMP and were filled with silicone gel. <br>  
+
<p>There was a slightly different family for ceramic that had forty-millimeter spacing; most PC boards had fifty-millimeter spacing. The reason why this is important is that it represents a major change that happened to the ECTC Conference focus. </p>
  
<br>  
+
<p>After we created this task force and hammered out all the details, it was just a matter of taking it to JEDEC, which is the Joint Electron Device Council, which then goes through an official standardization process. The activity that I was doing was a precursor consensus. Nepcon West introduced five different sizes in 1978, but without any standardization activity. </p>
  
There was a slightly different family for ceramic that had forty-millimeter spacing; most PC boards had fifty-millimeter spacing. The reason why this is important is that it represents a major change that happened to the ECTC Conference focus.  
+
<p>What was needed was to publicize the results of the task force standardization effort. It was obvious that we had this IEEE task force with a final report to give, and we needed to get the word out about what was going on. At that time I had not been consistently active in the ECTC, and was not in the committee structure for the ECTC. A man named Tom Matcovitch was, and he was a professor at Drexel University and worked in ceramic hybrids, but mostly in ceramic structures for ceramic capacitors for resistance. Remember the ECTC hadn’t gone beyond components then. </p>
  
<br>
+
==== Semi-conductors, interconnects, and system design  ====
  
<br>  
+
<p>'''Balde:''' </p>
  
After we created this task force and hammered out all the details, it was just a matter of taking it to JEDEC, which is the Joint Electron Device Council, which then goes through an official standardization process. The activity that I was doing was a precursor consensus. Nepcon West introduced five different sizes in 1978, but without any standardization activity.<br>  
+
<p>I decided that it was high time that the ECTC dealt with the package for semi-conductors as a component. So I got hold of Matcovitch, who had some stature in ECTC, and he agreed to give the talk, and that represented the first talk in IC packaging at the ECTC. It was funny that when he got asked questions he would turn to me and ask if I wanted to comment, and that was about the same year that Christian Bernard was on the lecture tour with his chauffeur and— </p>
  
<br>  
+
<p>'''Colburn:''' </p>
  
What was needed was to publicize the results of the task force standardization effort. It was obvious that we had this IEEE task force with a final report to give, and we needed to get the word out about what was going on. At that time I had not been consistently active in the ECTC, and was not in the committee structure for the ECTC. A man named Tom Matcovitch was, and he was a professor at Drexel University and worked in ceramic hybrids, but mostly in ceramic structures for ceramic capacitors for resistance. Remember the ECTC hadn’t gone beyond components then.
+
<p>He was the doctor fellow? </p>
  
==== Semi-conductors, interconnects, and system design ====
+
<p>'''Balde:''' </p>
  
'''Balde:'''<br>  
+
<p>Yes. After twenty lectures the chauffeur said, “I’ve heard that so often that I could give your talk,” so he said, “Go do it.” He did well on the talk and he got a question and he said, “Oh, that’s so easy, my chauffeur could answer it.” Well, that’s what happened in this thing. After that we were on a roll with respect to the IC packages. </p>
  
I decided that it was high time that the ECTC dealt with the package for semi-conductors as a component. So I got hold of Matcovitch, who had some stature in ECTC, and he agreed to give the talk, and that represented the first talk in IC packaging at the ECTC. It was funny that when he got asked questions he would turn to me and ask if I wanted to comment, and that was about the same year that Christian Bernard was on the lecture tour with his chauffeur and— <br>  
+
<p>This is a conference that had never done any systems work, and the next thing that happened was that there were people who began to use the ceramic packages or the multichip modules, notably IBM and NEC. Meanwhile the NTT people under Takakki Ohsaki had come up with a ceramic base, with applied layers of organic interconnections. It was the very first MCM-D multi-chip module. Papers from NEC, NTT and IBM appeared at the 1982 ECTC at Seattle. That was the next major turn in the expansion of the scope of ECTC focus, where all of a sudden we had system design sections as well as some of the components. </p>
  
<br>  
+
<p>Now that we had the system thing in place in 1982, and interconnects in general, it became increasingly apparent that the ECTC had to deal with the increasing number of papers being submitted that fell loosely in the category of interconnects. So we had to do something about that. If you are only one session of seven papers and you’ve got fifty-eight, you’re in deep trouble. </p>
  
'''Colburn:'''<br>He was the doctor fellow?<br>
+
==== Single chip packaging; multi-chip modules  ====
  
<br>  
+
<p>'''Balde:''' </p>
  
'''Balde:'''<br>Yes. After twenty lectures the chauffeur said, “I’ve heard that so often that I could give your talk,” so he said, “Go do it.” He did well on the talk and he got a question and he said, “Oh, that’s so easy, my chauffeur could answer it.” Well, that’s what happened in this thing. After that we were on a roll with respect to the IC packages.  
+
<p>I think the first thing we did was to create a section on single chip packaging, and I talked the very reluctant Mike McShane into agreeing to be assistant chair and he’s hung on there ever since. So that worked well. We were loaded with modeling and simulation papers from John Prince and his fellows at the University of Arizona, so once when we had about forty papers and could choose only ten, I made the motion to break away all the modeling simulation papers into a separate session. It was a very selfish motive, because that meant that we could be in control of what was left. In fact, we didn’t even plan it. It was in the session that we were doing the advance paper selection, and it went through before anybody knew what hit them. Sometimes it worked that way. And so it was broken off, the interconnect session from a session in multi-chip modules. </p>
  
<br>  
+
<p>Multi-chip module activity was an interesting excursion. What happened is we began to get thirty to forty papers in multi-chip modules. It was really hot stuff. </p>
  
<br>  
+
<p>'''Colburn:''' </p>
  
This is a conference that had never done any systems work, and the next thing that happened was that there were people who began to use the ceramic packages or the multichip modules, notably IBM and NEC. Meanwhile the NTT people under Takakki Ohsaki had come up with a ceramic base, with applied layers of organic interconnections. It was the very first MCM-D multi-chip module. Papers from NEC, NTT and IBM appeared at the 1982 ECTC at Seattle. That was the next major turn in the expansion of the scope of ECTC focus, where all of a sudden we had system design sections as well as some of the components.
+
<p>When was this? </p>
  
<br>  
+
<p>'''Balde:''' </p>
  
<br>  
+
<p>This would have been 1987 or 1988, something like that. When we had only fifteen, we could pick seven good ones, that was fine. One year—I’d have to do the research to find out which year—we had forty-two. </p>
  
Now that we had the system thing in place in 1982, and interconnects in general, it became increasingly apparent that the ECTC had to deal with the increasing number of papers being submitted that fell loosely in the category of interconnects. So we had to do something about that. If you are only one session of seven papers and you’ve got fifty-eight, you’re in deep trouble.
+
<p>'''Colburn:''' </p>
  
<br>  
+
<p>Do you remember what forces were driving that? </p>
  
==== Single chip packaging; multi-chip modules ====
+
<p>'''Balde:''' </p>
  
'''Balde:'''<br>  
+
<p>Oh, the multi-chip module was the way to go to increase the system packaging density. It’s still the major driving force today. They’re the most powerful force in all of the elements of manufacturing technology. It’s still multi-chip modules at the core. </p>
  
I think the first thing we did was to create a section on single chip packaging, and I talked the very reluctant Mike McShane into agreeing to be assistant chair and he’s hung on there ever since. So that worked well. We were loaded with modeling and simulation papers from John Prince and his fellows at the University of Arizona, so once when we had about forty papers and could choose only ten, I made the motion to break away all the modeling simulation papers into a separate session. It was a very selfish motive, because that meant that we could be in control of what was left. In fact, we didn’t even plan it. It was in the session that we were doing the advance paper selection, and it went through before anybody knew what hit them. Sometimes it worked that way. And so it was broken off, the interconnect session from a session in multi-chip modules.
+
==== Ceramics  ====
  
<br>  
+
<p>'''Colburn:''' </p>
  
<br>  
+
<p>Was there a particular breakthrough that made that suddenly more economical? </p>
  
Multi-chip module activity was an interesting excursion. What happened is we began to get thirty to forty papers in multi-chip modules. It was really hot stuff.<br>  
+
<p>'''Balde:''' </p>
  
<br>  
+
<p>Yes. The Takakki Ohsaki proposed the breakthrough for getting out of all ceramics and using copper/polyimide layers. It started with NTT. Again we had an IEEE task force which formalized IPC designations of MCM-D, which is the organic material deposited on usually a ceramic substrate. In MCM-C all the layers were ceramic; MCM-L all the layers were laminate multilayer. And if you had organically deposited layers over L as a substrate, you say D over L; and if you put organically deposited layers for the interconnect, you could have D over C. So you have got all these combinations, but they are now official. </p>
  
'''Colburn:'''<br>When was this?<br>  
+
<p>The problem with ceramic is that it was very expensive. As implemented by IBM, it was actually thirty-eight to forty-seven layers of ceramic in something three inches square. It cost a bundle, but it was the only way to do the job. If you weren’t an IBM or NEC, you couldn’t possibly afford it. There was some military work by John Bauer at RCA in Moorstown, but those were little ceramic single chip packages, half an inch square or one inch square. </p>
  
<br>  
+
<p>What happened is that though it took four to five years to do it, the breakthrough came out of the organics deposited on the base substrate of printed circuit laminate. There were a number of processes, the most important being IBM Yasu, which made their SLT, which is deposited organic, which can give you higher density because the lines can be much finer than printed circuit board or ceramic. So there was this whole rush of companies getting in the act, keeping up with the Joneses to be on the cutting edge. This resulted in forty-two papers. </p>
 
+
'''Balde:'''<br>This would have been 1987 or 1988, something like that. When we had only fifteen, we could pick seven good ones, that was fine. One year—I’d have to do the research to find out which year—we had forty-two.<br>
+
 
+
<br>
+
 
+
'''Colburn:'''<br>Do you remember what forces were driving that?<br>
+
 
+
<br>
+
 
+
'''Balde:'''<br>Oh, the multi-chip module was the way to go to increase the system packaging density. It’s still the major driving force today. They’re the most powerful force in all of the elements of manufacturing technology. It’s still multi-chip modules at the core.<br>
+
 
+
<br>
+
 
+
==== Ceramics ====
+
 
+
'''Colburn:'''<br>Was there a particular breakthrough that made that suddenly more economical?<br>
+
 
+
<br>
+
 
+
'''Balde:'''<br>Yes. The Takakki Ohsaki proposed the breakthrough for getting out of all ceramics and using copper/polyimide layers. It started with NTT. Again we had an IEEE task force which formalized IPC designations of MCM-D, which is the organic material deposited on usually a ceramic substrate. In MCM-C all the layers were ceramic; MCM-L all the layers were laminate multilayer. And if you had organically deposited layers over L as a substrate, you say D over L; and if you put organically deposited layers for the interconnect, you could have D over C. So you have got all these combinations, but they are now official.
+
 
+
<br>
+
 
+
<br>
+
 
+
The problem with ceramic is that it was very expensive. As implemented by IBM, it was actually thirty-eight to forty-seven layers of ceramic in something three inches square. It cost a bundle, but it was the only way to do the job. If you weren’t an IBM or NEC, you couldn’t possibly afford it. There was some military work by John Bauer at RCA in Moorstown, but those were little ceramic single chip packages, half an inch square or one inch square.
+
 
+
<br>
+
 
+
<br>
+
 
+
What happened is that though it took four to five years to do it, the breakthrough came out of the organics deposited on the base substrate of printed circuit laminate. There were a number of processes, the most important being IBM Yasu, which made their SLT, which is deposited organic, which can give you higher density because the lines can be much finer than printed circuit board or ceramic. So there was this whole rush of companies getting in the act, keeping up with the Joneses to be on the cutting edge. This resulted in forty-two papers.  
+
 
+
<br>  
+
  
 
==== Committee structure and session organization  ====
 
==== Committee structure and session organization  ====
  
The ECTC has a committee structure which did not turn out to be very effective for this dynamic or any flood of papers on a new subject. A committee is appointed on each of the major subjects. The way in which it is organized is both an asset and a detriment. They do not assemble a group of session chairs on various topics who have expertise in those areas; rather, they have standing committees; one on interconnection, one on polyimide, one on ceramics, and so on. The problem with that structure is that each committee head is concerned with building his own turf, so that if you have one session for the topic of the committee, each committee wants to apply for two, even if no new papers are coming in. Then a push is made to get the necessary papers to fill the extra slots. It makes it very, very difficult to kill anything, because these people are defending the subject for the session, even if there are few papers. <br>  
+
<flashmp3>348 - balde - clip 2.mp3</flashmp3>
  
<br>  
+
<p>The ECTC has a committee structure which did not turn out to be very effective for this dynamic or any flood of papers on a new subject. A committee is appointed on each of the major subjects. The way in which it is organized is both an asset and a detriment. They do not assemble a group of session chairs on various topics who have expertise in those areas; rather, they have standing committees; one on interconnection, one on polyimide, one on ceramics, and so on. The problem with that structure is that each committee head is concerned with building his own turf, so that if you have one session for the topic of the committee, each committee wants to apply for two, even if no new papers are coming in. Then a push is made to get the necessary papers to fill the extra slots. It makes it very, very difficult to kill anything, because these people are defending the subject for the session, even if there are few papers. </p>
  
But you also have a situation where there might be a session on multichip modules which use copper and polyimide as the interconnector on the surface. It is hard to assign them to a multichip module session because the polyimide people might ask for the best papers that belong in multichip module, because it’s polyimide dielectric. Well if you think about it logically, there is polyimide in everything these days, which means that the whole conference could be a subset of the polyimide session. And that’s obviously chaos.  
+
<p>But you also have a situation where there might be a session on multichip modules which use copper and polyimide as the interconnector on the surface. It is hard to assign them to a multichip module session because the polyimide people might ask for the best papers that belong in multichip module, because it’s polyimide dielectric. Well if you think about it logically, there is polyimide in everything these days, which means that the whole conference could be a subset of the polyimide session. And that’s obviously chaos. </p>
  
<br>  
+
<p>The other problem is that, again using polyimide as an example, the twenty multichip module papers are narrowed down to the best seven. Three or four of them are given to interconnect and they are accepted. The paper selection committee meeting resembles a horse trading bazaar, like an open auction or the stock market, “I’ll trade you one of these for two of those,” and so on, and really a free for all with yelling and shouting people calling down to the end of the room to somebody else. It is very much like the stock market. But the problem is that if a guy only has three papers and he needs four more to get a session, he’ll accept anything. He will go to the people who rejected papers and basically pick them up off the floor. </p>
  
<br>  
+
<p>Now some of them were rejected because you had eight good papers and could use only seven, but some of them were actually rejected because you had twenty papers and ten of them were not up to even a minimum standard. Then the papers which were really unacceptable end up in other sessions, which of course weakens the conference. The other consequence of that, (another weakness of the ECTC), is that it takes an extreme amount of courage on the part of the general chair to keep from adding sessions. You don’t want to offend the connector people, for instance, by taking away their precious connector sessions, but since you’ve got more papers on multichip modules than you are able to place, you have two or three sessions on multichip modules. </p>
  
The other problem is that, again using polyimide as an example, the twenty multichip module papers are narrowed down to the best seven. Three or four of them are given to interconnect and they are accepted. The paper selection committee meeting resembles a horse trading bazaar, like an open auction or the stock market, “I’ll trade you one of these for two of those,” and so on, and really a free for all with yelling and shouting people calling down to the end of the room to somebody else. It is very much like the stock market. But the problem is that if a guy only has three papers and he needs four more to get a session, he’ll accept anything. He will go to the people who rejected papers and basically pick them up off the floor. <br>  
+
<p>The Denver Multichip Module Conference was set up with fifteen sessions maximum. There are five different times of day—morning, afternoon; morning, afternoon and morning—and only three concurrent. The reason why three works is there is always one session you don’t care about. Whatever your interests, they are catholic enough to encompass all three, and you shuttle back between the other two and the time. </p>
  
<br>  
+
<p>Iwona Turlik and John Segelken and a couple of others tried to fight it, but before we knew it we had eighteen, twenty, twenty-two, twenty-four, twenty-six, twenty-eight, thirty, thirty-two sessions at the ECTC. If you go to a meeting like this, the most you can possibly get to is ten percent of the conference. And if you, as I do, report to clients what is important in what happened here, it makes my job impossible because I’m only going to see one paper out of ten, at best. So you end up with a conference with 800 people, but 800 people are dispersed between a number of talks. One of those sessions has forty people and a workshop has sixty to eighty. Thus there is a sort of fight going on between the minimalists who want to try to get back to fifteen or eighteen sessions and those who want to continue to expand the number. And that’s the problem. </p>
  
Now some of them were rejected because you had eight good papers and could use only seven, but some of them were actually rejected because you had twenty papers and ten of them were not up to even a minimum standard. Then the papers which were really unacceptable end up in other sessions, which of course weakens the conference. The other consequence of that, (another weakness of the ECTC), is that it takes an extreme amount of courage on the part of the general chair to keep from adding sessions. You don’t want to offend the connector people, for instance, by taking away their precious connector sessions, but since you’ve got more papers on multichip modules than you are able to place, you have two or three sessions on multichip modules.  
+
<p>[[C. P. Wong|C.P. Wong]] has belonged to the “grow it bigger and bigger school,” taking as an image the Materials Research conference in Boston. It does twenty-five simultaneous sessions in five hotels over a mile and a half. I went there once when I was invited to give a talk. Four or five of us were talking about packaging. One guy gave a talk I had heard so many times I could have given the talk myself, so I decided to go out and look to see what was happening elsewhere in the conference. One nearby session was on nuclear waste disposal at sea and the other one was on dental implants. No community of interest there whatsoever. There was one talk on fine line printed circuits, but it was a forty-five-minute walk away for a half-an-hour talk and I was on next, so I couldn’t go to that. </p>
  
<br>  
+
<p>I stayed for two days and there was not a single person I saw on Tuesday that stayed for Wednesday. I checked with other people. John Stafford came and gave a talk on Friday. He flew in at one o’clock, gave his talk at two o’clock, and left on the next flight. That’s how a huge conference runs, as a series of independent workshops. </p>
  
<br>  
+
<p>The academic types who go there only get off to one conference once a year so they want to make the most of it. But the industrial types that go to the ECTC don’t want a week-long conference because their job may not be there after a week if they are not defending their turf. You can be missed for two and a half days and get along very nicely, but four or five days and lose the next weekend is a whole week and it can be a big problem. So you have the [[C. P. Wong|C.P. Wongs]] who would like nothing better than to have twenty-five simultaneous sessions and those of us who say twenty-five sessions with zero people in attendance at each one doesn’t make any sense. You get more speakers, but less non-speaker attendees, because they can’t hear them all anyway, they may as well stay home and buy the proceedings. So you have this kind of interplay going on which really is a big open issue and needs to be resolved. Everything else works well, the organization, hierarchy, and so on and so forth. </p>
 
+
The Denver Multichip Module Conference was set up with fifteen sessions maximum. There are five different times of day—morning, afternoon; morning, afternoon and morning—and only three concurrent. The reason why three works is there is always one session you don’t care about. Whatever your interests, they are catholic enough to encompass all three, and you shuttle back between the other two and the time.
+
 
+
<br>
+
 
+
<br>
+
 
+
Iwona Turlik and John Segelken and a couple of others tried to fight it, but before we knew it we had eighteen, twenty, twenty-two, twenty-four, twenty-six, twenty-eight, thirty, thirty-two sessions at the ECTC. If you go to a meeting like this, the most you can possibly get to is ten percent of the conference. And if you, as I do, report to clients what is important in what happened here, it makes my job impossible because I’m only going to see one paper out of ten, at best. So you end up with a conference with 800 people, but 800 people are dispersed between a number of talks. One of those sessions has forty people and a workshop has sixty to eighty. Thus there is a sort of fight going on between the minimalists who want to try to get back to fifteen or eighteen sessions and those who want to continue to expand the number. And that’s the problem.
+
 
+
<br>
+
 
+
<br>
+
 
+
C.P. Wong has belonged to the “grow it bigger and bigger school,” taking as an image the Materials Research conference in Boston. It does twenty-five simultaneous sessions in five hotels over a mile and a half. I went there once when I was invited to give a talk. Four or five of us were talking about packaging. One guy gave a talk I had heard so many times I could have given the talk myself, so I decided to go out and look to see what was happening elsewhere in the conference. One nearby session was on nuclear waste disposal at sea and the other one was on dental implants. No community of interest there whatsoever. There was one talk on fine line printed circuits, but it was a forty-five-minute walk away for a half-an-hour talk and I was on next, so I couldn’t go to that. <br>
+
 
+
<br>
+
 
+
I stayed for two days and there was not a single person I saw on Tuesday that stayed for Wednesday. I checked with other people. John Stafford came and gave a talk on Friday. He flew in at one o’clock, gave his talk at two o’clock, and left on the next flight. That’s how a huge conference runs, as a series of independent workshops.
+
 
+
<br>
+
 
+
<br>
+
 
+
The academic types who go there only get off to one conference once a year so they want to make the most of it. But the industrial types that go to the ECTC don’t want a week-long conference because their job may not be there after a week if they are not defending their turf. You can be missed for two and a half days and get along very nicely, but four or five days and lose the next weekend is a whole week and it can be a big problem. So you have the C.P. Wongs who would like nothing better than to have twenty-five simultaneous sessions and those of us who say twenty-five sessions with zero people in attendance at each one doesn’t make any sense. You get more speakers, but less non-speaker attendees, because they can’t hear them all anyway, they may as well stay home and buy the proceedings. So you have this kind of interplay going on which really is a big open issue and needs to be resolved. Everything else works well, the organization, hierarchy, and so on and so forth.<br>  
+
  
 
=== Comparison of various conferences, workshops  ===
 
=== Comparison of various conferences, workshops  ===
  
'''Colburn:'''<br>Has it needed tuning over the years? Have ECTC and the CPMT Society grown well over the years with the needs of industry and researchers?<br>  
+
<p>'''Colburn:''' </p>
  
<br>  
+
<p>Has it needed tuning over the years? Have ECTC and the CPMT Society grown well over the years with the needs of industry and researchers? </p>
  
'''Balde:'''<br>CPMT is doing more at the workshop level. They are in competition, or sometimes cooperation with IMAPS. They never cooperated much with the International Society for Hybrid Macroelectronics (ISHM), an organization doing ceramic hybrids, but they cooperated some with the International Electronic Packaging Society (IEPS), an organization more involved with systems. The systems kind of papers you would get from NEC and Takakki Ohsaki and IBM would have been standard in IEPS.
+
<p>'''Balde:''' </p>
  
<br>  
+
<p>CPMT is doing more at the workshop level. They are in competition, or sometimes cooperation with IMAPS. They never cooperated much with the International Society for Hybrid Macroelectronics (ISHM), an organization doing ceramic hybrids, but they cooperated some with the International Electronic Packaging Society (IEPS), an organization more involved with systems. The systems kind of papers you would get from NEC and Takakki Ohsaki and IBM would have been standard in IEPS. </p>
  
<br>  
+
<p>IEPS and ISHM merged, and with the merger they have become stronger in running workshops. The [[IEEE History|IEEE]] has taken to doing that also. A workshop is an interestingly different sort of an event. There are some workshops, like the System Packaging Committee, which operates as if an attendance of 100 is a top limit. There used to be a rule in the IEEE that said that up to 100 you could ban tape recorders and representatives of the press and you didn’t have to write a proceedings. Therefore the talks would be off the record and patent free, so your licensing people couldn’t give you a hard time. </p>
  
IEPS and ISHM merged, and with the merger they have become stronger in running workshops. The IEEE has taken to doing that also. A workshop is an interestingly different sort of an event. There are some workshops, like the System Packaging Committee, which operates as if an attendance of 100 is a top limit. There used to be a rule in the IEEE that said that up to 100 you could ban tape recorders and representatives of the press and you didn’t have to write a proceedings. Therefore the talks would be off the record and patent free, so your licensing people couldn’t give you a hard time. <br>  
+
<p>From 100 to 300 were expected to call it a symposium, and generally at a symposium all the workshops are a one-ring circus. A lot of people like them. In the first place, they are informal. You interrupt the speaker right when you find that you don’t understand. If you want to take forty-five minutes with the answers to one guy, you do it and you either make up the time or you go to lunch late, no big deal. Whereas the minute you get to two simultaneous sessions you have the same problems you have at conferences because you have to keep in synch, and therefore they are not quite as free. My bias is that a good conference has three sessions and a terrible conference has six, simultaneously. </p>
  
<br>  
+
<p>At any rate, CPMT began to broaden its activities beyond that of merely sponsoring the ECTC. One of the first things it did was to split off a separate for conference for the manufacturing technology people. They began to want more and more sessions with manufacturing technology at the ECTC and under the old format if you wanted a manufacturing technology session you had to, heaven forbid, get rid of connectors, and so on and so forth. So obviously this was big warfare, and the thing to do was to break that off into a separate conference. That was pioneered when Mauro Walker did a conference in Japan which retroactively got labeled the Manufacturing Technology Conference. It had some elements of manufacturing technology and had some design elements as well, but then over the years as I served on the committee it became apparent to me that if you allowed too much in the way of design input it would be like homogenized milk, in that you can’t tell the difference between that IEMT (International Electronic Manufacturing Technology) and the ECTC. </p>
  
From 100 to 300 were expected to call it a symposium, and generally at a symposium all the workshops are a one-ring circus. A lot of people like them. In the first place, they are informal. You interrupt the speaker right when you find that you don’t understand. If you want to take forty-five minutes with the answers to one guy, you do it and you either make up the time or you go to lunch late, no big deal. Whereas the minute you get to two simultaneous sessions you have the same problems you have at conferences because you have to keep in synch, and therefore they are not quite as free. My bias is that a good conference has three sessions and a terrible conference has six, simultaneously.  
+
<p>We had to hold the line. I remember making Jean Joly of Honeywell Ball very unhappy. He was a designated hitter for Honeywell Ball to the Manufacturing Technology Conference and he had a design sort of paper. I told him we didn’t have a place for that; it’s not eligible. He wasn’t very happy about it, because he wasn’t eligible to present their paper at the ECTC because Kurzwell was the designated hitter to present it at the ECTC. But out of that came more of a shaping of IEMT as a Manufacturing Technology Conference. </p>
  
<br>  
+
<p>The IEMT Conferences, which were started by Mauro Walker of Motorola in Japan then began to be followed by additional ones in both the U.S. and Japan. The Manufacturing one in the United States never got off the ground, not over 100. It was really not much more than a big workshop. Bill Bekenbaugh had been active in that one. He oversees one that is done very well: IEMT and Kanagawa and other places in Japan that have 200 or so attendants. I have a staff member who was involved with that to some extent, Kanji Otsuka. </p>
  
<br>  
+
<p>The other kind of activity that came from CPMT was the VLSI workshops. That was started by George Harmon of the National Bureau of Standards and [[Oral-History:Ron Gedney|Ron Gedney]], who worked at IBM. Harmon was into reliability work, and the problem is that you can’t get people to go to reliability workshops. Reliability workshops have a reputation similar to Sominex: chances of getting through without falling asleep are about zilch. But from his position at the National Bureau of Standards he wanted to have a place where reliability papers could go. And Ron Gedney wanted to have a workshop that would deal with the VLSI packaging, the chip in the package. It wasn’t IC design, it wasn’t system design; it was that middle niche, the interface between step one and step two, very large scale integration packaging. And so they joined hands, Harmon and Gedney, to organize the VLSI conference. I think the first one was in Silver Springs, Maryland, or it might have Research Triangle Park, and out of that they had a couple up in Marlboro and a couple of others. Kanji Otsuka decided to pick that up for Japan, and ran a series of very successful workshops every other year in Kyoto. </p>
  
At any rate, CPMT began to broaden its activities beyond that of merely sponsoring the ECTC. One of the first things it did was to split off a separate for conference for the manufacturing technology people. They began to want more and more sessions with manufacturing technology at the ECTC and under the old format if you wanted a manufacturing technology session you had to, heaven forbid, get rid of connectors, and so on and so forth. So obviously this was big warfare, and the thing to do was to break that off into a separate conference. That was pioneered when Mauro Walker did a conference in Japan which retroactively got labeled the Manufacturing Technology Conference. It had some elements of manufacturing technology and had some design elements as well, but then over the years as I served on the committee it became apparent to me that if you allowed too much in the way of design input it would be like homogenized milk, in that you can’t tell the difference between that IEMT (International Electronic Manufacturing Technology) and the ECTC.  
+
<p>Then they picked it up in Europe, running it in France. But meanwhile Elaine Pope of Intel was the one running it in the United States, but she was using a formula where she was not getting enough papers from the U.S. The U.S. core group died in that area. I believe it could be brought back to life if it was done the right way, though I’m not sure it is worth the effort. So you have the VLSI workshop which is almost big enough to be a conference and the IEMT conference which is small enough to be a workshop. And then all the rest of the workshops are kind of ad hoc. There might be one ball grid array for instance held at Binghamton, New York maybe at the offices of SUNY and another one held by other places. </p>
  
<br>  
+
<p>The [[IEEE Computer Society History|IEEE Computer Society]] had an ongoing maverick group called the Computer Packaging Committee, and the Computer Society increasingly became dominated by administrative nitpicking. The computer packaging group, which was moving away from computers and calling it sell system packaging, couldn’t stand the hair splitting of the Computer Society and they came into the ECTC as TC-14. It runs workshops in Japan and in Oiso and Tsukuba and Kyoto, and in the United States in the Poconos, South Padre Island, Santa Barbara, San Diego, Sedona, Las Vegas, Austin, Texas and in Europe in places like Brussels and Cork and the island of Malta. This has been a successful thing and it’s been an operation in the area of System Packaging since 1968, which is basically eight to ten years before we brought the first packaging to ECTC. And many of the people in that committee are the ones who have organized almost everything else, so that if you look carefully you will see that they got their organizational start in this system packaging committee that began in the IEEE Computer Society. They never ran a conference and don’t want anything but workshops; doesn’t want a conference, doesn’t have a newsletter; in other words, a maverick. </p>
  
<br>  
+
<p>And it’s now one of the TC’s, TC-14. CPMT offered us the ability to escape from the rules of the Computer Society, so it was the Computer Society’s loss and CPMT’s gain. The rule that really got in our way was that if you have a group, you have to have a full-fledged election for the next officer, complete with a biography and a photo dossier and two candidates in a competitive elections. Well, when you have a small committee you are darned lucky to get somebody whose arm you can twist into taking the position. You are not about to have a credible election. If I ask you to take the position and you say yes, and then I say you are going to have to run against so-and-so, then you’ll be more likely to say, “Well, if he wants to run, he can have it. Thank you very much and goodbye,” you know. And they created a lot of senseless rules that say you have to have a committee in place with people in place before you book a hotel. If you don’t know what day it is, how can you get anybody to serve? </p>
  
We had to hold the line. I remember making Jean Joly of Honeywell Ball very unhappy. He was a designated hitter for Honeywell Ball to the Manufacturing Technology Conference and he had a design sort of paper. I told him we didn’t have a place for that; it’s not eligible. He wasn’t very happy about it, because he wasn’t eligible to present their paper at the ECTC because Kurzwell was the designated hitter to present it at the ECTC. But out of that came more of a shaping of IEMT as a Manufacturing Technology Conference.  
+
<p>I’m digressing, but it won’t take me long to digress. One of the worst rules, they had a mailing list which consisted of anybody who filled in a postcard to be on the mailing list. Their mailing list was 650 people, and that committee ignores it. And they said, “How dare you ignore it? That’s a designated mailing list.” So we said we found that it’s a waste of time. The people from Bangladesh and Cuba and Pakistan don’t come. So why should we waste our time and effort notifying them? They said, “Well, you’ve got to do it.” So we took that 650-name mailing list and found only 65 people who had ever come to a meeting. And of those 65 people, 38 were dead or retired. And the remaining ones were on our mailing list. We had an independent main mailing list. At any rate, Mickey Mouse rules. At the CPMT it’s a little more freewheeling, and that’s been important. </p>
  
<br>  
+
<p>Anyway, you really have only these two other activities other than minor workshops, but now this TC-14 runs some of its workshops in combination with the Computer Society which refused to give it up. They were killing it, but they refused to give it up. And so it is jointly sponsored between the major players. And we are going to see more of the co-sponsoring operations. The materials workshop at Ojai, and there are a couple of others that have worked very well.[some continuity lost as tape was turned over at this point...] </p>
  
<br>
+
=== Multi-chip modules sessions at ECTC; IEEE involvement  ===
  
The IEMT Conferences, which were started by Mauro Walker of Motorola in Japan then began to be followed by additional ones in both the U.S. and Japan. The Manufacturing one in the United States never got off the ground, not over 100. It was really not much more than a big workshop. Bill Bekenbaugh had been active in that one. He oversees one that is done very well: IEMT and Kanagawa and other places in Japan that have 200 or so attendants. I have a staff member who was involved with that to some extent, Kanji Otsuka.
+
<p>'''Balde:''' </p>
  
<br>  
+
<p>One continuing problem is the competition between scheduled sessions and poster sessions. Well, poster sessions are not the greatest thing since sliced bread. It’s pretty bad. It means that you don’t really give the paper and nobody is going to get to hear it, only the people that wander into that room are going to see it. If you are going to get it published, you’d be better off giving it to another conference and get it published. </p>
  
<br>  
+
<p>So when we received forty-two papers on multichip modules, I took twenty-five of them and we did multichip module sessions at Nepcon West, the first time three sessions and the next time five sessions, and it was going to be ten sessions on multichip modules. The IEPS appointed someone as the official liaison with Nepcon West, a man out of Rolling Springs, Texas who was not familiar with multichip modules, and who said he didn’t want the position. By this time I had ten sessions with the chairs named and all the speakers selected, and he said, “Go fly a kite” and threw it in the wastebasket. </p>
  
The other kind of activity that came from CPMT was the VLSI workshops. That was started by George Harmon of the National Bureau of Standards and Ron Gedney, who worked at IBM. Harmon was into reliability work, and the problem is that you can’t get people to go to reliability workshops. Reliability workshops have a reputation similar to Sominex: chances of getting through without falling asleep are about zilch. But from his position at the National Bureau of Standards he wanted to have a place where reliability papers could go. And Ron Gedney wanted to have a workshop that would deal with the VLSI packaging, the chip in the package. It wasn’t IC design, it wasn’t system design; it was that middle niche, the interface between step one and step two, very large scale integration packaging. <br>And so they joined hands, Harmon and Gedney, to organize the VLSI conference. I think the first one was in Silver Springs, Maryland, or it might have Research Triangle Park, and out of that they had a couple up in Marlboro and a couple of others. Kanji Otsuka decided to pick that up for Japan, and ran a series of very successful workshops every other year in Kyoto.  
+
<p>So Harry Charles of ISHM said, “Would you like to move into a conference slot? We’ve got a hotel reserved in Denver to put on a multichip module conference.And I said only under three conditions: (1) No more than three papers simultaneous, (2) no standing committees, you have a committee of the whole that you divide up to become the sessions chairs, fifty percent of them rotate every year whether you like it or not, and condition (3) co-sponsored by ISHM, CPMT and IEPS, the main reason being to head them off at the pass. Otherwise they will each produce concurrent things in competition. </p>
  
<br>  
+
<p>And so that worked for a number of years. IEEE’s participation was weak in the start, and not adequate, for reasons that had not to do with intent but with accidents of time. Les Fox was moving out of packaging work at DEC, DEC wasn’t supporting things, people couldn’t go to meetings, and so on and so forth, and between Les Fox moving out and someone else moving in, the ball got dropped, and they did not supply their end and didn’t help with the program. And this was to the point that giving them one-third of the monetary proceeds when they didn’t do anything was not going to fly. So what has happened is that became a strong conference, which means that the MCM activity coming in to ECTC has now weakened, and in fact no longer has a committee </p>
 
+
<br>
+
 
+
Then they picked it up in Europe, running it in France. But meanwhile Elaine Pope of Intel was the one running it in the United States, but she was using a formula where she was not getting enough papers from the U.S. The U.S. core group died in that area. I believe it could be brought back to life if it was done the right way, though I’m not sure it is worth the effort. So you have the VLSI workshop which is almost big enough to be a conference and the IEMT conference which is small enough to be a workshop. And then all the rest of the workshops are kind of ad hoc. There might be one ball grid array for instance held at Binghamton, New York maybe at the offices of SUNY and another one held by other places.
+
 
+
<br>
+
 
+
<br>
+
 
+
The IEEE Computer Society had an ongoing maverick group called the Computer Packaging Committee, and the Computer Society increasingly became dominated by administrative nitpicking. The computer packaging group, which was moving away from computers and calling it sell system packaging, couldn’t stand the hair splitting of the Computer Society and they came into the ECTC as TC-14. It runs workshops in Japan and in Oiso and Tsukuba and Kyoto, and in the United States in the Poconos, South Padre Island, Santa Barbara, San Diego, Sedona, Las Vegas, Austin, Texas and in Europe in places like Brussels and Cork and the island of Malta. This has been a successful thing and it’s been an operation in the area of System Packaging since 1968, which is basically eight to ten years before we brought the first packaging to ECTC. And many of the people in that committee are the ones who have organized almost everything else, so that if you look carefully you will see that they got their organizational start in this system packaging committee that began in the IEEE Computer Society. They never ran a conference and don’t want anything but workshops; doesn’t want a conference, doesn’t have a newsletter; in other words, a maverick. <br>
+
 
+
<br>
+
 
+
And it’s now one of the TC’s, TC-14. CPMT offered us the ability to escape from the rules of the Computer Society, so it was the Computer Society’s loss and CPMT’s gain. The rule that really got in our way was that if you have a group, you have to have a full-fledged election for the next officer, complete with a biography and a photo dossier and two candidates in a competitive elections. Well, when you have a small committee you are darned lucky to get somebody whose arm you can twist into taking the position. You are not about to have a credible election. If I ask you to take the position and you say yes, and then I say you are going to have to run against so-and-so, then you’ll be more likely to say, “Well, if he wants to run, he can have it. Thank you very much and goodbye,” you know. And they created a lot of senseless rules that say you have to have a committee in place with people in place before you book a hotel. If you don’t know what day it is, how can you get anybody to serve?
+
 
+
<br>
+
 
+
<br>
+
 
+
I’m digressing, but it won’t take me long to digress. One of the worst rules, they had a mailing list which consisted of anybody who filled in a postcard to be on the mailing list. Their mailing list was 650 people, and that committee ignores it. And they said, “How dare you ignore it? That’s a designated mailing list.” So we said we found that it’s a waste of time. The people from Bangladesh and Cuba and Pakistan don’t come. So why should we waste our time and effort notifying them? They said, “Well, you’ve got to do it.” So we took that 650-name mailing list and found only 65 people who had ever come to a meeting. And of those 65 people, 38 were dead or retired. And the remaining ones were on our mailing list. We had an independent main mailing list. At any rate, Mickey Mouse rules. At the CPMT it’s a little more freewheeling, and that’s been important.
+
 
+
<br>
+
 
+
<br>
+
 
+
Anyway, you really have only these two other activities other than minor workshops, but now this TC-14 runs some of its workshops in combination with the Computer Society which refused to give it up. They were killing it, but they refused to give it up. And so it is jointly sponsored between the major players. And we are going to see more of the co-sponsoring operations. The materials workshop at Ojai, and there are a couple of others that have worked very well.<br>[some continuity lost as tape was turned over at this point...]<br>
+
 
+
<br>
+
 
+
=== Multi-chip modules sessions at ECTC; IEEE involvement ===
+
 
+
'''Balde:''' <br>One continuing problem is the competition between scheduled sessions and poster sessions. Well, poster sessions are not the greatest thing since sliced bread. It’s pretty bad. It means that you don’t really give the paper and nobody is going to get to hear it, only the people that wander into that room are going to see it. If you are going to get it published, you’d be better off giving it to another conference and get it published. <br>
+
 
+
<br>
+
 
+
So when we received forty-two papers on multichip modules, I took twenty-five of them and we did multichip module sessions at Nepcon West, the first time three sessions and the next time five sessions, and it was going to be ten sessions on multichip modules. The IEPS appointed someone as the official liaison with Nepcon West, a man out of Rolling Springs, Texas who was not familiar with multichip modules, and who said he didn’t want the position. By this time I had ten sessions with the chairs named and all the speakers selected, and he said, “Go fly a kite” and threw it in the wastebasket. <br>
+
 
+
<br>
+
 
+
So Harry Charles of ISHM said, “Would you like to move into a conference slot? We’ve got a hotel reserved in Denver to put on a multichip module conference.” And I said only under three conditions: (1) No more than three papers simultaneous, (2) no standing committees, you have a committee of the whole that you divide up to become the sessions chairs, fifty percent of them rotate every year whether you like it or not, and condition (3) co-sponsored by ISHM, CPMT and IEPS, the main reason being to head them off at the pass. Otherwise they will each produce concurrent things in competition. <br>
+
 
+
<br>
+
 
+
And so that worked for a number of years. IEEE’s participation was weak in the start, and not adequate, for reasons that had not to do with intent but with accidents of time. Les Fox was moving out of packaging work at DEC, DEC wasn’t supporting things, people couldn’t go to meetings, and so on and so forth, and between Les Fox moving out and someone else moving in, the ball got dropped, and they did not supply their end and didn’t help with the program. And this was to the point that giving them one-third of the monetary proceeds when they didn’t do anything was not going to fly. So what has happened is that became a strong conference, which means that the MCM activity coming in to ECTC has now weakened, and in fact no longer has a committee  
+
 
+
<br>  
+
  
 
=== Conference and workshop diversification  ===
 
=== Conference and workshop diversification  ===
  
'''Balde:'''<br>
+
<p>'''Balde:''' </p>
 
+
Not every conference should deal with every issue. If you do, then conferences become homogenized. Workshops are a completely different breed of animal. You can have a workshop on, let’s say, solder connection, and it really is completely independent from the workshop on ceramic substrates, which is clearly independent from the one about the printed circuit boards and fine line or maybe flex. In other words, there is almost an unlimited demand for workshops, because the guy who goes to the one workshop doesn’t want to go to the other one anyway. But at conferences, we haven’t gotten that idea across as well that you should have a conference that’s clearly different from anybody else’s conference. We have to resist a tendency for each conference to be all things to all people.
+
 
+
<br>
+
 
+
<br>
+
 
+
One of the things that bothers me is the fact that they want to bring in semiconductor design into the ECTC. Those are perfectly good IC components. Semiconductor design and manufacturing doesn’t belong. The goal of twenty-five simultaneous sessions is not a wise one. I mean, why go? You may as well stay home. The guy who goes to give a talk arrives about a half an hour before the talk, gives his talk, and leaves, and there are twenty-five people in the audience. The smaller the number of sessions there are, the higher the attendance in each session. So that’s one of the major current issues, with the business of the standing committee with its activity of perpetuating dead issues. The connector people didn’t want to give it up even when they only had two papers being submitted, and there is another who doesn’t want to give up multichip modules even though he has only two papers, because he wants to beat the bushes. He has a committee that says, “Give it up already,” but it’s this age-old battle that continues, which doesn’t deal well with obsolescence.<br>
+
 
+
<br>
+
 
+
[Transcriber note: The background noise level here increases substantially.]
+
 
+
<br>  
+
  
<br>  
+
<p>Not every conference should deal with every issue. If you do, then conferences become homogenized. Workshops are a completely different breed of animal. You can have a workshop on, let’s say, solder connection, and it really is completely independent from the workshop on ceramic substrates, which is clearly independent from the one about the printed circuit boards and fine line or maybe flex. In other words, there is almost an unlimited demand for workshops, because the guy who goes to the one workshop doesn’t want to go to the other one anyway. But at conferences, we haven’t gotten that idea across as well that you should have a conference that’s clearly different from anybody else’s conference. We have to resist a tendency for each conference to be all things to all people. </p>
  
I’m supposed to be applying my comments to ECTC. But comments on IEMT, VSLI, and workshops are interwoven and interchangeable. The business of spawning workshops, the organization of CPMT could be construed to be, in some sense, the ECTC in different places. It’s part of the whole—a combination of workshops and conferences. Workshops do not require a paper. In a way it’s like the bush leagues; you can try out a talk at a workshop and if it’s successful you write it after that and give it at a conference. In a way it’s more cutting edge. It gets the paper that isn’t approved for publication yet. Sometimes you get the paper that’s not quite ready for prime time, but that’s okay too. You get work-in-progress kind of papers. And I say, the IEMT is a conference, but it’s workshop size. And there are other workshops, such as the VLSI in Kyoto that are conference size. So we blur the label once again, and people fight to call it a workshop even though it isn’t. Some ground rules would help there I think.<br><br>  
+
<p>One of the things that bothers me is the fact that they want to bring in [[Semiconductors|semiconductor]] design into the ECTC. Those are perfectly good IC components. Semiconductor design and manufacturing doesn’t belong. The goal of twenty-five simultaneous sessions is not a wise one. I mean, why go? You may as well stay home. The guy who goes to give a talk arrives about a half an hour before the talk, gives his talk, and leaves, and there are twenty-five people in the audience. The smaller the number of sessions there are, the higher the attendance in each session. So that’s one of the major current issues, with the business of the standing committee with its activity of perpetuating dead issues. The connector people didn’t want to give it up even when they only had two papers being submitted, and there is another who doesn’t want to give up multichip modules even though he has only two papers, because he wants to beat the bushes. He has a committee that says, “Give it up already,” but it’s this age-old battle that continues, which doesn’t deal well with obsolescence. </p>
  
<br>  
+
<p>[Transcriber note: The background noise level here increases substantially.] </p>
  
<br>  
+
<p>I’m supposed to be applying my comments to ECTC. But comments on IEMT, VSLI, and workshops are interwoven and interchangeable. The business of spawning workshops, the organization of CPMT could be construed to be, in some sense, the ECTC in different places. It’s part of the whole—a combination of workshops and conferences. Workshops do not require a paper. In a way it’s like the bush leagues; you can try out a talk at a workshop and if it’s successful you write it after that and give it at a conference. In a way it’s more cutting edge. It gets the paper that isn’t approved for publication yet. Sometimes you get the paper that’s not quite ready for prime time, but that’s okay too. You get work-in-progress kind of papers. And I say, the IEMT is a conference, but it’s workshop size. And there are other workshops, such as the VLSI in Kyoto that are conference size. So we blur the label once again, and people fight to call it a workshop even though it isn’t. Some ground rules would help there I think. </p>
  
<br><br>
+
[[Category:People and organizations|Balde]] [[Category:Engineers|Balde]] [[Category:Engineering profession|Balde]] [[Category:Professional communication|Balde]] [[Category:Engineering disciplines|Balde]] [[Category:Components, circuits, devices & systems|Balde]] [[Category:Electronic components|Balde]] [[Category:Standardization|Balde]] [[Category:Power, energy & industry applications|Balde]] [[Category:Packaging|Balde]] [[Category:Multichip modules|Balde]] [[Category:Resistors|Balde]] [[Category:Capacitors|Balde]] [[Category:Engineered materials & dielectrics|Balde]] [[Category:Conductivity & superconductivity|Balde]] [[Category:Semiconductor materials|Balde]] [[Category:News|Balde]]

Revision as of 13:50, 13 November 2013

Contents

About Jack Balde

Balde received his Bachelors of Electrical Engineering from RPI. He then worked at Western Electric on airborne radar, underwater sonar, and navigational systems autopilots, hearing aids, and other military subjects up to ca. 1958. From 1956 on he taught at Western Electric’s in-house graduate engineering program, where he taught radar systems designing, engineering statistics, human factors engineering, and other systems design. He then moved to Western Electric’s Engineering Research Center, where he worked on tantalum thin film circuits. He worked with the IEEE to help create standards for package for semiconductors. He mentions his association with the Electronic Components and Technology Conference. He spends extensive time discussing the structure of various conferences, how they aid or resist technological dynamism, and the way association of conferences unifies fields, the creation of new conferences creates new fields.

About the Interview

JACK BALDE: An Interview Conducted by Robert Colburn, IEEE History Center, 2 June 1999

Interview # 348 for the IEEE History Center, The Institute of Electrical and Electronics Engineers, Inc.

Copyright Statement

This manuscript is being made available for research purposes only. All literary rights in the manuscript, including the right to publish, are reserved to the IEEE History Center. No part of the manuscript may be quoted for publication without the written permission of the Director of IEEE History Center.

Request for permission to quote for publication should be addressed to the IEEE History Center Oral History Program, 39 Union Street, New Brunswick, NJ 08901-8538 USA. It should include identification of the specific passages to be quoted, anticipated use of the passages, and identification of the user.

It is recommended that this oral history be cited as follows:

Jack Balde, an oral history conducted in 1999 by Robert Colburn, IEEE History Center, New Brunswick, NJ, USA.

Interview

INTERVIEW: Jack Balde

INTERVIEWER: Robert Colburn

DATE: 2 June 1999

PLACE: San Diego, California

Education and Western Electric career overview

Colburn:

If I could start with your early academic training, schooling and how you got into the field?

Balde:

My background is a Bachelors of Electrical Engineering from RPI. I had worked at Bell Laboratories for a couple of summer jobs, and I went to Western Electric when I came back out of school working on radar and navigational systems autopilot, hearing aids, up until about 1958. I worked on mostly airborne radar and navigational systems, but also some Navy systems as well, including underwater sonar, and a variety of military things.

In about 1956 Tim Shea, vice president of Western Electric Co decided that Western Electric was going to have a graduate engineering training kind of activity very much like M.J. Kelly had done for Bell Laboratories, and I was drafted into being on the academic staff for that. I taught courses in radar systems designing and engineering statistics. (It was obvious that someone who thought statistics was a branch of calculus and didn’t have the faintest idea of how to use it had taught the average college graduate.) And I taught human factors engineering and radar systems engineering and basically system design kind of things, and wrote a couple of textbooks.

Western Electric Engineering Research Center

Balde:

Then I got tagged to go to the Engineering Research Center, which was something a Vice President Tim Shea set up basically to do manufacturing research. Bell Laboratories did not serve the combined Western Electric/Bell Laboratories team well in anything but design, and many of the designs were not manufactured at the lowest cost, by a long shot. So the idea was to do manufacturing research to improve things.

I worked in the area of interconnections, and the most important interconnection that we needed to deal with was of Tantalum thin film circuits, which was very much needed for telephone equipment in order to produce the frequency determining filters for touch tone dialing and things like that. That’s enough background to get you to where I first began going to the ECTC.

Colburn:

Yes.

Balde:

The goal for Tantalum thin films was to make both resistors and capacitors, but not so much inductors. The man who led the activity at Bell Laboratories was Dave McLean. Others working on the team were Bob Barry, John Fisher and others at Allentown, and others including myself at the Princeton Engineering Research Center. I was working at the Engineering Research Center, and of my sixteen-eighteen patents about eleven-twelve of them were in the Tantalum film area, and we did some really pioneering work in Tantalum thin films.

Electronic Components and Technology Conference

Early involvement, IEEE standards task force

Balde:

I started going to the ECTC when it was at the Marriott Twin Bridges. I guess I went as early as 1951 or 1952, and I went pretty consistently, although I missed a couple of years in the late fifties, and then went consistently from the middle sixties on. In the early stages ECTC was only a components conference. By components I mean resistors, capacitors, and maybe inductors and not much else.

By about the mid-‘70s we had had a need to move packages, on printed circuit boards; from the dual inline package, which needs a mass of through-holes, to a surface mount package, the square chip packages were called chip carriers, because other people had used the term chip carrier.

In combination with Dan Amy of Sperry Univac, we proposed to bring some kind of order out of the chaos using an IEEE task force. At that stage there were over 3,500 different chip carriers, most of them within 1/8th to 1/10th of an inch of another per side. The leads were spaced differently, the sizes were spaced differently, and so on and so forth. So we proposed bringing a standardized footprint for all packages of ceramic or plastic. We then organized a task force of some twenty-seven companies, most of them competitors, who agreed to cooperate on the grounds that they had to do something about the chaos. We used the old rule saying that no two sizes should be separated any less than by the square root of two, to give a decent family of sizes. We hammered out what the sizes were and what the lead spacings of fifty millimeters were and the commonality between ceramic with edge clips or ceramic without edge clips, or the molded plastic packages. Molded plastic cases with open lids were proposed by Dimitry Grabbe of AMP and were filled with silicone gel.

There was a slightly different family for ceramic that had forty-millimeter spacing; most PC boards had fifty-millimeter spacing. The reason why this is important is that it represents a major change that happened to the ECTC Conference focus.

After we created this task force and hammered out all the details, it was just a matter of taking it to JEDEC, which is the Joint Electron Device Council, which then goes through an official standardization process. The activity that I was doing was a precursor consensus. Nepcon West introduced five different sizes in 1978, but without any standardization activity.

What was needed was to publicize the results of the task force standardization effort. It was obvious that we had this IEEE task force with a final report to give, and we needed to get the word out about what was going on. At that time I had not been consistently active in the ECTC, and was not in the committee structure for the ECTC. A man named Tom Matcovitch was, and he was a professor at Drexel University and worked in ceramic hybrids, but mostly in ceramic structures for ceramic capacitors for resistance. Remember the ECTC hadn’t gone beyond components then.

Semi-conductors, interconnects, and system design

Balde:

I decided that it was high time that the ECTC dealt with the package for semi-conductors as a component. So I got hold of Matcovitch, who had some stature in ECTC, and he agreed to give the talk, and that represented the first talk in IC packaging at the ECTC. It was funny that when he got asked questions he would turn to me and ask if I wanted to comment, and that was about the same year that Christian Bernard was on the lecture tour with his chauffeur and—

Colburn:

He was the doctor fellow?

Balde:

Yes. After twenty lectures the chauffeur said, “I’ve heard that so often that I could give your talk,” so he said, “Go do it.” He did well on the talk and he got a question and he said, “Oh, that’s so easy, my chauffeur could answer it.” Well, that’s what happened in this thing. After that we were on a roll with respect to the IC packages.

This is a conference that had never done any systems work, and the next thing that happened was that there were people who began to use the ceramic packages or the multichip modules, notably IBM and NEC. Meanwhile the NTT people under Takakki Ohsaki had come up with a ceramic base, with applied layers of organic interconnections. It was the very first MCM-D multi-chip module. Papers from NEC, NTT and IBM appeared at the 1982 ECTC at Seattle. That was the next major turn in the expansion of the scope of ECTC focus, where all of a sudden we had system design sections as well as some of the components.

Now that we had the system thing in place in 1982, and interconnects in general, it became increasingly apparent that the ECTC had to deal with the increasing number of papers being submitted that fell loosely in the category of interconnects. So we had to do something about that. If you are only one session of seven papers and you’ve got fifty-eight, you’re in deep trouble.

Single chip packaging; multi-chip modules

Balde:

I think the first thing we did was to create a section on single chip packaging, and I talked the very reluctant Mike McShane into agreeing to be assistant chair and he’s hung on there ever since. So that worked well. We were loaded with modeling and simulation papers from John Prince and his fellows at the University of Arizona, so once when we had about forty papers and could choose only ten, I made the motion to break away all the modeling simulation papers into a separate session. It was a very selfish motive, because that meant that we could be in control of what was left. In fact, we didn’t even plan it. It was in the session that we were doing the advance paper selection, and it went through before anybody knew what hit them. Sometimes it worked that way. And so it was broken off, the interconnect session from a session in multi-chip modules.

Multi-chip module activity was an interesting excursion. What happened is we began to get thirty to forty papers in multi-chip modules. It was really hot stuff.

Colburn:

When was this?

Balde:

This would have been 1987 or 1988, something like that. When we had only fifteen, we could pick seven good ones, that was fine. One year—I’d have to do the research to find out which year—we had forty-two.

Colburn:

Do you remember what forces were driving that?

Balde:

Oh, the multi-chip module was the way to go to increase the system packaging density. It’s still the major driving force today. They’re the most powerful force in all of the elements of manufacturing technology. It’s still multi-chip modules at the core.

Ceramics

Colburn:

Was there a particular breakthrough that made that suddenly more economical?

Balde:

Yes. The Takakki Ohsaki proposed the breakthrough for getting out of all ceramics and using copper/polyimide layers. It started with NTT. Again we had an IEEE task force which formalized IPC designations of MCM-D, which is the organic material deposited on usually a ceramic substrate. In MCM-C all the layers were ceramic; MCM-L all the layers were laminate multilayer. And if you had organically deposited layers over L as a substrate, you say D over L; and if you put organically deposited layers for the interconnect, you could have D over C. So you have got all these combinations, but they are now official.

The problem with ceramic is that it was very expensive. As implemented by IBM, it was actually thirty-eight to forty-seven layers of ceramic in something three inches square. It cost a bundle, but it was the only way to do the job. If you weren’t an IBM or NEC, you couldn’t possibly afford it. There was some military work by John Bauer at RCA in Moorstown, but those were little ceramic single chip packages, half an inch square or one inch square.

What happened is that though it took four to five years to do it, the breakthrough came out of the organics deposited on the base substrate of printed circuit laminate. There were a number of processes, the most important being IBM Yasu, which made their SLT, which is deposited organic, which can give you higher density because the lines can be much finer than printed circuit board or ceramic. So there was this whole rush of companies getting in the act, keeping up with the Joneses to be on the cutting edge. This resulted in forty-two papers.

Committee structure and session organization

The ECTC has a committee structure which did not turn out to be very effective for this dynamic or any flood of papers on a new subject. A committee is appointed on each of the major subjects. The way in which it is organized is both an asset and a detriment. They do not assemble a group of session chairs on various topics who have expertise in those areas; rather, they have standing committees; one on interconnection, one on polyimide, one on ceramics, and so on. The problem with that structure is that each committee head is concerned with building his own turf, so that if you have one session for the topic of the committee, each committee wants to apply for two, even if no new papers are coming in. Then a push is made to get the necessary papers to fill the extra slots. It makes it very, very difficult to kill anything, because these people are defending the subject for the session, even if there are few papers.

But you also have a situation where there might be a session on multichip modules which use copper and polyimide as the interconnector on the surface. It is hard to assign them to a multichip module session because the polyimide people might ask for the best papers that belong in multichip module, because it’s polyimide dielectric. Well if you think about it logically, there is polyimide in everything these days, which means that the whole conference could be a subset of the polyimide session. And that’s obviously chaos.

The other problem is that, again using polyimide as an example, the twenty multichip module papers are narrowed down to the best seven. Three or four of them are given to interconnect and they are accepted. The paper selection committee meeting resembles a horse trading bazaar, like an open auction or the stock market, “I’ll trade you one of these for two of those,” and so on, and really a free for all with yelling and shouting people calling down to the end of the room to somebody else. It is very much like the stock market. But the problem is that if a guy only has three papers and he needs four more to get a session, he’ll accept anything. He will go to the people who rejected papers and basically pick them up off the floor.

Now some of them were rejected because you had eight good papers and could use only seven, but some of them were actually rejected because you had twenty papers and ten of them were not up to even a minimum standard. Then the papers which were really unacceptable end up in other sessions, which of course weakens the conference. The other consequence of that, (another weakness of the ECTC), is that it takes an extreme amount of courage on the part of the general chair to keep from adding sessions. You don’t want to offend the connector people, for instance, by taking away their precious connector sessions, but since you’ve got more papers on multichip modules than you are able to place, you have two or three sessions on multichip modules.

The Denver Multichip Module Conference was set up with fifteen sessions maximum. There are five different times of day—morning, afternoon; morning, afternoon and morning—and only three concurrent. The reason why three works is there is always one session you don’t care about. Whatever your interests, they are catholic enough to encompass all three, and you shuttle back between the other two and the time.

Iwona Turlik and John Segelken and a couple of others tried to fight it, but before we knew it we had eighteen, twenty, twenty-two, twenty-four, twenty-six, twenty-eight, thirty, thirty-two sessions at the ECTC. If you go to a meeting like this, the most you can possibly get to is ten percent of the conference. And if you, as I do, report to clients what is important in what happened here, it makes my job impossible because I’m only going to see one paper out of ten, at best. So you end up with a conference with 800 people, but 800 people are dispersed between a number of talks. One of those sessions has forty people and a workshop has sixty to eighty. Thus there is a sort of fight going on between the minimalists who want to try to get back to fifteen or eighteen sessions and those who want to continue to expand the number. And that’s the problem.

C.P. Wong has belonged to the “grow it bigger and bigger school,” taking as an image the Materials Research conference in Boston. It does twenty-five simultaneous sessions in five hotels over a mile and a half. I went there once when I was invited to give a talk. Four or five of us were talking about packaging. One guy gave a talk I had heard so many times I could have given the talk myself, so I decided to go out and look to see what was happening elsewhere in the conference. One nearby session was on nuclear waste disposal at sea and the other one was on dental implants. No community of interest there whatsoever. There was one talk on fine line printed circuits, but it was a forty-five-minute walk away for a half-an-hour talk and I was on next, so I couldn’t go to that.

I stayed for two days and there was not a single person I saw on Tuesday that stayed for Wednesday. I checked with other people. John Stafford came and gave a talk on Friday. He flew in at one o’clock, gave his talk at two o’clock, and left on the next flight. That’s how a huge conference runs, as a series of independent workshops.

The academic types who go there only get off to one conference once a year so they want to make the most of it. But the industrial types that go to the ECTC don’t want a week-long conference because their job may not be there after a week if they are not defending their turf. You can be missed for two and a half days and get along very nicely, but four or five days and lose the next weekend is a whole week and it can be a big problem. So you have the C.P. Wongs who would like nothing better than to have twenty-five simultaneous sessions and those of us who say twenty-five sessions with zero people in attendance at each one doesn’t make any sense. You get more speakers, but less non-speaker attendees, because they can’t hear them all anyway, they may as well stay home and buy the proceedings. So you have this kind of interplay going on which really is a big open issue and needs to be resolved. Everything else works well, the organization, hierarchy, and so on and so forth.

Comparison of various conferences, workshops

Colburn:

Has it needed tuning over the years? Have ECTC and the CPMT Society grown well over the years with the needs of industry and researchers?

Balde:

CPMT is doing more at the workshop level. They are in competition, or sometimes cooperation with IMAPS. They never cooperated much with the International Society for Hybrid Macroelectronics (ISHM), an organization doing ceramic hybrids, but they cooperated some with the International Electronic Packaging Society (IEPS), an organization more involved with systems. The systems kind of papers you would get from NEC and Takakki Ohsaki and IBM would have been standard in IEPS.

IEPS and ISHM merged, and with the merger they have become stronger in running workshops. The IEEE has taken to doing that also. A workshop is an interestingly different sort of an event. There are some workshops, like the System Packaging Committee, which operates as if an attendance of 100 is a top limit. There used to be a rule in the IEEE that said that up to 100 you could ban tape recorders and representatives of the press and you didn’t have to write a proceedings. Therefore the talks would be off the record and patent free, so your licensing people couldn’t give you a hard time.

From 100 to 300 were expected to call it a symposium, and generally at a symposium all the workshops are a one-ring circus. A lot of people like them. In the first place, they are informal. You interrupt the speaker right when you find that you don’t understand. If you want to take forty-five minutes with the answers to one guy, you do it and you either make up the time or you go to lunch late, no big deal. Whereas the minute you get to two simultaneous sessions you have the same problems you have at conferences because you have to keep in synch, and therefore they are not quite as free. My bias is that a good conference has three sessions and a terrible conference has six, simultaneously.

At any rate, CPMT began to broaden its activities beyond that of merely sponsoring the ECTC. One of the first things it did was to split off a separate for conference for the manufacturing technology people. They began to want more and more sessions with manufacturing technology at the ECTC and under the old format if you wanted a manufacturing technology session you had to, heaven forbid, get rid of connectors, and so on and so forth. So obviously this was big warfare, and the thing to do was to break that off into a separate conference. That was pioneered when Mauro Walker did a conference in Japan which retroactively got labeled the Manufacturing Technology Conference. It had some elements of manufacturing technology and had some design elements as well, but then over the years as I served on the committee it became apparent to me that if you allowed too much in the way of design input it would be like homogenized milk, in that you can’t tell the difference between that IEMT (International Electronic Manufacturing Technology) and the ECTC.

We had to hold the line. I remember making Jean Joly of Honeywell Ball very unhappy. He was a designated hitter for Honeywell Ball to the Manufacturing Technology Conference and he had a design sort of paper. I told him we didn’t have a place for that; it’s not eligible. He wasn’t very happy about it, because he wasn’t eligible to present their paper at the ECTC because Kurzwell was the designated hitter to present it at the ECTC. But out of that came more of a shaping of IEMT as a Manufacturing Technology Conference.

The IEMT Conferences, which were started by Mauro Walker of Motorola in Japan then began to be followed by additional ones in both the U.S. and Japan. The Manufacturing one in the United States never got off the ground, not over 100. It was really not much more than a big workshop. Bill Bekenbaugh had been active in that one. He oversees one that is done very well: IEMT and Kanagawa and other places in Japan that have 200 or so attendants. I have a staff member who was involved with that to some extent, Kanji Otsuka.

The other kind of activity that came from CPMT was the VLSI workshops. That was started by George Harmon of the National Bureau of Standards and Ron Gedney, who worked at IBM. Harmon was into reliability work, and the problem is that you can’t get people to go to reliability workshops. Reliability workshops have a reputation similar to Sominex: chances of getting through without falling asleep are about zilch. But from his position at the National Bureau of Standards he wanted to have a place where reliability papers could go. And Ron Gedney wanted to have a workshop that would deal with the VLSI packaging, the chip in the package. It wasn’t IC design, it wasn’t system design; it was that middle niche, the interface between step one and step two, very large scale integration packaging. And so they joined hands, Harmon and Gedney, to organize the VLSI conference. I think the first one was in Silver Springs, Maryland, or it might have Research Triangle Park, and out of that they had a couple up in Marlboro and a couple of others. Kanji Otsuka decided to pick that up for Japan, and ran a series of very successful workshops every other year in Kyoto.

Then they picked it up in Europe, running it in France. But meanwhile Elaine Pope of Intel was the one running it in the United States, but she was using a formula where she was not getting enough papers from the U.S. The U.S. core group died in that area. I believe it could be brought back to life if it was done the right way, though I’m not sure it is worth the effort. So you have the VLSI workshop which is almost big enough to be a conference and the IEMT conference which is small enough to be a workshop. And then all the rest of the workshops are kind of ad hoc. There might be one ball grid array for instance held at Binghamton, New York maybe at the offices of SUNY and another one held by other places.

The IEEE Computer Society had an ongoing maverick group called the Computer Packaging Committee, and the Computer Society increasingly became dominated by administrative nitpicking. The computer packaging group, which was moving away from computers and calling it sell system packaging, couldn’t stand the hair splitting of the Computer Society and they came into the ECTC as TC-14. It runs workshops in Japan and in Oiso and Tsukuba and Kyoto, and in the United States in the Poconos, South Padre Island, Santa Barbara, San Diego, Sedona, Las Vegas, Austin, Texas and in Europe in places like Brussels and Cork and the island of Malta. This has been a successful thing and it’s been an operation in the area of System Packaging since 1968, which is basically eight to ten years before we brought the first packaging to ECTC. And many of the people in that committee are the ones who have organized almost everything else, so that if you look carefully you will see that they got their organizational start in this system packaging committee that began in the IEEE Computer Society. They never ran a conference and don’t want anything but workshops; doesn’t want a conference, doesn’t have a newsletter; in other words, a maverick.

And it’s now one of the TC’s, TC-14. CPMT offered us the ability to escape from the rules of the Computer Society, so it was the Computer Society’s loss and CPMT’s gain. The rule that really got in our way was that if you have a group, you have to have a full-fledged election for the next officer, complete with a biography and a photo dossier and two candidates in a competitive elections. Well, when you have a small committee you are darned lucky to get somebody whose arm you can twist into taking the position. You are not about to have a credible election. If I ask you to take the position and you say yes, and then I say you are going to have to run against so-and-so, then you’ll be more likely to say, “Well, if he wants to run, he can have it. Thank you very much and goodbye,” you know. And they created a lot of senseless rules that say you have to have a committee in place with people in place before you book a hotel. If you don’t know what day it is, how can you get anybody to serve?

I’m digressing, but it won’t take me long to digress. One of the worst rules, they had a mailing list which consisted of anybody who filled in a postcard to be on the mailing list. Their mailing list was 650 people, and that committee ignores it. And they said, “How dare you ignore it? That’s a designated mailing list.” So we said we found that it’s a waste of time. The people from Bangladesh and Cuba and Pakistan don’t come. So why should we waste our time and effort notifying them? They said, “Well, you’ve got to do it.” So we took that 650-name mailing list and found only 65 people who had ever come to a meeting. And of those 65 people, 38 were dead or retired. And the remaining ones were on our mailing list. We had an independent main mailing list. At any rate, Mickey Mouse rules. At the CPMT it’s a little more freewheeling, and that’s been important.

Anyway, you really have only these two other activities other than minor workshops, but now this TC-14 runs some of its workshops in combination with the Computer Society which refused to give it up. They were killing it, but they refused to give it up. And so it is jointly sponsored between the major players. And we are going to see more of the co-sponsoring operations. The materials workshop at Ojai, and there are a couple of others that have worked very well.[some continuity lost as tape was turned over at this point...]

Multi-chip modules sessions at ECTC; IEEE involvement

Balde:

One continuing problem is the competition between scheduled sessions and poster sessions. Well, poster sessions are not the greatest thing since sliced bread. It’s pretty bad. It means that you don’t really give the paper and nobody is going to get to hear it, only the people that wander into that room are going to see it. If you are going to get it published, you’d be better off giving it to another conference and get it published.

So when we received forty-two papers on multichip modules, I took twenty-five of them and we did multichip module sessions at Nepcon West, the first time three sessions and the next time five sessions, and it was going to be ten sessions on multichip modules. The IEPS appointed someone as the official liaison with Nepcon West, a man out of Rolling Springs, Texas who was not familiar with multichip modules, and who said he didn’t want the position. By this time I had ten sessions with the chairs named and all the speakers selected, and he said, “Go fly a kite” and threw it in the wastebasket.

So Harry Charles of ISHM said, “Would you like to move into a conference slot? We’ve got a hotel reserved in Denver to put on a multichip module conference.” And I said only under three conditions: (1) No more than three papers simultaneous, (2) no standing committees, you have a committee of the whole that you divide up to become the sessions chairs, fifty percent of them rotate every year whether you like it or not, and condition (3) co-sponsored by ISHM, CPMT and IEPS, the main reason being to head them off at the pass. Otherwise they will each produce concurrent things in competition.

And so that worked for a number of years. IEEE’s participation was weak in the start, and not adequate, for reasons that had not to do with intent but with accidents of time. Les Fox was moving out of packaging work at DEC, DEC wasn’t supporting things, people couldn’t go to meetings, and so on and so forth, and between Les Fox moving out and someone else moving in, the ball got dropped, and they did not supply their end and didn’t help with the program. And this was to the point that giving them one-third of the monetary proceeds when they didn’t do anything was not going to fly. So what has happened is that became a strong conference, which means that the MCM activity coming in to ECTC has now weakened, and in fact no longer has a committee

Conference and workshop diversification

Balde:

Not every conference should deal with every issue. If you do, then conferences become homogenized. Workshops are a completely different breed of animal. You can have a workshop on, let’s say, solder connection, and it really is completely independent from the workshop on ceramic substrates, which is clearly independent from the one about the printed circuit boards and fine line or maybe flex. In other words, there is almost an unlimited demand for workshops, because the guy who goes to the one workshop doesn’t want to go to the other one anyway. But at conferences, we haven’t gotten that idea across as well that you should have a conference that’s clearly different from anybody else’s conference. We have to resist a tendency for each conference to be all things to all people.

One of the things that bothers me is the fact that they want to bring in semiconductor design into the ECTC. Those are perfectly good IC components. Semiconductor design and manufacturing doesn’t belong. The goal of twenty-five simultaneous sessions is not a wise one. I mean, why go? You may as well stay home. The guy who goes to give a talk arrives about a half an hour before the talk, gives his talk, and leaves, and there are twenty-five people in the audience. The smaller the number of sessions there are, the higher the attendance in each session. So that’s one of the major current issues, with the business of the standing committee with its activity of perpetuating dead issues. The connector people didn’t want to give it up even when they only had two papers being submitted, and there is another who doesn’t want to give up multichip modules even though he has only two papers, because he wants to beat the bushes. He has a committee that says, “Give it up already,” but it’s this age-old battle that continues, which doesn’t deal well with obsolescence.

[Transcriber note: The background noise level here increases substantially.]

I’m supposed to be applying my comments to ECTC. But comments on IEMT, VSLI, and workshops are interwoven and interchangeable. The business of spawning workshops, the organization of CPMT could be construed to be, in some sense, the ECTC in different places. It’s part of the whole—a combination of workshops and conferences. Workshops do not require a paper. In a way it’s like the bush leagues; you can try out a talk at a workshop and if it’s successful you write it after that and give it at a conference. In a way it’s more cutting edge. It gets the paper that isn’t approved for publication yet. Sometimes you get the paper that’s not quite ready for prime time, but that’s okay too. You get work-in-progress kind of papers. And I say, the IEMT is a conference, but it’s workshop size. And there are other workshops, such as the VLSI in Kyoto that are conference size. So we blur the label once again, and people fight to call it a workshop even though it isn’t. Some ground rules would help there I think.