Interview with Dick Somes, PICMG’s long-time VP of Technology
By Lance A. Leventhal, ATCA Newsletter Technical Editor.

What would you cite as the most important accomplishments of your tenure as PICMG’s VP of Technology?

My first accomplishment, back in 1994, was the formalization of PICMG’s Policies and Procedures. The CompactPCI spec had just been revised, but only a few people had been involved. Clearly, we needed to guarantee access to both development and review of specifications to all members. Our basic principle has always been to define a way for doing high-quality technical work in a consensus oriented environment. We just adopted the third revision of the Policies and Procedures, so in a sense it represents my three most important accomplishments.

In fact, I can’t claim principal credit for the other things I’ve done. I believe I made an important contribution to the development of our intellectual property policy, formally adopted four years ago and updated earlier this year. Other major players in that effort were our President, Joe Pavlat, and our attorney, Andy Updegrove.

My most significant contribution to an individual spec was the first revision of AdvancedTCA. I was officially the editor for that revision, and I have fond (!) memories of leading day long page by page reviews.

What were the biggest disappointments of your tenure?

Several specifications simply stalled and were never completed, including a couple I led myself. I learned that strong leadership driving the activity forward is as important as consensus over the content in open specification development. We now have clear criteria for inactivity and an explicit process for dealing with it. We thus get earlier warnings of problems, and a trigger to take corrective action.

Another major disappointment is truly painful. It is the soured relationship between PICMG and VITA. I was chairman of the VITA Standards Organization the year before I was elected to my current PICMG job, and continued to be active there for years afterward. The goals and methods of the two organizations are very similar. An old saying is that imitation is the sincerest form of flattery, and I must acknowledge that PICMG owes a great deal to VITA as a model.

VITA has adapted the AdvancedTCA platform management architecture in some specifications, and at one point was also adapting the AMC spec for Eurocard carriers. There’s a lot we could be doing together, and I’m still hopeful that relations between the two groups will improve. We have many common members who can force the issue.

What was the biggest surprise of your tenure?

Well, I certainly did not foresee the growth of PICMG in terms of the number of specs or the range of technologies we would be supporting. We started out with a very narrow range of interests, namely PCI/ISA based passive backplane industrial computers using standard PCI SIG peripheral boards. CompactPCI was the first step in broadening our interests, defining a new form factor and increasing the slot count for PCI platforms.

CompactPCI grew to encompass a range of telecom applications, and reduce the role of the PCI bus.

Obviously, the transition to switched serial interconnects on AdvancedTCA and MicroTCA was another big change.

We have software and protocol oriented specs as well now.

It was all incremental, and no step was particularly startling when it happened. After all, we’ve always been a bottoms up organization whose course is charted more by what our members want to do than by initiatives defined by the leadership. Looking back, though, it is surprising how it all turned out.

How would you characterize (in just a few words) the current status of the following PICMG specs:

CompactPCI: Mature in most areas, but still evolving in others. The PICMG 2.16 Gigabit Ethernet platform is probably the dominant variant, but there are new PCI Express variants under development that promise continued growth.

AdvancedTCA: With Revision 3.0 on the street, it’s a pretty complete body of work with just a few niceties to be added. Its centerpiece is its platform management architecture, which really has more general application.

MicroTCA: Still evolving, especially in terms of environmental hardening. It probably covers the broadest range of applications of any architecture we have developed.

What would you say are the most important things that need to be done with regard to the AdvancedTCA and MicroTCA specifications?

Consolidate the platform management architectures into a single coherent specification. This is among the most important work PICMG has done, with much credit due to Mark Overgaard and Pigeon Point Systems for their leadership.

Define minimum cost platforms in each family. For MicroTCA in particular, there is a need for a “medium availability” platform without full redundancy but enhanced maintainability.

Limit the proliferation of PICMG defined PCI Express platforms to a complementary set.

What are the biggest challenges you see facing AdvancedTCA and MicroTCA?

That’s almost the same question. I would re-iterate the importance of cost. It seems that standards based platforms are inherently more expensive than those optimized for a specific purpose. We need a better understanding of life cycle costs so as to appreciate when features we don’t really need in a standards based platform are cheaper than the non-recurring cost to eliminate them.

I think we should invest in a better understanding of the positioning of the PICMG architectures relative to one another and to outside architectures, proprietary and standard.

I think we should invest more in understanding the environmental capabilities of our architectures, especially AMC and MicroTCA, and what we can do to enhance their properties.

Which markets do you think are especially promising for AdvancedTCA and MicroTCA?

AdvancedTCA has a place in the data center if we deal with the power distribution issue. An AdvancedTCA platform that integrates AC input capability would be a major step in that direction.

MicroTCA has a place in transportation, industrial automation, and the test and measurement markets.

What are the most interesting (or most unusual) applications you have seen for AdvancedTCA and MicroTCA?

The most unusual use I’ve seen of AdvancedTCA is the undersea monitoring application reported by SAIC at last year’s AdvancedTCA Summit.

The prospect of using AMCs and MicroTCA in a ruggedized form in tactical vehicles and even aircraft is interesting, and somewhat unexpected. There’s still feasibility work to be done, obviously.

What advice would you give to your successor at PICMG?

Look ahead. The process I’ve left you is obviously a response to past issues. That is both a virtue and a vice. It guarantees access, but sometimes does not deliver critical mass. It confirms PICMG as the bottom-up organization we started out to be, one in which any three Executive Members can start something. However, it does not necessarily provide the top down guidance sometimes needed to prevent proliferation of competing specifications or identify new directions.

Don’t be too quick to change the status quo, but don’t be afraid to do so either. Try to internalize the current process so you can understand how all its elements fit together. You will need to guide subcommittee chairs and potential sponsors, and that is easier if you really understand the principles on which the process is based. No one else in PICMG really has an incentive to do that. It will also help you understand when something is out of place, or a situation is not explicitly covered.

If you’re the employee of a member company, as I was for 10 of my 12 years as PICMG Technical Officer, be aware of the tensions between your corporate and PICMG obligations and keep them in balance. I actually kept two hats in my office as a reminder.

How would you describe the current state of the embedded systems business to a top manager who had one minute to devote to the subject? What is the most important point you would want to make?

I would not even try to describe the embedded systems business to a top manager in a minute. It would simply be a waste of everyone’s time. There is no pure and compelling argument for companies to engage in the business. The investments are high, the volumes are relatively low, the adoption ramps are long, and the cost pressures are greater than ever. Big companies have always been impatient with the embedded systems segments of their businesses and ready to divest them when the opportunity offered itself.

To succeed in the embedded systems business, I think you must love it for what it is, which is the opportunity to make commercially available technology do things it wasn’t designed to do. It is the place where you make computers, which would just as soon do things according to their schedule, interact with the real world. It’s the place where you get to deal with new challenges, whether in the way a military force handles threats or in the way an automobile operates on a fuel other than gasoline.

That’s what has fascinated me since the first time I tried to make a computer controlled measurement of transmitter frequency or receiver sensitivity.

There, now you know what you may have suspected. I’m a romantic of sorts. I believe the heart leads and the head follows in any worthy enterprise.

What are your future plans?

I’ve been living in Arizona for almost five years in an age 55+ “active adult community”. My wife and I are having a great time here. This is a wonderful place, with many mentally and physically active people. They all come from somewhere else and are eager to make friends and have a good time together. We’ve become heavily involved in several recreational activities. My wife and I, for instance, have rediscovered contract bridge after 25+ years of almost total inactivity. We expect to widen our range of activities when I don’t have any more professional commitments.

We have 27 holes of golf in our section of the community, and 36 more in the community as a whole. I’m hoping to sharpen my game, which has plenty of room for improvement. Keeping the ball on the short grass is one challenge, and developing better touch around the greens is another.

We’ve been traveling a fair amount, but always with the laptop packed and daily looks at my email. I’m looking forward to traveling without the need to do that. We’re headed for Hawaii for four weeks early next year. That’s the longest trip we’ve taken. I’d like to do a cross-country road trip one summer as well.

There are also technical interests I’d like to pursue. Some areas that interest me, such as speech processing, date back to graduate school. I’m not sure the techniques I was interested in then apply to modern problems, but they have the feel of unfinished business that I’d like to wrap up. At least, the computational tools needed are easier and cheaper to come by than they were 40 years ago.

I’d also like to do some writing and video production. First and foremost I want to capture what I know about my family history and my wife’s for our children and grandchildren. There are things we both wish we had asked our folks about, but never did. All but one of them are gone now. I’d like to anticipate and answer questions the next generation may someday wish they had asked.

I don’t think I’ll be bored.

Commentary

Dick Somes has been Vice-President of Technology for PICMG since 1994. Upon his retirement this year, it is surely time for all of us to thank him for his commitment to high-quality specifications, his insistence on providing the level of detail needed for actual implementations, and his energy in moving activities ahead when they appeared to be moribund if not dead. Everyone in the embedded systems business owes him a tremendous debt for his efforts, commitment, and dedication.

The only thing that really concerns me here, besides the fact that I should have asked him better questions, is that he and I are exact contemporaries. We graduated from college in the same year. I therefore see no reason why he should retire and enjoy life as he apparently plans! Seriously, his long record is why we have given him a special Industry Technology Leadership award at this year’s AdvancedTCA Summit.