Featured Engineer

Interview with Mike Steinberger

Mike Steinberger

Mike Steinberger - Lead Architect, Serial Channel Products, SiSoft

What are your favorite hardware tools that you use?

Although I’ve spent a lot of time in the lab over the years, I don’t have access to a lab right now.

My two favorite pieces of equipment are a sampling oscilloscope and a VNA. A well equipped sampling scope gives me TDR, TDT, eye diagrams and data pattern waveforms. I get a lot of good information from each of these outputs. The VNA is by far the most reliable way to measure passive components, and I depend on those measurements to drive the development of our electromagnetic models.

What are your favorite software tools that you use?

Valgrind. This program has literally added months to my life. It works exactly as advertised, straight out of the box, with no fiddling around or false steps. That also makes it one of the highest quality pieces of software I’ve ever encountered.

Numerical Recipes in C++. This is the classic book and software for numerical processing. The treatment of the subject is rigorous and yet readable and practical. The algorithms are robust and bug-free (although I do have a quibble with the sign in the exponent in their definition of the Fourier transform). If the numerical technique isn’t in Numerical Recipes, you don’t need it.

What is the hardest/trickiest bug you have ever fixed?

The most astonishing debugging I ever saw was when I first joined Cray Inc. I was only peripherally involved, but through system level experiments and exceptionally disciplined experimental technique, we isolated one wire on an IC than ran parallel to two other wires for such a long distance that it generated sporadic errors.

In what I do now, the hardest debugging I do is on our circuit solver. The algorithms aren’t all that complex, but they’re very subtle. It takes me a long time to get my head into that code, and a lot of rigorous thinking to make sure the fix doesn’t break something.

What is on your bookshelf?

There are certain classics on my bookshelf that I either refer to on a regular basis or else have learned a great deal from:

  • Gold and Rader, Digital Processing of Signals
  • H.W. Ott, Noise Reduction Techniques in Electronic Systems, 2nd ed.
  • Bell Laboratories, Transmission Systems for Communications
  • Matthei, Young and Jones, Microwave filters, Impedance-Matching Networks, and Coupling Structures
  • Ramo, Whinnery and Van Duzer, Fields and Waves in Communication Electronics, 3rd ed.
  • Press, Teukolsky, Vetterling and Flannery, Numerical Recipes in C++
  • Feynman, Leighton and Sands, The Feynman Lectures on Physics
  • Wozencraft and Jacobs, Principles of Communication Engineering
  • Wayland, Differential Equations Applied in Science and Engineering
  • Ellis and Stroustrup, The Annotated C++ Reference Manual
  • Lippman, C++ Primer
  • Bentley, Programming Pearls
  • Meyers, Effective C++,
  • Aho, Sethi and Ullman, Compilers: Principles, Techniques and Tools
  • Papoulis, Probability, Random Variables, and Stochastic Processes
  • Luenberger, Optimization by Vector Space Methods
  • MacLane and Birkhoff, Algebra
Do you have any tricks up your sleeve? (special way to analyze circuits, special process you use to make something, etc.)

Well, yeah. Every illusionist does. (Pay no attention to the man behind the curtain.)

What has been your favorite project?

One of my more satisfying projects recently has been the simple via experiment we performed in 2008 and published in 2009. With the exception of the need for a VNA, the technology was low tech- experimental samples I made in my shop, closed form solutions to Maxwell’s equations. And yet the match between measurement and theory was excellent and we were able to prove a lot about the physics of vias. There’s still a lot more knowledge we can extract from that data, and I hope to do a lot more of that analysis this year.

(Repeat after me in the worst Mexican accent you can produce: “3D field solver? We don’t got no 3D field solver. We don’t need no stinking 3D field solver.”)

Do you have any note-worthy engineering experiences? (blowing up things, getting shocked, etc.)

My best stories are not ones I was involved in directly, and not ones I could share publicly.

The only story that’s even sort of interesting is too self-congratulatory to include in an interview, but I’ll let you know what it is. I began my career at Hughes Aircraft Co., and one day my supervisor asked me whether I could etch a 3dB interdigitated directional coupler on a ceramic substrate. Using rather crude equipment (Polaroid camera with manual massaging during development, for example), I produced the unit, and the first copy had exactly the right coupling. The lab manager wanted to shake my hand for that, and all I was thinking about were the next ten copies. I didn’t realize it couldn’t be done, so I did it.

What are you currently working on?

I’m working on a lot of different things, but the research I’m most excited about is to improve the depth of our understanding of the electromagnetics of passive interconnects. Through a combination of measured data and equation-based analysis, I think we can get the kind of engineering insight that will be required to competently design 25 Gb/s serial links.

What direction do you see your business heading in the next few years?

I think the theme will be 25 Gb/s interconnects, and I expect us to be deeply involved in supplying the tools needed to do competent engineering at these data rates.

What challenges do you foresee in our industry?

I don’t claim to see any more than is already obvious to a lot of people. The fact of the matter is that we truly are getting to the end of Moore’s Law. I’ve seen at least one excellent talk on this topic. What we’re now seeing is that we have to put in a lot more engineering to maintain the current level of progress.

  • For mobile devices, the challenge is to get more functionality for a given power consumption. I’m not directly involved in this, so I don’t know much more about it.
  • For stationary devices (desktop PCs, data routing, high performance computing, …) the fundamental limitation is the data rate of the interconnect. To my certain knowledge, that’s been true for at least the last 15 years. The trouble is, physical subassemblies aren’t getting any smaller, materials and interconnect structures are improving marginally, and transistors aren’t getting that much faster. The net result is that every single element of a data path between ICs has to be fully optimized and integrated into an end to end solution. This, in turn, requires that many disciplines must be seamlessly integrated into a comprehensive capability. That’s what makes my job fun.
Visit Michael's EEWeb Profile
Previous Next

Previous Spotlights

Featured Products

ad1

TechXchange General Community

Technical questions? Find your answers here! Digi-Key experts available 24-7!

ad1

Maxim Control and Automation

Broad offering of high-performance analog, robust communications and innovative application-specific ICs

ad1

Honeywell MICRO SWITCH V15W

Designed for harsh-duty wash down areas, the V15W series of basic switch offers rugged, precise on/off control

sidebar
 
Click Here