Short History

This is me and my father in 1959. My father was a civil engineer and probably worked on the design of this bridge. In this picture my father shows his newborn son a bridge under construction. He is apparently too excited to close the car door (Volkswagen bug in background). I suppose his goal was to acclimate his son to a future role as an engineer. It worked, as I am an engineer now. 

This is a 1967 photo of my mother, brother Mach, brother Lance, and me. My father is in the previous picture, and my sister Helaine is not shown in this album. The photo is at Primrose campground, 17 miles north of Seward Alaska. 

This is a semi-automatic wire wrap machine at Caltech that started a sequence of events that has formed about half the activity my career. In the fall of 1981, I got this idea to create a parallel computer of a new design and called at the time a “homogeneous machine.” The red blob slightly right of the center of the picture is wire on a wire-wrap circuit board. The machine has a moveable pointer to tell the operator where to put wires. 

Four circuit boards wired with the machine from the previous picture formed a four-processor parallel computer. This computer was programmed for Quantum ChromoDynamics (QCD) calculations by Eugene Brooks and we wrote a paper. 

The design worked so we made a printed circuit board for the computer and a 64 processor chassis. This system was later renamed the “Cosmic Cube” after I left Caltech. 

This picture shows the “hypercube” wiring on the back. If you look, you can see horizontal wire bundles span 2, 4, 8, etc. card slots. Words like Cosmic Cube, Hypercube, and quantum have popular appeal that I’ve learned is important to making things successful! 

The Cosmic Cube became the basis for a whole series of progressively larger computers. This series continues to the present day with the world’s #1 and #2 most powerful supercomputers — the Livermore/IBM Blue Gene and Sandia Red Storm. These supercomputers are of the same design as the one shown in earlier pictures but 100,000,000 times more powerful (approaching a Petaflops or 10^15 floating point operations per second) and costing around $100M. I worked on some of these machines over the years, including the Red Storm system shown here. 

So what do you do after having a role in the biggest supercomputers in the world? As careers move along, people seem to go from actually doing stuff to attending meetings. So I’ve set up a workshop series (with Thomas Sterling and Horst Simon) on Zettaflops (that would be 10^21 floating point operations per second). Here I am running the workshop in 2005. Another such workshop will take place in October 2007. 

For some diversification, I’ve started to “work” with quantum computing. “Work” in this case referring to attending meetings, writing proposals, and so forth. I took the picture in Australia of an experimental system for creating quantum computers atom-by-atom. The lady is a colleague of mine (whose face has been obscured intentionally) who constructed the apparatus. For those unfamiliar with quantum computers, much of the current experimental apparatus looks like stainless steel pipe. The actual devices are inside and often require high vacuum and temperatures near absolute zero. What you see is the container for creating that sort of environment. 

Back to regular life, this is our family: Erika, me, and Beverly at Exit Glacier in Alaska in 2007. This was one of those self-timer photos. I clicked the camera and then ran really fast to get into the picture before the camera went off.

— with Erika Alden DeBenedictis.