Pages

Categories

Search

 

Context on D-Wave Systems from President and CEO Vern Brownell

by
May 12, 2014
CTOvision
No Comment

By Bob Gourley

I recently caught up with D-Wave Systems President and CEO Vern Brownell and asked him to provide a bit more context for our readers on what this very unique system is and does.

Vern has a very interesting background that instantly grabbed my attention. He has terrific past performance leading a world class enterprise IT company (Egenera, he was the founder at CEO). He was also the Chief Technology Officer at Goldman Sachs.

Having CTO roots and experience founding and running a very successful firm seemed to have positioned Vern well for leadership at D-Wave. His business savvy and ability to understand both enterprise mission needs and technologies enable him to think through the priority use cases for the D-Wave solutions. I also found he had an amazing ability to talk both technology and business, which is also something I would expect a CEO with CTO roots to excel at.

Now here is more on D-Wave:

  • The company D-Wave was founded in 1999 (and the science it is built upon has been well known since the 1990′s). Vern’s focus as CEO has been to commercialize what was known to be a powerful idea that worked in labs. Clearly he has done that.
  • The D-Wave computer (current version is known as the D-Wave Two) exploits quantum mechanical effects. The system is designed around qubits rather than bits.
  • Leveraging these laws of quantum mechanics, some of the most fundamental laws of the universe. The promise of this is an unprecedented speedup over classical computing. Key problems they are focused on are the complex problems that classical computing is not able to address well.
  • Doing this has required significant advancements in science. The D-Wave Two operates at an extreme environment where all known environmentals are controlled and accounted for and external influences are minimized. The compute structure itself is supercooled.
  • This very powerful compute platform is had to be engineered to be efficient and to operate cold (very cold, like -273 degrees Celsius, which is just about as close to the theoretical position of absolute zero as you can get (.02 degrees above absolute zero, in case you were wondering).
  • The processor is shielded from magnetic effects. It is also in a very high vacuum, like a bigger vacuum than that found in our solar system. The pressure over the chip is 10 billion times lower than atmospheric pressure.
  • It is interesting that because of this design, the power consumed by this device is so low that it saves energy, especially when compared to modern supercomputers. The total power consumption is about 100 times less than a supercomputer that might be tasked with similar problems.
  • Although the D-Wave approach has been to leverage the power of some very advanced science, the team consciously decided design it in ways conducive to manufacturing of real, fieldable, operable systems. From day one of the company they have focused on commercialization of the technology and that means thinking efficient and effective (this is certainly not a commodity piece of hardware, but the principles of good commodity IT were applied where possible).
  • The current use of the system is on optimization problems. A user models a problem and then the Dwave-Two optimizes for the best solution or solutions. The processor considers all possibilities simultaneously to determine the optimal solution.
  • Applications for the solution are what we have long called “NP-hard” problems (if you don’t have a computer science or decision-theory background you can think of these as really big ugly problems that traditional math and computers have a very hard time solving). One NP-hard we all learned in school is the old “traveling salesman problem” where you try to optimize a route a salesman takes. That old problem is a very simple way to understand how hard a problem can get as the inputs to the problem gets larger.  Optimization can apply to use cases in healthcare, medical research, finance (monte carlo simulations of risk, for example). Other use cases center around aviation, avionics and design of platforms for air and space. Optimization is also critically important for air traffic control and flight schedules (and most of us in the traveling public would love to see improvements there).
  • Programming the computer takes skill. But documentation and concepts and plenty of research is there to assist any programmer. D-Wave provides services around their solution too, and several partners are working with them to build out the delivery ecosystem (including Lockheed Martin, for example).  For insights into software see http://www.dwavesys.com/software

Concluding thoughts: Like any great CEO or seasoned CTO Vern knows that the technology is not what this is all about. It is about the “so what” of the technology.  I believe the next thing we will see from D-Wave will be more published use cases and concepts for applications. That will help fuel demand for the capability and grow the market, and perhaps accelerate solutions that serve us all better. 

For more see the website of D-Wave Systems

CTOvision Pro readers can find more here: