3 Comments

Summary:

D-Wave Systems is moving closer all the time to commercializing its cloud quantum computing service. In the process, it is demonstrating just how fast quantum computing science is developing.

Vern Brownell, President and CEO, D-Wave Systems, speaks at Structure Data 2014.

Classical computers are only as good as the number of transistors they have. According to Moore’s Law, that number doubles every two years.

That pace may soon slow, but there’s an alternative: quantum computing, which relies not on transistors but on particles to perform calculations much more quickly. Speaking at the Gigaom Structure Data conference Wednesday, D-Wave Systems president and CEO Vern Brownell said that many of the major computing breakthroughs over the next decade will be dedicated to quantum computing, and computers will improve much more rapidly than the world is used to seeing.

“You will see many orders of magnitude improvement on each generation, rather than the 2X or 5X that we typically see in classical computing,” Brownell said.

D-Wave has sold computing systems to Google and Lockheed Martin, but it is not yet ready to fully commercialize its computing services. Eventually, it would like to rent cloud cycles to anyone.

To get there, it will have to make working with quantum computers much easier. Brownell equated D-Wave’s current progress to Intel in the 1970s, when the company had just developed a microprocessor. You still need to program to perform any kind of task.

D-Wave is also continuously improving the speed of its computer. Brownell said it follows what the company likes to call “Rose’s Law,” after its founder: every two years, it will quadruple the number of qubits (particles coaxed into a specific quantum state) its computers run on, “more or less like clockwork.” Like transistors, more qubits equals more power. And D-Wave has already progressed from a computer with 128 qubits to an upcoming machine with 1,000.

 

Structure Data Ticker

Photo by Jakub Mosur

  1. I wonder how many years until Moore’s law comes to a complete end. He predicted this back in 1960″s while working at Intel. However year each intel find ways to upgrade their chipset by at least 2X the speed.

    Share
  2. John Sellers Friday, March 21, 2014

    You will notice he is actually saying and not saying.

    If one can build on Qbit, then you don’t have to do much to put 128 or 1000 in the same box. But that doesn’t make them work.

    He said nothing about an entanglement of 128 or 1000 Qbits in a single entanglement. This lapse is telling., because D-Wave would be shouting it from the roofs if they had actually done an entanglement of 128 or 1000 Qbits.

    You see if you entangle 128 Qbits you could do about 3.4 times 10 raised to the 38 power calculations at the same time.

    If you entangle 1000 Qbits you could do about 10 raised to the 301 power calculations at the same time.

    I am positive that D-Wave has never done this, and I think that they never will.

    Share
  3. John Sellers, what the heck are you talking about. 128 qubits is 2 raise to the power of 128, and 1024 is 2 raise to 1024. Heres a tip, get a math textbook and try to learn counting in binaries. And its not Qbit its qubits.

    Dwave has evidence for large scale entanglements. The processor was not designed for full entanglement but then no one expects that even from other experiments or scientists. Full entanglement and its performance is theoritical. In reality you can only develop a machine to be as close as possible to the theoritical limit, and theres strong indication that Dwave is achieving towards that.

    Share

Comments have been disabled for this post