Re:New Google lab working on quantum computing hardware
by gaze on Thursday September 04, 2014 @09:02AM
(#8267943)
"I hope Martinis considers NV (nitrogen vacancy) centers in synthetic diamond as quantum computing elements." He's built his career around superconducting circuits. Switching to optics would put him outside his area of expertise, so it's unlikely.
Re:New Google lab working on quantum computing hardware
by bkcooper on Wednesday September 03, 2014 @11:03PM
(#8266561)
I'm pretty intrigued by this. Martinis is very good, and I already thought it was interesting when I saw earlier today that Google had picked him up. That he will be trying to improve D-Wave style designs (by dropping in transmons, from the sound of this article) is not at all what I would have guessed.
I think that will be fairly tough. It's been a while since I looked at D-Wave, but aside from the whole "is it quantum" question about their stuff, their fabrication always seemed like a real strength. They had tons of on-chip circuits for addressing all of the qubits, and building devices with that many Josephson junctions that work is not easy. The comment that some of their material choices are probably limiting coherence is almost certainly true (this was one of Martinis's big contributions from ~10 years ago). But translating processes of that complexity to cleaner setups seems like it's going to require some real fabrication fu. I certainly wouldn't put it past Martinis, but it's ambitious for sure.
Re:New Google lab working on quantum computing hardware
by archgoon on Thursday September 04, 2014 @01:46AM
(#8266953)
Millikelvin _is_ quite cold. 2 mK is the limit of dilution refrigeration technologies (achieved by mixing the liquids Helium-3 and Helium-4), so 20 mK is about the best you can do continuously. Since they want to operate a device for an unspecified amount of time, this is probably what they're going with.
(that's the paper that summarizes the results that won the 2001 Nobel Prize in Physics)
However, in 2003, the coldest temperature of a full system was achieved when a group at MIT cooled a BEC down to 500 picoKelvin (half a billionth of a Kelvin).
That said, if you're willing to cut some hairs and only count _part_ of a thermal system (say, the nuclear spins of rhodium atoms), then you can count Tauno Knuuttila's achievement of 100 picokelvin.
Re:New Google lab working on quantum computing hardware
by vidarh on Thursday September 04, 2014 @04:54AM
(#8267272)
The article does not claim that "someone's vaporware" is better than a shipping system. It states that they believe there may be opportunities in doing things differently. With no promises that they'll manage to do better. There's nothing of substance there for D-Wave to respond to. Of course there are opportunities in doing things differently.
There's considerable work that indicates diamond NV centers could be valuable in the next step beyond D-Wave's systems. A few of many:
Photonic Architecture for Scalable Quantum Information Processing in Diamond http://arxiv.org/abs/1309.4277
Quantum Correlation Between Distant Diamonds http://www.sciencemag.org/content/334/6060/1213.summary
Quantum computing: Diamond and Silicon Converge http://211.144.68.84:9998/91keshi/Public/File/34/479-7371/pd...
Homoepitaxial Growth of Single Crystal Diamond Membranes for Quantum Information Processing http://dash.harvard.edu/bitstream/handle/1/11859326/Homoepit...
I think that will be fairly tough. It's been a while since I looked at D-Wave, but aside from the whole "is it quantum" question about their stuff, their fabrication always seemed like a real strength. They had tons of on-chip circuits for addressing all of the qubits, and building devices with that many Josephson junctions that work is not easy. The comment that some of their material choices are probably limiting coherence is almost certainly true (this was one of Martinis's big contributions from ~10 years ago). But translating processes of that complexity to cleaner setups seems like it's going to require some real fabrication fu. I certainly wouldn't put it past Martinis, but it's ambitious for sure.
Martinis makes his qubits from aluminum circuits built on sapphire wafers and chills them to 20 millikelvin
I wasn't even aware that "millikelvin" was a thing.
http://en.wikipedia.org/wiki/Dilution_refrigerator
That said, if you're willing to consider non-continuous cooling methods such as laser cooling, then you can get quite a bit colder.
Back in 1995, the first Bose Einstein Condensate was created in a laboratory by achieving a temperature at about 1 microkelvin.
http://cua.mit.edu/ketterle_group/Projects_1995/Pub_95/davi9...
(that's the paper that summarizes the results that won the 2001 Nobel Prize in Physics)
However, in 2003, the coldest temperature of a full system was achieved when a group at MIT cooled a BEC down to 500 picoKelvin (half a billionth of a Kelvin).
http://www.sciencemag.org/content/301/5639/1513.abstract
That said, if you're willing to cut some hairs and only count _part_ of a thermal system (say, the nuclear spins of rhodium atoms), then you can count Tauno Knuuttila's achievement of 100 picokelvin.
http://lib.tkk.fi/Diss/2000/isbn9512252147/
<yawn>
The article seemed to be very unbalanced. Normally you would expect at least a quote from the Dwave side.