Microsoft's Gooseberry is a dish best served really, really cold: Progress made on silicon quantum computing
- Reference: 1611904054
- News link: https://www.theregister.co.uk/2021/01/29/microsofts_gooseberry_is_a_dish/
- Source link:
Complementary metal–oxide–semiconductor (CMOS) technology – an integrated circuit manufacturing process – is used to make a variety of computer components. And many of the scientists developing quantum computers would prefer to use this familiar approach to build machines that realize quantum bits, or qubits, instead of relying on more exotic mechanisms that have been explored like liquid-state nuclear-magnetic-resonance or ion traps.
[1]
Qubits are to quantum computers what binary bits are to classical computers, the state measured for computation. Qubits represent the state of a quantum system, which can be determined by measuring subatomic particles like the spin of electrons or the polarization of photons.
The idea of a silicon-based quantum computer dates back to a [2]1998 paper from Bruce Kane, then a researcher at Australia’s University of New South Wales. But developing a practical implementation has proven to be a challenge because, among other things, CMOS components tend to throw off heat and that interferes with qubits. The [3]first programmable quantum processor using silicon was described by researchers in the Netherlands in 2018.
[4]
Useful quantum computers will be impossible without error correction. Good thing these folks are working on it [5]READ MORE
Qubits in such machines depend on storage at near-zero Kelvin temperatures, a feat that has required extensive refrigeration hardware and wiring for each qubit. That becomes a problem when you need millions of them, more or less the number of qubits estimated to be necessary for the sort of general-purpose quantum computer that research scientists hope to build.
In [6]a paper published this week in Nature Electronics, researchers from Microsoft and the University of Sydney, Australia, explain how they've developed a chip called Gooseberry that can support thousands of qubits – significantly more than the 65-qubit design touted by IBM last year – though they haven't actually made such a device.
"The chip powering this platform, called Gooseberry, resolves several issues with I/O in quantum computers by operating at 100 milliKelvin (mK) while dissipating sufficiently low power so that it does not exceed the cooling power of a standard commercially-available research refrigerator at these temperatures," explained Chetan Nayak, general manager of quantum hardware at Microsoft in a [7]blog post . "This sidesteps the otherwise insurmountable challenge of running thousands of wires into a fridge."
[8]
Microsoft's Gooseberry quantum chip ... Click to enlarge
The researchers, led by University of Sydney professor David Reilly, have also devised a cryo-compute core that sits above Gooseberry in the quantum stack to relay information from the quantum layer. The core, a general-purpose CPU, is designed to operate at a higher though still rather chilly temperature – 2 K as compared to 100 mK. According to Nayak it deals with some triggering functions, data handling, and branching decision logic and thus requires more circuit blocks and transistors.
"The core acts as an intermediary between Gooseberry and executable code that can be written by developers, allowing for software-configurable communication between the qubits and the outside world," explains Nayak.
[9]
In the same way, this research represents an intermediary step toward a functional, general purpose quantum device. Nayak concedes other breakthroughs have to happen – like [10]error correction – but he insists Microsoft is in it for the long haul. ®
Get our [11]Tech Resources
[1] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_emergent_tech/front&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=2&c=2YBPq04oFCqb4DNdAdbS@3AAAAAE&t=ct%3Dns%26unitnum%3D2%26raptor%3Dcondor%26pos%3Dtop%26test%3D0
[2] https://www.nature.com/articles/30156
[3] https://www.nature.com/articles/nature25766
[4] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_emergent_tech/front&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33YBPq04oFCqb4DNdAdbS@3AAAAAE&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[5] https://www.theregister.com/2020/12/09/quantum_computing_correction/
[6] https://www.nature.com/articles/s41928-020-00528-y
[7] https://www.microsoft.com/en-us/research/blog/full-stack-ahead-pioneering-quantum-hardware-allows-for-controlling-up-to-thousands-of-qubits-at-cryogenic-temperatures/
[8] https://regmedia.co.uk/2021/01/29/microsoft_gooseberry.jpg
[9] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_emergent_tech/front&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44YBPq04oFCqb4DNdAdbS@3AAAAAE&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[10] https://www.theregister.com/2020/12/09/quantum_computing_correction/
[11] https://whitepapers.theregister.com/
Re: Perhaps the time is right
Seconded. Too often, this quantum stuff has the feeling of the 'don't look behind the curtain'.
Re: Perhaps the time is right
@AMBxx" I've been peeking behind the curtain during lockdown. I got a couple of books on quantum mechanics and I've been following online lectures from MIT* and Prof Susskind**. I think that even the people who look behind the curtain and really understand it (not me) can't explain it in the context of the "real world", other than to say that it is the real world (probably!).
Take tunnelling (the thing that makes tunnel diodes and electron microscopes work). The maths behind it isn't that hard to follow if you've got engineering maths or undergraduate calculus. You put a particle in a potential energy "box", calculate at the states it can occupy, solve the equations and, lo and behold, there's a probability that the particle can sometimes be outside the "box". Impossible, obviously, because the particle hasn't got enough energy to get out of the box, but do a few experiments and the particle does what the maths predicts. Classical mechanics says it can't happen.
We all understand classical mechanics in a very practical way. Most of us can predict the path of a moving object and catch, kick or hit it. Quantum mechanics isn't like this at all. Even Einstein believed that the really odd aspects of it - like entanglement - couldn't be true.
*The Allan Adams Quantum Physics I, free from MIT. This is a really good course, not too difficult to follow and if you work through the course work and mid-terms then you should come out with a very good understanding of basic quantum physics. The problems and tests (and some solutions I think) are also available.
**Susskinds lectures are on the web and iTunes. There are lots, but the filming is a bit annoying and the sound poor in the earlier ones. However, if you're interested in the subject they are good lectures and he's a bit of a legend.
Re: a handy-dandy guide to quantum computer concepts
Perhaps have fun [1] at the same time:
https://content.minetest.net/packages/JavaFXpert/qiskitblocks/
.
[1] For some personally determined values of "fun" :-)
Australia has invested heavily in quantum computing
Australia's investment in silicon-based quantum computing (QC) started in the 90s with Bob Clark at the University of New South Wales. They now have the Centre for Quantum Computation and Communication Technology (at cqc2t.org). Checking out some of the pages about their research should give some idea of the amount of investment (and some of the problems they face).
Sometimes you need long-term investment in basic research before you see any returns. QC appears to be one of those.
Re: Australia has invested heavily in quantum computing
Basic research:
A good example is the transistor. First created in the lab in 1947, it did not become important until the integrated circuits of the late 1960s. But it is based on the quantum physics developed in universities in the 1920s.
So the transistor demonstrates a forty-year gap between basic research and day to day application. Controlled nuclear fusion is taking even longer. Then there is QC.
Interesting what Intel is doing
Intel managed to stay out of the headlines of quantum computing, but they are working on it: https://www.intel.com/content/www/us/en/research/quantum-computing.html
Maybe they are betting the farm on it, that could explain why they as super rich market leader in CPU design are a bit laggy in the sub 10nm game of CPU's.
For the moment it is like the "Self driving car, High Capacity EV batteries that don't become trash after six years and Nuclear Fusion" hype, everybody talks about it, but no one produces something that can be unleashed into the wild in the foreseeable future.
Re: Interesting what Intel is doing
Everyone is throwing money at QC just in case anyone else gets it to work. I'm quite confident it will not result in the advances people think it will.
Perhaps the time is right
For El Reg to provide a handy-dandy guide to quantum computer concepts and implementation for elderly engineers like me?
I dislike magic incantations (the quantum states collapse...) and have absolutely no idea how or why one might want such an animal - apart, apparently, to break the entire financial world.