Current issues in Quantum Computing
Prabhas Chongstitvatana, 4 January 2016
Recently there is a report from "Google Quantum Artificial Intelligence"
research laboratory.
https://plus.google.com/+QuantumAILab
The team presents a paper that claims that their D-Wave Quantum Annealing
machine has a speed gain of 100 million times over classical
computers. I will discuss this report.
http://arxiv.org/pdf/1512.02206v3.pdf
Here is what is said in the news: You can see the picture to the D-Wave
quantum annealing machine.
http://www.thestar.com/news/insight/2016/01/02/google-takes-quantum-leap-into-artificial-intelligence.html
First, please watch a short video explaining what is a quantum computer.
https://www.youtube.com/watch?v=w_-_H9eBte8
Now we will have a look at the report.
The Google Quantum AI lab has the D-Wave 2X machine with around 1000 qubits
memory.
http://www.technologyreview.com/news/544421/googles-quantum-dream-machine/
The effort to build a quantum computer is concentrated on trying to build a
practical "quantum bit".
Here is the work of IBM.
http://www.technologyreview.com/news/537041/ibm-shows-off-a-quantum-computing-chip/
Now, we can discuss the claim from Google team.
It is a great achievement to have 100 million times speed up over a single
core processor.
To discuss this result, we can scale the baseline (single core) to multiple
core.
Modern GPU
Nvidia recent product is the Tesla board. The NVIDIA Tesla M40 is a
dual-slot 10.5 inch PCI Express Gen3 card with a single NVIDIA Maxwell GM200
graphics processing unit (GPU). The Tesla M40 has 12 GB GDDR5 on-board
memory and a 250 W maximum power limit. It contains 3072 cores, at approx 1
GHz.
http://images.nvidia.com/content/tesla/pdf/tesla-m40-product-brief.pdf
Top 500 supercomputer
Tianhe-2 tops the list with a performance of 33.86 petaflop/s (quadrillions
of calculations per second or Pflop/s) on the Linpack benchmark. No. 2 spot
is Titan, a Cray XK7 system installed at DOE’s Oak Ridge National
Laboratory. Titan achieved 17.59 petaflop/s on the Linpack benchmark.
Tianhe-2 has around 2 million cores. The system contains Intel Xeon E5-2692
12C 2.200GHz, TH Express-2, Intel Xeon Phi 31S1P. Titan has around
600,000 cores. It has Cray XK7 , Opteron 6274 16C 2.200GHz, Cray Gemini
interconnect, NVIDIA K20x.
http://top500.org/lists/2015/11/
So, if we compare absolute computing power, classical machines are still
very powerful.
However, the question about D-Wave machine is not how fast it is. It
is really fast as demonstrated, but the question remains whether it has
"exponential speed up" over classical machines?
Here is a simple story about this issue:
http://www.wired.com/2015/09/googles-quantum-computer-just-got-a-big-upgrade-1000-qubits/
There are several investigation into this question. One of the report
is here:
http://www.sciencemag.org/content/345/6195/420
And here is another view from the researcher in the field:
http://phys.org/news/2015-12-scott-aaronson-google-quantum-computing-paper.html
If you are interested in programming a quantum computer, I have one paper
published in 2012. My
result achieved an exponential speed up.
Yingchareonthawornchai,
S., Aporntewan, C., and Chongstitvatana, P., "An Implementation of Compact
Genetic Algorithm on a Quantum Computer," Int. Joint Conf. on Computer
Science and Software Engineering (JCSSE), 30 May - 1 June 2012,
pp.131-135.
http://www.cp.eng.chula.ac.th/~piak/paper/2012/jcsse-quantum-cga.pdf
In summary,
Quantum computer is here. It has been demonstrated its potential.
However, D-Wave is a special purpose computer that can do only a few limited
type of problems. Practical quantum computer is believed to be
possible in a few years. What we know is that Tech companies poured
their resources into this promising field. (IBM, Google, Microsoft, etc.)
Here is the link to D-wave company http://www.dwavesys.com/
You can learn more about quantum computing by visiting this website. It has
some nice simulator. (best browsing on Chrome)
http://www.quantumplayground.net/#/home
Enjoy!