Advertisement

Blog

Will Quantum Computing Enhance Analog Design? Part 3

In May, quantum computing was in the news as Google helped fund NASA's Quantum Artificial Intelligence Laboratory to house a D-Wave 2 quantum computer, by paying for the device. In a blog that I wrote at the time, I speculated that “I can imagine a validation system using a quantum computer that would avoid number-of-bits precision limitations during a simulation of an analog IC.”

Like another possible game changer, silicon photonics, quantum computing is an appealing topic in part because of the unknown potential. In fact, there is enough interest that market forecasts are appearing for quantum computing. Research firm Research and Markets released a new report earlier this month, succinctly titled “Photonic Integrated Circuit (IC) & Quantum Computing Market (2012 – 2022): By Application (Optical Fiber Communication, Optical Fiber Sensors, Biomedical); Components (Lasers, Attenuators); Raw Materials (Silica on Silicon, Silicon on Insulator).”

The title alone is a mouthful, but in a press release, Research and Markets finds that “quantum computing is another application of PICs which is forecasted to be commercialized in 2017. This technology is expected to completely revolutionize the computing industry.” PICs stands for Photonic Integrated Circuits, yet another hot topic in communications. Research and Markets estimates the total market for PICs will be more than $1.5 billion by 2022, growing at more than 26 percent a year between now and then.

The thing to keep in mind is that photons are the poster children for research in quantum states, quantum entanglement, and other unusual physics that may enable quantum computers. Thus, if you are forecasting a possible future where quantum computers are the workhorses, it makes sense to closely follow PICs and related technologies.

Quite a bit has happened since May. In that same month, Charles Margolis, writing on investor website Seeking Alpha, suggested that investing in quantum computing was fairly risky. The theme of Margolis's article was to explore any potential threat to companies like Intel, which might be presented by the evolution of quantum computing. He quoted an Intel source stating that Intel was not involved in quantum computing. Essentially this was all in response to the press about D-Wave at the time.

In June, physics.com reviewed quantum computing, devoting a lot of page space to the boson sampling devices I mentioned in my May blog. While they noted that boson sampling devices working on 20 to 30 photons are needed to fully see quantum interference effects, they also noted those barriers did not stop the formation of a $100 million (Canadian) private equity fund for quantum physics.

Just a few weeks later, Photonics Online reported on an MIT advancement in so-called “optical transistors.” Although not essential for quantum computing per-se, a true optical transistor is considered key to a fully photonic computing scheme.

MIT's work had to do with creating a “cloud” of super-cooled atoms, which were initially transparent but were entangled. Quantum entanglement is another very non-intuitive quantum behavior in which two atoms or photons become matched in their quantum states (i.e., they are entangled ), and remain so even when separated by relatively large distances.

This means that if you change the state of one entangled partner, the other also changes state, even though it is not close to the first partner. This allowed the MIT researchers to switch the state of the cloud with a single photon, changing the cloud from transparent to 80 percent opaque to light. Possibly the switching photons could then be sent around a PIC, switching photonic gates and leading to an actual photonic processor.

In early September, Bristol University in the United Kingdom announced Quantum in the Cloud, an interesting program wherein you can use a simulator of a boson sampling-like device to try out quantum computing experiments. Once you are ready for the real thing, you can register to use a 2-qubit quantum computer over the web.

Later in September, the University of California at Santa Barbara announced a new paper on Nature Physics Letters, describing a nanomechanical transducer device capable of converting electrical quantum states to optical quantum states. In the press release, the researchers say their device “uses an optomechanical crystal implemented in a piezoelectric material in a way that is compatible with superconducting qubits.” The following figure shows a color-coded SEM (scanning electron micrograph) of the device.

Scanning electron micrograph of the device showing the mechanically suspended optomechanical crystal (blue) with electrodes (yellow) and the photonic circuit (red).(Source: UCSB)

Scanning electron micrograph of the device showing the mechanically suspended optomechanical crystal (blue) with electrodes (yellow) and the photonic circuit (red).
(Source: UCSB)

The UCSB team plans to eventually connect the device to super-cooled quantum devices. Andrew Cleland, an author of the UCSB paper and associate director of the California Nanosystems Institute at UCSB, is quoted in the press release, saying, “We believe that combining optomechanics with superconducting quantum devices will enable a new generation of on-chip quantum devices with unique capabilities.”

What I believe is that creating qubits and entangled quantum states is becoming relatively common, and we may be entering an era of applications within a decade or so. As I said in May, such computing devices may be well suited to simulate and design analog systems, which inherently deal with continuous phenomena instead of digital ones.

Related posts:

21 comments on “Will Quantum Computing Enhance Analog Design? Part 3

  1. Davidled
    November 19, 2013

    I wonder when Quantum computer is popular in the market so that every engineering student buys it in order to calculate the complex formulation less than a few second. I guess that it is not easy to store information into memory by Quantum Computer since it is very sensitive in the temperature and it is kindly unstable.

  2. eafpres
    November 19, 2013

    @DaeJ–thanks for reading and offering an imaginative view of the future.  It is easy to laugh at such scenarios, but over the span of my career (30+ years) we have gone from computers being mainframes to having super-computer like power in a phone.  I certainly didn't have the vision 30 years ago to come close to what is common today.

    In many ways I feel we live in wondrous times.  While a lot of computing advances are used early in things like fluid dynamics, weather modeling, simulating the big bang and other such pursuits, it is also true that electronics design has made good use of computing and is often an early adopter.  I think that if a company could produce a dedicated quantum processor and design analog simulation software for it, they would really have something. I hope I get to see that!

  3. samicksha
    November 20, 2013

    This shall allow a quantum computer to decrypt many of the cryptographic systems that we are using today, i.e. it would be a in the number algorithm for solving the problem, parallel to same i am curious to see universal gate set for same.

  4. etnapowers
    November 20, 2013

    @Blaine: really nice BLOG! The optomechanical crystal is a really interesting device that could determine the definitive diffusion of the Silicon Photonics technology. I suppose that the characterization and the reliability of this crystal is really a key point for the success of this technology, which holds terrific promises of revolution in the electronics field.

  5. eafpres
    November 20, 2013

    @etnapowers–I agree there is a lot of work to do in order to prove some of these features are stable enough to make in a semiconductor fab and not just in the lab, and as you note reiiabilty will also be a key factor.  But just the idea of combining micro mechanical, electrical, adn photonic functions in one device leads to incredible possibilities.  I've written before  about how MEMS is making its way into analog front ends for tunable matching etc.  I feel we are still only at the very beginning of another round of revolution in electronic integration.

  6. eafpres
    November 20, 2013

    @samicksha–The Bristol University work lets you consider some types of standard structures for quantum computing.  However, there are still many different roads being pursued so what the final quantum computing device may be is still very open, if it even is possile to scale up.  So the gate set is still an open question.

  7. amrutah
    November 21, 2013

    @etnapowers:

        With so much of development happening in the field of semiconductors we are at a inflection point where a lot will change by 2020-2030.  With new technology like carbon nanotubes, new semiconductor material (compounds of materials) we will see new devices.  With new semiconductor technology we can hope to see these quantum computers to be small and easily available for everybody.

  8. etnapowers
    November 26, 2013

    @Blaine: I fully agree on incredible possibilities for this technology, moreover I think that only if the combination of “micro mechanical, electrical, adn photonic functions” , as you correctly said , will be stable and reliable this technology will be widely diffused.

  9. etnapowers
    November 26, 2013

    @amrutah: the scenario that you described is really exciting to me! As soon as new technologies will be producing new quantum computers there will be an intellectual revolution concerning the way to design , test , integrate and produce the electronic components useful to this scope.

  10. eafpres
    November 26, 2013

    @etnapowers–that is a very important point.  Even today's semiconductors can suffer from slow diffusion of atoms which can cause errros in a chip's function.  As each circuit element becomes smaller, we reach a point where movement of only a few atoms coudl disrupt a transistor.  This is made worse by higher density leading to higher core temperatures within a 3-D chip.

    For MEMS, I think we are early in the life of various MEMS applications and don't know how long these devices will really last, and what the failure modes might be.  I'm sure people are looking at it.  

  11. etnapowers
    November 26, 2013

    @Eafpres,

    that's absolutely true. I guess that new standards of reliability will be developed for these atomic systems and i think that in such a future scenario the engineers with a good knowledge of quantum physics will be very valued.

    The MEMS are devices based on the piezoelectric effect that is a electromechanical property, I wonder if there is a similar effect on atomic scale , for example on a layer of atoms.

     

  12. amrutah
    November 27, 2013

    @eafpres:

       I agree, with very little MEMS devices hitting the market we don't have a clear assessment of their reliability.  These being mechanical structures are prone vibration and have higher failure rates.

  13. etnapowers
    November 27, 2013

    MEMS systems are utilized in many applications and many industry sectors: consumer (gyroscopes for smartphone, tablets…),  vibrational sensors (safety, ), automotive (motor control) etc…

    Due to this wide diffusion the reliability of these devices will be tested very soon and I believe that the result will be good.

  14. yalanand
    November 30, 2013

    @Blaine, thanks for the post. I am curious to know if optical transistors will completely replace traditional transistor and how long quantum computing will take to come market ?

  15. yalanand
    November 30, 2013

    with very little MEMS devices hitting the market we don't have a clear assessment of their reliability.

    @amrutah, I totally agree with you. We still have limited knowledge on how MEMS devices fail because limited tools and models are available. How to model the reliability of MEMS is a challenge.

  16. yalanand
    November 30, 2013

     With new technology like carbon nanotubes, new semiconductor material (compounds of materials) we will see new devices. 

    @amrutah, I think nanotubes will completely change the way we implement systems because carbon nanotubes have totally different characteristics. I think VLSI engineers should start studying about nanotubes because this is going to be the next big technology.

  17. amrutah
    December 1, 2013

    @Yalanand: “carbon nanotubes have totally different characteristics…”

       Yes the characteristics of carbon nanotubes is different, it is proving to be a great material to redefine battery technology.  The portable devices will have a long lasting power, which might further reduce the size of cellphones (may get rid of charger module), surveillance drones.  Also it is good material for storage or memory devices.

  18. amrutah
    December 1, 2013

    @yalanand: Yes, improving the yield and reliability of the MEMS device is complex because of the existing knowledge of the MEMS device models.  

        Having said that, the 2007 ITRS (International Technical Roadmap for Semiconductors) release has already laid the roadmap for NEMS devices which might eventually replace MEMS devices.  The future for semiconductor is fast changing and help integrate more and more analog within a small area.

  19. eafpres
    December 2, 2013

    @yalanand–My guess is we are about 10 years away from large scale photonic transistor integration, maybe even longer.

    However, in the field of so-called Silicon Photonics there are commercialized devices with varying levels of integration of optical functions on chips.  A near-term goal is to replaced the so-called “light engine” in datacom and telecom equipment which uses fiber optics to send high data rates.  Today, most such equipment uses lasers, one interface which converts electrical to optical is expensive, and uses lots of power.  An example is a Si Photonic company called Luxtera provides a light engine to Molex for an “active optical cable”.

  20. Davidled
    December 2, 2013

    I think that silicon photonics could be used in the high data rate transmitter and receiver of Satellite. I believe that the cost is a little bit expensive now. Once cost is the acceptable point, this technology would be popular in other application. Then I expect that there is no more jitter and cross talking with noise in the PCB Board.

  21. eafpres
    December 2, 2013

    @DaeJ–good point; possibly a high level of photonic operation might be more radiation resistant which is an issue for space operations.

    However, I do not think jitter will be eliminated by Si Photonics.  I think that fundamental limits will mean there is always some level of jitter.

Leave a Reply