My Interests in Quantum Computing: Its Capabilities, Limitations, and Issues

I’m a technologist rather than an application developer, so I have no personal application for quantum computing per se, but as a technologist I’m interested in the nature of the technology itself — the capabilities of quantum computers, their limitations, and whatever issues might interfere with the exploitation of the technology by real-world application developers. This informal paper will outline my personal focus on the capabilities, limitations, and issues with quantum computing.

Although I couch this informal paper in terms of my own personal interests, I am also saying that this is my personal view of the opportunities and challenges that almost everyone will encounter with this new technology. As such, the reader can and should presume that my views expressed herein represent a rough summary of the topic areas relevant to the state of the art of quantum computing, not so much as a snapshot of the moment, but the likely trajectory of the sector in the coming years, not so much in terms of specific numbers and specific technical features, but in more general terms of capabilities, limitations, and issues to be addressed.

It’s not my intention to detail all aspects of my interests in quantum computing, but simply to offer a broad outline with just enough detail and color to make it clear where my interests lie. Details on my interests can be found in various informal papers which I have posted on Medium:

In a nutshell

Put most simply, this informal paper addresses two basic questions:

  1. What is my interest in quantum computing?
  2. What interests me about quantum computing?

I divide my interests into three levels:

  1. Primary interests. What I care about the most. My primary focus. Where most of my attention and energy goes.
  2. Secondary interests. Matters I’m still interested in, but not as my top priority.
  3. Tertiary interests. Matters that are still important to somebody, but not to me, at least not so much.

My three main areas of interest — primary interests:

  1. Capabilities. What can a quantum computer do.
  2. Limitations. What can’t a quantum computer do, or do only with some degree of difficulty.
  3. Issues. Impediments and obstacles which need to be addressed before people can successfully exploit quantum computers — to achieve widespread deployment of production-ready, production-scale, practical, real-world applications which achieve a dramatic level of quantum advantage.

Subsequent sections will suggest some of the detail of those areas of my interest.

To be clear, all aspects of quantum computing, and quantum information science in general, are of interest to me, but some more than others. This informal paper outlines what aspects are of most interest to me.

Capabilities

Just to provide some color and not intending to be comprehensive or exhaustive, here are some illustrative examples of topics of primary interest to me with regards to the capabilities of quantum computing:

  1. What can a quantum computer do. (See Limitations for what a quantum computer can’t do.)
  2. What makes a quantum computer tick.
  3. How does a quantum computer work.
  4. Underlying STEM which enables the technology, including the quantum effects of quantum mechanics and physics which are the central enablement of quantum computing.
  5. Theory of quantum computing.
  6. Appropriate levels of abstraction. Exploit raw physics and the hardware, but facilitate application solutions. How many levels of abstraction make most sense? Unknown at this stage. Will vary depending on needs.
  7. Quantum computing in the larger overall context of quantum information science.
  8. Design of a quantum computer.
  9. Architecture of a quantum computer.
  10. Features.
  11. Functions.
  12. Tasks which can be accomplished with quantum computers.
  13. What functions of a classical computer are supported.
  14. What can a quantum computer do better than a classical computer (quantum advantage)?
  15. How much better can a quantum computer be than a classical computer for a particular algorithm or application (actual quantum advantage)?
  16. What can a quantum computer do that a classical computer can’t (quantum supremacy)?
  17. Algorithms.
  18. Academic papers pointing to potential advances or evaluating possibilities, or simply for historical reference or perspective. Preferably posted as preprints online and free on arXiv.org.
  19. Online (and free) tutorials, lecture notes, and slide decks. The best and easiest ways to access knowledge on quantum computing.
  20. Programming models.
  21. Programming languages tailored to the non-classical aspects of quantum computing.
  22. Principles of operation. What does a developer need to know about a particular quantum computer to successfully develop an application which fully exploits the features and power of that machine.
  23. Design patterns.
  24. Metaphors for mapping application problems to quantum solutions.
  25. Algorithmic building blocks and libraries.
  26. Mapping high-level computational models to the raw physics of quantum processors.
  27. Application frameworks.
  28. Applications which are enabled and can be supported by the functions of a quantum computer.
  29. Broad categories for applications suitable for quantum computing.
  30. Business problems which can be addressed using quantum computing.
  31. STEM problems which can be addressed using quantum computing. Simulating physics and chemistry, for example.
  32. Solving optimization problems in general, where combinatorial explosion precludes direct classical solutions.
  33. Quantum computer as a function, subroutine, or coprocessor for a classical computer.
  34. Pure quantum computing as well as hybrid quantum/classical computing.
  35. Hybrid quantum/classical computing.
  36. Universal quantum computing. Merger or blending of classical and quantum computing in a single, unified machine.
  37. Types of problems which can be solved with quantum computers.
  38. Performance — what you can do and how long it will take.
  39. Scaling. Ability to handle larger problems. Qubit requirements and Big-O complexity as a function of input size.
  40. Where and how you can get exponential speedup.
  41. Specifications and documentation — what you can do, all of the details.
  42. What might the capabilities of future quantum computers look like compared to those of today and those expected over the next couple of years?
  43. Industry trends, in terms of broad trends and expectations as they relate to capabilities.
  44. How much can you really do with NISQ machines?
  45. How much more could you do with a post-NISQ machine?
  46. Competition to the extent that it results in alternative approaches as a hedge against the prospect that a single, chosen, preferred technological approach to quantum computing might fail or be less capable than competing alternatives.
  47. Specific customers and specific applications — to the extent that they provide dramatic evidence of the utility of the capabilities of quantum computing.
  48. Tools. To the extent that they introduce, enable, or greatly facilitate the use of capabilities which might otherwise be much too difficult to use for most application developers.
  49. Simulators. To provide the capabilities of quantum hardware which does not yet exist.

Limitations

Just to provide some color and not intending to be comprehensive or exhaustive, here are some illustrative examples of topics of primary interest to me with regards to the limitations of quantum computing:

  1. What a quantum computer can’t do, or do only with some degree of difficulty.
  2. Factors which limit what quantum computers can do, or how people can use them.
  3. What can’t you do (well) with NISQ machines.
  4. What limitations of NISQ machines might be relaxed for post-NISQ machines.
  5. What limitations would remain even for post-NISQ machines.
  6. What functions of a classical computer are not supported.
  7. Which types of problems cannot be so readily solved using quantum computers.
  8. Hybrid quantum/classical computing — drawbacks and limitations.
  9. Performance — what you can’t do or takes too long to satisfy technical and business requirements.
  10. Where and how you can’t get exponential speedup.
  11. Specifications and documentation — what you can’t do and limits to what you can do.
  12. Limitations of the underlying STEM which enables quantum computing.
  13. All the Turing machine features that are missing.
  14. Lack of rich datatypes.
  15. Lack of I/O, database access, big data, network access, and network services.
  16. Difficulty of debugging since quantum state collapses upon measurement.
  17. Too few qubits.
  18. Coherence time too short.
  19. Gate fidelity errors.
  20. Limited connectivity between qubits, and cost and fidelity of swap networks.
  21. Not all algorithms can achieve exponential speedup. For example, the much-touted Grover search algorithm achieves only a quadratic speedup — not even close to exponential speedup.
  22. Most touted algorithms are not viable based on limitations of current quantum computers and even those expected over the next few years: phase estimation, quantum Fourier transform, order finding, and Shor’s algorithm for factoring large semiprime numbers.
  23. Tools. To the extent that they can help to mitigate limitations in a dramatic rather than superficial manner.
  24. Extent to which simulators cannot simulate current real hardware or desired configurations of future hardware, such as raw number of qubits or total quantum states. Or not adequately model the actual limitations of the actual hardware, such as error rates and intermittent errors or environmental interference, or gradual decay of calibration.
  25. Extent to which shots (quantum circuit repetitions) must be very high to compensate for significant error rate in hardware. And may not scale well as size of data increases.
  26. Extent to which shots (quantum circuit repetitions) must be very high to compensate for inherent probabilistic nature of qubit hardware. And may not scale well as size of data increases.
  27. Lack of staff with sufficient aptitude, knowledge, and skills.
  28. Limited talent pool for attracting staff with sufficient aptitude, knowledge, and skills.
  29. Insufficient organizational budget to attract and retain staff with sufficient aptitude, knowledge, and skills.
  30. Level of difficulty for on-the-job training.
  31. Degree to which educational institutions may be unprepared to adequately educate students in quantum information science in sufficient numbers to satisfy organizational hiring needs.
  32. Quantum advantage is likely to be proprietary, bestowing benefits on early movers.” — IBM report. Reduces extent of sharing and learning from others.

Issues

Just to provide some color and not intending to be comprehensive or exhaustive, here are some illustrative examples of topics of primary interest to me with regards to the issues of quantum computing relevant to achieving widespread deployment of production-ready, production-scale, practical, real-world applications which achieve a dramatic level of quantum advantage:

  1. Overwhelming hype. Tends to be more harmful than helpful. Overall, it’s just disruptive without adding much positive value. Given that the hype is overwhelming, one of my personal goals is to try to overwhelm — and even transcend — the hype with pragmatic realism.
  2. Timing. A lot of the hype implies that much can be done today. Lack of realistic clarity for timing of most promises of what may come in the years ahead. Lack of clarity for less than five years vs. more than five years, if not ten years.
  3. Understanding the personas, use cases, and access patterns for quantum computing. Who might use or depend on quantum computing, what they expect to achieve, and how they might go about doing it.
  4. Impediments and obstacles which need to be addressed before people can successfully exploit quantum computers.
  5. Factors which need to be addressed.
  6. Some open issues are fundamental, others are just a matter of time, and others are resource issues.
  7. Degree of documentation — features, functions, specifications.
  8. Describing and explaining quantum computing in plain English.
  9. Describing and explaining quantum parallelism in plain English.
  10. Understanding how real-world problems can be mapped to solutions exploiting quantum parallelism.
  11. Degree of open source as a general goal or at least free online access.
  12. Degree of progress which may remain hidden from wide public view and access due to proprietary intellectual property. “Quantum advantage is likely to be proprietary, bestowing benefits on early movers.” — IBM report. Reduces extent of sharing and learning from others. How can this issue be overcome?
  13. Probabilistic rather than deterministic nature of quantum computing.
  14. Difficulty of mapping or converting classical algorithms to quantum.
  15. Difficulty of analyzing problems to derive a quantum solution.
  16. Achieving exponential speedup is no slam dunk cake walk.
  17. Hybrid algorithms are tedious to design and develop, and deliver less than stellar results and performance.
  18. The greatest challenges for quantum computing are hardware and algorithms.
  19. Need for off the shelf solutions (applications) to real-world problems. So organizations can immediately deploy quantum solutions without the need to become quantum experts.
  20. Levels of skills and aptitude required for the many different roles of staff involved with the many aspects of quantum computing. Who will need a PhD in physics, who will need an MS in quantum information science, undergraduate exposure to quantum computing, and who can get by with a simple two-week training course. Or a two-day course for technical managers. Or a two-hour seminar for senior managers and executives.
  21. Education. For research — Phd, postdoc. For masters degree — awareness of quantum vs. focus on quantum. For undergraduate — awareness vs. minor vs. major. High School — awareness, introduction, and some actual use. Continuing education and lifelong learning — all levels.
  22. Training. So many possibilities, but what makes the most sense for each particular situation — one size does not fit all.
  23. At what stage might we expect to see staff who are quantum native — more at home in the quantum world than the classical world. When will that be needed? When can it be exploited for dramatic effect.
  24. How to integrate quantum computing infrastructure and staff with current IT infrastructure and staff. Or, how and when to keep it separate.
  25. Unclear what hardware technology will yield the best qubits.
  26. Scaling. How well will current algorithms scale, for larger problems, for more qubits, for raw performance — Big-O, for quantum advantage, or given any connectivity limitations.
  27. Need for methodology and rules of thumb for determining number of shots (quantum circuit repetitions) to compensate for significant error rate in hardware.
  28. Gaps, holes, weaknesses, and uncertainties in general.
  29. Roadmaps. For hardware, for algorithms, for support software, for applications.
  30. Implementations. Ideas are great and essential, but transforming them into reality is another matter.
  31. When will quantum computing be ready for prime-time, production-scale, practical, real-world applications that deliver dramatic quantum advantage and real business value?
  32. Promises and hype vs. reality and deliveries.
  33. Need for much better documentation.
  34. Need for a rich collection of quantum algorithmic building blocks.
  35. Need for a rich portfolio of real-world example algorithms and applications.
  36. Need for much more comprehensive technical specifications.
  37. Need for fully documented principles of operations for each machine detailing everything a programmer needs to know about a particular quantum computer to successfully develop an algorithm or application for that particular machine.
  38. Details of open source. As much code and documentation as possible should be open source, or at least with free online access. This includes specifications, training materials, tutorials, lecture notes, online courses, and even the full text of books whenever possible.
  39. Getting authors to put all of their code and relevant configuration details into publicly accessible GitHub repositories.
  40. Transitioning from the early stage of experimental, test, demonstration, toy, research, prototype, and mockup efforts to the stage of production-ready, production-scale, real-world practical algorithms and applications.
  41. Transitioning to a mature sector where widespread production development and deployment is commonplace.
  42. Ease of design.
  43. Ease of implementation.
  44. Breakthroughs. Rather than mere slow incremental progress.
  45. Tools. To the extent that they can mitigate issues — in a dramatic rather than superficial manner.
  46. When will we move beyond NISQ machines?
  47. How perfect can we make qubits.
  48. Do we really need quantum error correction. When might we get it.
  49. How much of current efforts will bear fruit or will have to be discarded and started over as the technology progresses over the coming years. What will efforts look like five to ten years from now compared to today — how much of today’s efforts and technology will even be recognizable.
  50. When might it be true that a particular application or algorithm won’t be dramatically faster using a quantum computer than sticking with purely classical computing.

Secondary interests

Some examples of my secondary interests, which are secondary matters which I still care about but are not a personal high priority for me:

  1. Different technologies for implementing qubits. I’m more interested in what you can do with a qubit than how it is implemented.
  2. How various qubit technologies are implemented.
  3. Who are the major vendors, university research labs, consulting firms, and other players in the sector.
  4. What can be implemented over the next few years or even 5–10 years, as opposed to implemented today. I’m more interested in what’s coming than what’s here now.
  5. Who has the most powerful quantum computer — from the perspective of how much can be accomplished using it.
  6. Industry trends, in terms of run of the mill announcements and relatively minor incremental short-term progress.
  7. Tools in general. Unless they dramatically enhance capability, dramatically mitigate limitations, or otherwise dramatically address issues.
  8. An introduction to quantum computing that actually sets the stage for conceptualizing practical applications.
  9. How to successfully get past the complexity of the math and physics which underlies quantum computing.
  10. An introductory tutorial which actually enables the development of practical applications.
  11. Online (and free) courses. I’d prefer to read raw materials at my own pace and on my own trajectory of interests, but sometimes online (and free) courses can be more efficient.
  12. Websites dedicated to quantum computing. Can be helpful, sometimes, but much of the hard-core material is distributed around the Internet on sites not dedicated to quantum computing per se.
  13. Academic journals. I prefer academic journal papers posted as preprints on arXiv.org, but sometimes papers are only available in the formal journals themselves or online, sometimes accessible for free but often hidden behind paywalls.

Tertiary interests

Tertiary matters are those that are likely important to others, but I simply don’t have much interest in. Ultimately, I care about just about everything related to quantum computing, but it’s a question of where I should focus the bulk of my attention.

Some examples of my tertiary interests:

  1. Marketing.
  2. Branding.
  3. Awareness building. Needed, but not clear how much real value it has. One exception being management and policymakers needing enough awareness to approve projects and fund budgets for quantum, especially research projects.
  4. General media. Man, most of this stuff really makes me cringe!! Even in finance, business, and technology media. The hard-science media is somewhat better, but still prone to hype.
  5. What can be implemented today, as opposed to over the next few years.
  6. What’s here now rather than what might be coming over the next few to five years.
  7. How much does a quantum computer cost? Pricing. Pricing of services. Pricing of consulting.
  8. What is the business model for quantum computing? Models, plural.
  9. Economics. Critical, but I’m a technology guy. Impact on the overall economy, impact on specific sectors of the economy.
  10. Which vendors compete with which other vendors and how?
  11. Funding of vendors.
  12. Government grants to fund research projects.
  13. Who has the most powerful quantum computer — I don’t care who it is, just that it exists and what it can and can’t do.
  14. Research projects which are not so likely to find their way into commercial offerings.
  15. Specific details of documentation.
  16. Specific details of algorithms, libraries, and frameworks.
  17. Specific details of the architecture of a particular quantum computer.
  18. Logistical details of a particular quantum computer. Physical size, power requirements, operating instructions, calibration, troubleshooting.
  19. Specific performance of a particular quantum computer.
  20. Pictures and images of a particular quantum computer.
  21. Specific details of APIs and libraries for a particular quantum computer.
  22. Who’s who — the personalities and egos. The cult of personality is not a real interest of mine.
  23. Academic research. Much of it is too esoteric to have much of an immediate or even longer-term impact for capabilities. But the good stuff, having real potential is of primary interest.
  24. Competition. Not particularly of interest to me except to the extent that it results in offerings with superior capabilities.
  25. Specific customers and specific applications. More interested in broad and abstract qualities of applications.
  26. Tools in general.
  27. Layers of software and hardware which do not add significant value (from the perspective of an application developer) to the raw capabilities of the machine.
  28. General availability of commercial machines. Eventually this is important, but for now the priority is just seeing the concepts proven at all, even once, such as the ENIAC moment and the FORTRAN moment.
  29. Exact timing and error rates. Good to know, but the priority now is simply to get error rates down, and timing is not a priority if circuits can only be very shallow due to error rates.
  30. Hands-on coding. I have no interest in developing code or actually running algorithms and applications. It can be helpful to read and study code, but my priority is on ideas and design, architecture, and higher-level abstractions.
  31. Experimental, test, demonstration, toy, research, prototype, and mockup efforts that have little prospect of transitioning to production-ready, production-scale, real-world practical algorithms and applications.
  32. Details of particular programming and execution environments.
  33. Geopolitical concerns. Various countries wish to lead and nobody wants to be left behind.
  34. Books. Some people love or even need them, but I find books to be an archaic holdover in the age of the Internet. I prefer online — and free — text.
  35. Conferences. As with books, some people love or even need them, but I find conferences to be an archaic holdover in the age of the Internet. I prefer online — and free — material and forums for interaction. I’d rather view a slide deck online at my own pace than sit through a conference session.
  36. Blogs. They can be helpful and some people really love them, but usually I’m looking for deeper knowledge and original sources or simply answers to questions without all of the narrative.
  37. Quantum volume. More of a marketing buzzword. No real utility for application developers who need to know more of the specific details of the machine (Principles of Operation — number of qubits available, maximum circuit depth, specific connectivity constraints, coherence time, gate error rate) — quantum volume hides those essential details. Still, it does have some limited value for indicating progress in enhancing the capabilities of new machines — a progress indicator. Even so, not as useful as MIPS or FLOPS for classical computers.
  38. History of quantum computing. I am fairly interested in it, but more as a minor background interest since it rarely has much of an impact on how I think about either the here and now or future of quantum computing.

My background and experience

My background is as a software developer of system software for classical computers, including programming languages, compilers, interpreters, software tools, graphics engines, search engines, database engines, and application frameworks, but my interest now is to understand quantum computing as deeply and broadly as possible to assess what it can do and what it can’t do as well as what is easy and what is hard.

I’m also interested in the boundary between classical computation and quantum computation, what portions of an application or algorithm belong on the classical side of that boundary, and what can directly exploit a quantum computer, and how application and algorithm designers can decide what goes on which side of that boundary to most fully and effectively exploit the full power of quantum computers.

And in some cases as much as one might wish to utilize a quantum computer, it might still be true that a particular application or algorithm won’t be dramatically faster using a quantum computer than sticking with purely classical computing.

My profile on LinkedIn:

Details on my interests

Details on my interests in quantum computing can be found in various informal papers which I have posted on Medium:

What’s next?

With two years of study and writing behind me now, my primary focus is simply to keep up on progress in the field, waiting impatiently for interesting breakthroughs, including:

  • Dramatic hardware advances. 64 qubits, 128 qubits, 256 qubits, 512 qubits, 1K qubits.
  • Interesting new algorithms. Quantum leaps ahead of the tired old stuff in “Mike and Ike.”
  • More powerful simulators while we wait for the hardware to catch up.
  • Arrival of The ENIAC Moment. A real-world app.
  • Arrival of The FORTRAN Moment. When any organization with semi-competent software developers can readily develop production applications exploiting quantum computing for a truly dramatic quantum advantage.
  • Arrival of quantum advantage. Initially ten to a hundred or a thousand times faster, but ideally a million, a billion, even a trillion times faster.
  • Arrival of quantum supremacy — for real-world applications. Google gets credit for their accomplishment, but we need to see quantum supremacy for real-world applications.
  • The unexpected advances that really inspire us. You never know. Nobody expected the many advances in classical computing as they happened. I’m waiting.

Or…

Should I take a break?

For a number of months I have been contemplating taking a break, maybe even up to two years, to give the sector time to catch up to my own expectations. I’d still put in a background level of effort monitoring progress, just in case any truly dramatic breakthroughs occur. Actually, that’s mostly what I am doing right now, for the most part, plus a little writing now and then.

Breakthroughs are of course quite welcome, but I don’t expect so many over the next two years (whether that is 18 months, 30 months, or even three years.) I do expect significant dollops of incremental progress, many minor breakthroughs, but true, dramatic breakthroughs, not so much. I expect a relative dearth of truly dramatic breakthroughs over the next two years. But, I am very ready and willing to be surprised, at any moment.

For now, I’ll continue plodding along, monitoring the incremental progress of this nascent sector.

Written by

Freelance Consultant

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store