Key Developments in Quantum Computing to Watch for in 2023

This informal paper lists a wide variety of key developments in quantum computing which may or may not happen in 2023. Most of the developments are strictly technical, but some are non-technical. Developments on the list are not guaranteed to happen nor even necessarily likely to happen. Some of the developments are fairly unlikely, but still remotely possible, so they will be especially notable if they do happen.

The list is not necessarily exhaustively complete (although exhausting!), but intended to be fairly comprehensive and at least representative of the lion’s share of areas needing attention. The point is to present a broad view of developments that are needed and worth watching for.

This is my own personal list. Others might have differing lists of their own. Each organization which wants to be serious about quantum computing should have a list of its own. My list can be viewed as a sample, suggesting items which others may choose to include in their own lists.

Ultimately, such a list can be used as a checklist to judge progress in the quantum computing sector, either overall or relative to the needs of a particular organization.

Seriously, if you don’t have your own list of key developments in quantum computing which you are watching for in 2023 (or beyond), you are not being serious about quantum computing.

Potential developments in 2023

This list is in no particular order, and neither structured nor prioritized, although newer items will tend to be added to the end. All items are implicitly for 2023, although many will relate to subsequent years as well.

1. Will there be any truly blockbuster breakthroughs in quantum computing? What areas might be ripe?
2. Will there be many notable breakthroughs? Even if not truly blockbuster. What areas might be ripe?
3. Will the year be more about plodding and incremental progress than breakthroughs or blockbuster breakthroughs?
4. Will the year be more about reality trying to catch up to the hype (unfulfilled promises of past years) rather than many bold new promises? What areas might require the most catch-up?
5. Will IBM make Osprey available early in 2023? Presumably in the first quarter.
6. Will IBM fully disclose the technical details and performance of Osprey early in 2023?
7. Will IBM’s Osprey be yet another flop or will it include some notable surprises?
8. Will IBM introduce Condor on schedule? Presumably in November. No big surprise or big deal if full availability is delayed a few months as with Osprey.
9. Will IBM introduce Heron on schedule? Presumably in November. No big surprise or big deal if full availability is delayed a few months as with Osprey.
10. Will IBM introduce a quantum processor which can consistently achieve over three nines of qubit fidelity? They’ve suggested that Heron could achieve three nines.
11. Will IBM’s Condor be yet another flop or will it include some notable surprises?
12. Will IBM’s Heron be yet another flop or will it include some notable surprises?
13. Will IBM finally add qubit quality to their quantum roadmap? And will they clearly mark milestones for three, 3.25, 3.5, 3.75, four, 4.5, and five nines of qubit fidelity? Where will qubit quality end up?
14. Will anybody introduce or make truly notable progress on full, transparent, and automatic quantum error correction?
15. Will IBM or any of the other superconducting transmon qubit vendors finally acknowledge that full any to any qubit connectivity is mandatory? And clearly detail milestones to get there.
16. Will Rigetti up their game dramatically and introduce a truly notable offering that makes them look substantially better than a mediocre also-ran? Much better qubit fidelity and full any to any connectivity could do the trick, but is either really in the cards for them in the next year or two?
17. Will Microsoft finally produce a workable topological qubit in the lab? And implement single qubit Clifford gates and measurements? Or two or three qubits so they can demonstrate two-qubit gates?
18. Will Intel finally demonstrate silicon spin qubits in the lab? And implement single qubit Clifford gates and measurements? Or two or three qubits so they can demonstrate two-qubit gates?
19. Will qubits introduced by Microsoft and Intel be very high fidelity and with full any to any connectivity by definition from the get-go? It would be a real shame to wait for so long to see practical qubits from Microsoft and Intel but then have them be of low quality and of no real utility for larger and more sophisticated quantum algorithms.
20. Will use of SWAP networks to emulate full qubit connectivity wane? This can only happen with the advent of full any to any connectivity.
21. Will Google introduce a new quantum computer?
22. Will Google finally officially acknowledge and document the purported 72-qubit quantum computer mentioned in a paper last July, but not formally announced or documented?
23. Will Google introduce a new quantum computer that actually demonstrates quantum error correction for one or more full logical qubits?
24. Will a new qubit technology be introduced? Even if only partially demonstrated in the lab.
25. Will any of the trapped-ion qubit vendors achieve support for full 32-qubit algorithms?
26. Will any of the trapped-ion qubit vendors achieve support for more than 32-qubit algorithms? 36? 40? 48?
27. Might any hardware vendor actually implement my proposal for 48 fully-connected near-perfect qubits capable of supporting a 20-bit quantum Fourier transform? Finally at the lower end of the range for practical quantum computing and a credible level of significant quantum advantage. I don’t expect it in 2023, but maybe somebody will come close or commit to achieving it in 2024. I don’t care whether it is superconducting transmon qubits or trapped-ion qubits — but even better if both!
28. Will algorithm designers finally focus on simulators rather than real quantum computer hardware so that they can demonstrate 24, 28, 32, 36, and 40-qubit algorithms? Presumably based on near-perfect qubit fidelity and full any to any qubit connectivity.
29. Will we finally start seeing a significant number of papers on 24, 28, 32, 36, and 40-qubit algorithms?
30. Will 24, 28, 32, 36, and 40-qubit algorithms become the norm? When will…? Might we hit the end of the year without achieving this norm?
31. Will quantum algorithm designers finally abandon variational methods in favor of focusing on quantum phase estimation? Implies a shift in focus from the limited capabilities of current quantum computer hardware in favor of simulators with near-perfect qubit fidelity and full any to any qubit conductivity.
32. Will we at least see a dramatic or even moderate reduction of the reliance on variational methods in quantum algorithms?
33. Will we see a dramatic and consistent emphasis on quantum Fourier transform and quantum phase estimation for advanced quantum algorithms?
34. Will people finally begin to realize that NISQ is an unproductive distraction and won’t lead to practical quantum applications which can achieve a truly dramatic quantum advantage? Or will we continue to hear a chorus of people continuing to misguidedly tout the merits of NISQ for practical applications in the near term?
35. Will all of the neutral-atom qubit vendors finally come clean publicly about their programming models? Are they really just special purpose software-configurable physics experiments rather than general-purpose quantum computers? QuEra has already stated that they use analog processing mode. ColdQuanta/Infleqtion and Atom Computing have not yet done so. Pasqal has implied that they will eventually support both modes, but they need a more clear statement with timing to clarify what they are today as opposed to what they might be in so many years.
36. Will ColdQuanta (Infleqtion) publicly disclose technical details of their 100-qubit Hilbert quantum computing device? Is it a true general purpose quantum computer or a special purpose device, possibly simply a software-controllable physics experiment? How does it compare or stack up against superconducting transmon qubits and trapped-ion qubits?
37. Will all of the photonics quantum computing vendors finally come clean publicly about their programming models? Are they really just special purpose software-configurable physics experiments rather than general-purpose quantum computers?
38. Will PsiQuantum finally exit stealth mode? Or will they continue burning through large amounts of cash and generating large amounts of hype? Might they even burn out and crash and burn this year (or next)? Will PsiQuantum become the Theranos of quantum computing?
39. Will there be further IPOs and SPACs for quantum computing companies this year?
40. Will the existing stock of publicly-traded quantum computing companies recover this year? Or continue to decline?
41. Will we start seeing some medium-level and higher-level algorithmic building blocks so that individual quantum logic gates become a rarity rather than the norm?
42. Will the potpourri of error mitigation and circuit splitting techniques and tools finally bear some fruit or be a complete or relatively complete flop? It’s truly bizarre to expect average quantum algorithm designers and application developers to shoulder the full burden of poor qubit design. Will average quantum algorithm designers be able to master and marshal all of these techniques and tools or will only the most super-elite be able to do so? Will they deliver a sufficiently lower error rate and be easily scalable enough to be useful for larger complex algorithms?
43. Will coherence time and gate execution time progress enough to make significantly larger circuits a reality? Much lower error rate is needed as well. How much larger will circuits be permitted to be than last year?
44. Will hype diminish, stay roughly at the same level, or even increase? Will vendors have learned their lessons at some point this year and change their ways, or just stay on auto-pilot right through the end of the year?
45. Will hype be the area of greatest growth of all of the areas of quantum computing?
46. Which, if any, of the significant (or insignificant) vendors will shutdown?
47. Which, if any, of the significant (or insignificant) vendors will merge with larger firms to avoid shutdown?
48. Which, if any, of the significant (or insignificant) vendors will merge with larger firms? Not implying that they are failing or in a weak condition.
49. Which independent quantum computing firms will be the largest in terms or either or both revenue and profit?
50. Will there be greater demand or interest in hardware-oriented quantum computing firms or software-oriented firms? Will interest in hardware firms increase, be roughly the same, or decline? Will interest in software firms increase, be roughly the same, or decline? Should developer tools, algorithms, and applications be broken out as separate categories, distinct from infrastructure software?
51. On the software side, will there be greater demand or interest in infrastructure software, tools, algorithms, or applications?
52. Will quantum-inspired classical algorithms be in high demand or interest?
53. Will IBM’s Quantum Volume (QV) metric for qubit quality last the year? Requires simulation of the generated circuit, so more than 50-qubit circuits are out of the question. Even 40-qubit circuits are unlikely to be simulatable. Even 32-qubits could be problematic. Superconducting transmon qubits are unlikely to get even to 32 qubits for quantum volume, although 32 qubits is already in range for some trapped-ion machines.
54. Will a new metric for qubit and circuit quality emerge? To replace IBM’s Quantum Volume (QV) which is only valid up to about 50 qubits — however large a circuit that can be supported by classical quantum simulator software. And will a consensus of support emerge to support it?
55. Will D-Wave gain in popularity, stay the same, or diminish in popularity?
56. Will a consensus emerge as to potential use cases for D-Wave? Including input data size which can be supported.
57. Is quantum computing technology getting so complex that it’s rapidly becoming a Tower of Babel? And more likely to crumble and collapse than to thrive? Or at least get much worse and less manageable even if it doesn’t fully crumble and collapse?
58. Is the shared vocabulary of quantum computing diverging so that increasingly we are unable to adequately discuss the technology with each other?
59. How stable will the definitions of quantum computer, quantum computing device, general-purpose quantum computer, practical quantum computer, special-purpose quantum computer, and special-purpose quantum computing device be? The shared vocabulary seems to be in a state of significant flux. Will boson sampling devices, photonic computing devices, quantum annealing devices, and neutral-atom quantum computing devices be considered as quantum computers or even general-purpose quantum computers as the year progresses?
60. Will we be able to judge the progress and advancement of quantum computing on a quarterly basis? Will we be able to detect palpable and notable progress? Or is six to nine months or even a year a more credible time basis for detecting palpable and notable progress? How much progress should we be able to detect at the end of Q1 or Q2?
61. Will transparency and technical disclosure improve for quantum computing technology, stay the same, or even get worse? Will it be a relatively minor change (positive or negative) or relatively major (positive or negative)? Including roadmaps as well as the actual technology.
62. Will metrics and benchmarks improve significantly, remain about the same, or deteriorate over the year?
63. Will we see any major new metrics and benchmarks for measuring the technical capabilities of quantum computers? Will something replace Quantum Volume (QV)? Will something measure quantum Fourier transform and quantum phase estimation? Is there much else that is really worth measuring at this stage, prior to the advent of a truly practical quantum computer? You can’t have much in the way of application benchmarks if you can’t have much in the way of applications
64. Will there be any or any widespread adoption of my Proposal for a Quantum Capabilities Label for Quantum Computers, Algorithms, and Applications?
65. Is there any chance of a practical quantum computer emerging? Or even close? How close? For a definition of practical quantum computer see my Elevator Pitch for Quantum Computing paper.
66. What will be the general state of the art for qubit fidelity? Including measurement. By the end of the year, by mid-year, and even by quarter. Might three nines be common by the end of the year? How soon might we see three nines for more than a single machine or two? Is 2.75 nines likely to be common by late in the year? Is 2.5 nines likely to be common by the third quarter? Is two nines (99%) or even 1.8 nines (98%) likely to be quite common even in the second half of the year?
67. Will near-perfect qubits become widely available or even available at all? Four to five nines of qubit fidelity, or at least 3.5 nines or so. Unlikely this year, but… See my What Is a Near-perfect Qubit? paper.
68. Might we achieve The ENIAC Moment or at least come close or close enough to feel that it is really within reach within another year or two?
69. What level of quantum advantage might be achieved by the end of the year and by the end of each quarter? Dramatic quantum advantage? SIgnificant or substantial quantum advantage? What fraction? See my papers on What Is Dramatic Quantum Advantage? and Fractional Quantum Advantage — Stepping Stones to Dramatic Quantum Advantage.
70. Will qutrits gain any traction?
71. Will qudits gain any traction? Value of d is other than 2 or 3 (which are qubits or qutrits.)
72. Will any new programming model appear?
73. Will any new programming model gain any traction?
74. Will there be any significant improvement in the capacity of classical quantum simulators? How many qubits and what maximum circuit size? What capacity can the average quantum circuit designer depend on for circuits with say 250, 500, 1,000, 2,000, or 2,500 gates — 16, 20, 24, 28, 32, 36, 40, 42, 44, 46, 48, 50 qubits, or what?
75. Will there be any significant improvement in performance of classical quantum simulators?
76. Will there be much progress implementing Shor’s factoring algorithm? What might be the largest semiprime that can actually be factored by Shor’s algorithm on a real quantum computer or even on a simulator?
77. What progress might be made on fine granularity of phase and probability amplitude? Which qubit technologies will do best at it? What granularity will become most common and a de facto standard that quantum algorithms can rely on, such as for quantum Fourier transform and quantum phase estimation? When will any vendors document what granularity they support for their qubits and gates?
78. Will open source thrive and expand? Including hardware as well as software, algorithms, applications, and tools. Will people be truly committed to it, across the board, or only pay lip service when it’s convenient and offers the vendor some opportunity for direct financial profit?
79. Will the U.S. Department of Energy (DOE) national laboratories produce any truly dramatic research breakthroughs to boost commercial quantum computing efforts?
80. Will the U.S. DOE national labs introduce any partnerships with academic research labs and commercial firms that have a real potential for dramatic breakthroughs in quantum computing research? Either in the near term, medium term, or longer term.
81. Will academic and government research labs be able to lure talented researchers back from financially tempting but technically boring or unproductive positions at commercial firms? We need more research, a lot more research. What we don’t need is premature commercialization, such as we’ve seen over the past five years. Let the researchers do the kind of research that they excel at and that we desperately need!
82. Will Congress continue to up the ante for funding of research in quantum information science? And make sure enough of that gets to quantum computing.
83. Will investment and funding for quantum computing from private venture capital, government labs, government grants, and internal corporate investment continue to grow? Will it grow faster, a little faster, a lot faster, or about the same, or will it actually decline, by a little or a lot?
84. Will NIST take a more active role in quantum computing, or will their role continue to diminish? The DOE national labs seemed to be picking up a larger role even as the role of NIST has diminished. NIST has a strong role in cryptography and quantum sensors and measurement, but not in quantum computing per se, which is odd since Dave Wineland of NIST won the Nobel Physics prize in 2012 for his work in quantum computing. Maybe the DOE national labs are better positioned since they have a greater need to actually use quantum computers, at least for scientific applications, but I still think NIST should have a stronger role, especially for commercial applications — since NIST is an arm of the Department of Congress and their mission is to… “promote American innovation and industrial competitiveness.” Or Maybe DOC needs a research arm for commercial computing separate from NIST.
85. Will IBM put out a revised quantum roadmap? I presume so, as they have done in the past two years.
86. Who will have the most detailed quantum roadmap?
87. Who will make the most progress on their quantum roadmap?
88. How will organizations involved in the design of quantum algorithms and the development of quantum applications react if insufficient progress occurs on the hardware front?
89. How will budgets for quantum computing evolve for organizations involved in the design of quantum algorithms and the development of quantum applications?
90. Will people stop referring to democratizing quantum? This is a poor and offensive use of an important term from the political sphere. Typically all people mean is assuring access, especially low-cost access. Nothing to do with democracy or politics — or putting the people in charge.
91. Will more physicists give up on mediocre near-term general-purpose quantum computers in favor of special-purpose quantum computing devices or software-controlled physics experiments? Analog-style computing is not as general, precise, and powerful as general-purpose computers, but can be acceptable if they get the job done (calculating approximations) while we all wait for quantum computers to become a lot more functional with higher fidelity qubits and full any to any qubit connectivity.
92. Who will see the greatest progress in quantum computing, the U.S., Europe, or China? Yes, there are other, smaller regions and countries, including Australia and Japan — and others.
93. Will India be a relatively smaller player in quantum computing in 2023 or zoom up to be a top-tier player? Will they chart their own quantum course, or relegate themselves to being outsourced contractors for the global top-tier quantum firms?
94. Might we see at least a hint of an intellectual property winter as patents begin to interfere with development of quantum computing products? Might onerous licensing fees or blocking patents cause a pause in product development as licenses are negotiated or while technical workarounds are researched and developed?
95. Will published papers on quantum algorithms and applications pay a lot more attention to rigorously analyzing and quantifying shot counts (circuit repetitions) for varying input data and parameter values?
96. Will published papers on quantum algorithms and applications pay a lot more attention to rigorously analyzing and characterizing scalability for varying input data and parameter values?
97. Will we tumble into a Quantum Winter? No, or at least very unlikely. We should have a free pass to explore possibilities for quantum computing technologies, products, and services on a broad basis for at least the next year, 15 months, or maybe even 18 months or even two years or even a little longer before we run the real risk of a Quantum Winter. See my paper for a deeper discussion: Risk Is Rising for a Quantum Winter for Quantum Computing in Two to Three Years. That was written almost nine months ago, so the clock is ticking.
98. Will being a full-stack quantum company be a feature or a bug? Is the broad diversity of being full-stack — hardware, software, tools, algorithms, and applications a key strength, or is it a distraction and loss of focus that risks undermining the breadth and depth of the advantages of a full-featured quantum ecosystem? Maybe a very few firms can thrive as full-stack, but even then I am skeptical. We certainly don’t need to see a large number of full-stack quantum computing firms.
99. Will the quantum ecosystem continue to evolve in a very strong, very healthy, very diverse, and very vibrant manner? Will it be even stronger, healthier, diverse, and vibrant? Or will some decay and confusion set in?
100. Will Intel or some other vendor begin producing components so that others can easily build their own quantum computers? See my proposal: Call for Intel to Focus on Components for Others to Easily Build Their Own Quantum Computers.
101. Will co-design of quantum hardware and algorithms be a big deal? Will it be more popular, about the same popularity, or less popular? It may have been more popular in the early days when machines were significantly less capable, but as general-purpose quantum computers get more capable, co-design may have less to offer.

This is a live list and will continue to be updated. New entries will generally be added at the end of the list.

For more of my writing: List of My Papers on Quantum Computing.