My Quantum Computing Wish List for Christmas 2020 and New Year 2021
Here are my top wishes for developments in quantum computing for Christmas in 2020. Okay, that’s too tall an order with too little time left, so this informal paper lists the quantum computing developments I really want to see in the coming year, 2021. It’s a fun list, but it’s also a real list — these are advances that have a realistic chance of occurring over the coming year.
For reference, here’s my Christmas wish list from last year:
I still want everything from my 2019 wish list since so little of my wishes came true, but this year I tried to come up with a more abbreviated list which was a little more practical and achievable.
All I want to see for Christmas in 2020 and in the New Year 2021 are…
My Top 5 Christmas 2020 wishes
- Greater progress towards full, automated, and transparent quantum error correction (QEC) — also known as fault-tolerant quantum computing (FTQC). More research papers for sure, but some experimental work as well. By the end of the year (2021), there should be a sense that we could have full QEC for an 8-qubit (logical qubits!) machine within another two years. People begin to talk more seriously about moving beyond NISQ sooner (within 2–3 years for 8 to 16 qubits.) More algorithm papers focus on use of logical qubits.
- Algorithms exploiting 32–45 qubits for practical applications. Published papers.
- Push the upper bound for classical quantum simulators by at least a couple of qubits, hopefully to 45 qubits. Try to make 24, 32, and 40-qubit simulation much more common and readily (and cheaply) accessible.
- Quantum volume of at least 4096 on at least one machine. Actually measured, not simply estimated. That’s 12 qubits for at least a circuit depth of 12 gates. Trapped-ion machines may do a lot better (need to be actual, measured results!), but superconducting transmon qubits need to push to reach this 4096 goal — the raw qubits are there, but coherence and gate errors (including swap networks) are a challenge. If not 4096, at least 2048. If not 2048, at least 1024.
- 100+ qubits. At least one machine in the 100+ qubit range.
My other Christmas 2020 wishes
- Much more fundamental R&D. Particularly for new qubit technologies and new programming models. And basic and advanced algorithms, algorithmic building blocks, and application frameworks.
- Another doubling of government R&D spending on both quantum information science (QIS) overall and quantum computing in particular.
- At least four or five notable new hardware vendors. Or at least two or three.
- At least one more trapped-ion vendor with a real machine and published results or at least a proposal to produce one within a year.
- A 40 or 50 or even 64-qubit trapped-ion quantum computer. Is there a good (or better yet, great) reason that trapped-ion machines should be playing catch-up rather than leading the way?
- Much more transparency. Public technical specifications and documentation. Quality needs dramatic improvement.
- New capacity and performance benchmarks to replace quantum volume, especially for quantum computers with more than 50 qubits.
- Quantum volume performance of at least 256 on more than a couple of machines. That needs to become the new low-end minimum expected performance.
- Record and publish a detailed full report for each claim of quantum volume. Detail qubits and circuit depth, and exactly what the limiting factors are. Report swap network overhead, and whether that is the limiting factor. Report all factors used or calculated for each claim of quantum volume. Detail all configuration settings, parameters, and tuning (static and dynamic) needed to complete each claim of quantum volume.
- Dramatic increase in circuit depth. Dramatic increase in coherence time and reduced gate errors, including swap networks needed to overcome connectivity issues for machines without native full connectivity.
- At least one more machine in the 64 to 72 qubit range.
- Many more GitHub repositories, complete with technical specifications, full documentation, sample data, and sample results, for quantum programs, algorithms, and applications. Especially hybrid applications. This should be mandatory for published papers claiming experimental results.
- An order of magnitude reduction in hype. Right, like that’s going to happen! Still, hope does spring eternal!
- All published papers available on arXiv. None hidden behind paywalls or even requiring registration to access.
- Stop using Grover’s search algorithm as an example of success in quantum computing. A mere quadratic performance improvement won’t get us to dramatic quantum advantage. Anything less than full exponential improvement is not worth writing and talking about.
- Published papers should cease and desist from using Shor’s algorithm for factoring large semiprime numbers as an example of the success of quantum computing. Maybe someday Shor’s algorithm will indeed be able to be implemented to crack 4096-bit public encryption keys, but that is the distant future, not the here and now, not the next few to five years.
- A decent Hello World quantum circuit, which actually does something readily apparent and practical and useful — something non-quantum technical staff can immediately relate to. Also needs to demonstrate at least a fraction of quantum advantage — easy to see how it could be scaled — without redesign of the source code — from 4 or so qubits to 8, 16, 24, 32, 40, 50, 64, 72, an 96 qubits as hardware advances. Lay off the quirky algorithms that have no practical application or can’t be implemented on near-term hardware. Needs to be something that really impresses people, like “Wow, a quantum computer can actually do THAT?!!”
Reviewing my old 2019 Christmas wish list
For reference, here are some of my wishes from last year, 2019, not as formal wishes for this year, but as disappointments from last year:
- Will Rigetti ever get their act together and produce a larger machine, if not the 128 qubits they promised over two years ago, at least 64 or 48 qubits? Guys, go big or go home! What’s the problem? Your competitors are leaving you behind in the dust!
- Still waiting for Google to deliver their 72-qubit machine. Or at least something beyond 53 or 54 qubits.
- Still waiting for IonQ to deliver a 79-qubit machine — or anything over 20 to 32 qubits.
- Honeywell has delivered 6 to 10 qubits, but when will they break out and compete with IonQ and IBM, like with 16, 20, 32, 48, or 64 qubits?
- Some announced progress towards a semiconductor-based quantum computer by Intel.
- Some announcement of progress towards a general-purpose photonic quantum computer by Xanadu.
- At least one published paper for quantum phase estimation or quantum Fourier transform for at least 8 qubits of input data and at least 8 bits of result precision. Preferably on a real machine, but simulated at a minimum. Will we have to wait for full quantum error correction (QEC), or will incremental qubit advances get us within reach much sooner, possibly even on current trapped-ion machines?
- At least one published paper on simulating the quantum chemistry of a molecule more complex than water and nitrogen. Something requiring more than 20 qubits and more than 150 gates.
- Several free, online, high-quality books on quantum computing and algorithm design. Bootleg PDFs don’t count.
- More free, online and high-quality lecture notes on quantum computing.
- At least one published paper for Shor’s algorithm to factor a semiprime larger than 15 and 21, like 6 or 7 bits, on a real quantum computer. The full, original Shor’s algorithm, not some stripped-down derivative customized for particular input data.
- At least one published paper for quantum phase estimation or quantum Fourier transform for at least 16 qubits of input data and at least 10–16 bits of result precision. Preferably on a real quantum computer, but simulated as well. Will we have to wait for full quantum error correction (QEC), or will incremental qubit advances get us within reach much sooner, possibly even current trapped-ion machines?
Incremental progress? Boring!
Seriously, we really do need simply a lot of relatively boring but incremental progress in so many areas. Incremental progress is very needed, but is also… very boring!
I’ll let other people settle for incremental progress, like 20% to 30% improvements over a year, but I’ll focus my wishes on dramatic gains and outright breakthroughs.
Seriously, there must be at least a few serious researchers out there who can produce something which really surprises me. Just do it!
That’s it! These are all I ask for for Christmas 2020 and the coming New Year 2021. Is that a lot to ask?
I sincerely expect many of those developments to occur in the coming year, but I’ll be quite satisfied if even half of those developments come to pass over the coming year.
Need commitment for 2021
Even if few of my wishes can be granted by Christmas 2020 or early in New Year 2021, what I really would like to see is widespread commitment to tackle all of my wishes in 2021. Maybe not all of them will come to fruition, but I expect that quite a few will, or at least come close.
Still a mere laboratory curiosity
I know people are working really hard and great progres is being made, but even with all of my Christmas and New Year wishes, quantum computing will still remain a mere laboratory curiosity as 2021 ends and 2022 begins. We probably need another two or three years. And we probably need full, automated, transparent quantum error correction — or qubits which are close enough to perfect that QEC is not needed for many algorithms and applications.
For more of my writing: List of My Papers on Quantum Computing.