32 ms·

> full-stack quantum computing Anyone familiar with the term? Haven't heard it before. > leader in scalable quantum processor technology I thought quantum co

> full-stack quantum computing Anyone familiar with the term? Haven't heard it before. > leader in scalable quantum processor technology I thought quantum computing was still in the research phase? What's there to scale?

cameronperot 5y ago
> Anyone familiar with the term? Haven't heard it before. Full-stack probably means they deal with all areas of QC, from theoretical research to hardware implementation. > I thought quantum computing was still in the research phase? What's there to scale? IMO it is still largely in the research phase, but actual quantum computers do exist. The thing to scale would be the number of physical and logical (error-corrected) qubits. Quantum error correction is difficult though, and will take a lot of work to get scalable QCs with error correction. With some of the current schemes being pursued today, the number of physical qubits per logical qubit is quite high (e.g. 17 physical per one logical in Surface-17).
krastanov 5y ago
"Full-stack" is not an official term, but it is a reasonable way to refer to "working on both creating the low level hardware (e.g. transmons), the intermediate layer (error correction that makes the many physical qubits into a few reliable logical qubits), and the higher levels of algorithms for useful applications". There is no "scalable" quantum computer in existence yet. If we can make a million qubits of quality 10x better than the ones we have now, then we will definitely be able to talk about scaling and useful applications. Today we have machines with 10s to 100s of qubits that are extremely exciting technology demonstrators, but too small and too noisy to fulfill any of the promises of quantum computing from the 90s and 00s.
- 1001101 5y ago
  I would imagine Shor's algorithm adds some friction to scaling, and who can be permitted to use scaled systems in the near term.
  - krastanov 5y ago
    I do not see how Shor's algorithm, just one among my quantum algorithms, brings anything to the question of scaling.
    - krastanov 5y ago
      Pardon the typo: "many", not "my"
    - 1001101 5y ago
      No worries on the typo, and sorry if I said something dumb :) - I know I have something to learn here. I'm a layperson whose knowledge is basically the first chapter of an introductory undergrad text (although when I have time, I will dig into more). Perhaps I didn't get the terminology right, but I was thinking in terms of their total addressable market, depending on the quantity and quality of the qubits, they wont' be able to hand the keys to the castle of public key cryptography (so Shor's algorithm) to just anyone. Maybe that's baked into their business plan. Then I thought of some ways one might still probably get a bit done in this respect with a smaller number of qubits if SAT is in BQP (very half-baked idea), but from what I googled, it sounds like that's an open question.
      - krastanov 5y ago
        I can not know for sure, but I doubt that this worry is on their mind. By the time these machines are reliable enough to do prime factorization, we would have had plenty of opportunities to move to "post quantum" public cryptography schemes that are as resilient to quantum attacks as they are to classical (cloudflare and google chrome have already implemented experimental HTTPS variants that are resilient). I doubt that Shor's algorithm would ever be a "killer app", because only a small fraction of encryption algorithms are susceptible to such an attack (it is not too far fetched to say that it is just a historical happenstance that our standard public encryption schemes happen to be susceptible).
        1001101 5y ago
cblconfederate 5y ago
they are going with the flow. I bet QuantumOps generalists will be a thing too.

cameronperot 5y ago

> Anyone familiar with the term? Haven't heard it before. Full-stack probably means they deal with all areas of QC, from theoretical research to hardware implementation. > I thought quantum computing was still in the research phase? What's there to scale? IMO it is still largely in the research phase, but actual quantum computers do exist. The thing to scale would be the number of physical and logical (error-corrected) qubits. Quantum error correction is difficult though, and will take a lot of work to get scalable QCs with error correction. With some of the current schemes being pursued today, the number of physical qubits per logical qubit is quite high (e.g. 17 physical per one logical in Surface-17).

krastanov 5y ago

"Full-stack" is not an official term, but it is a reasonable way to refer to "working on both creating the low level hardware (e.g. transmons), the intermediate layer (error correction that makes the many physical qubits into a few reliable logical qubits), and the higher levels of algorithms for useful applications". There is no "scalable" quantum computer in existence yet. If we can make a million qubits of quality 10x better than the ones we have now, then we will definitely be able to talk about scaling and useful applications. Today we have machines with 10s to 100s of qubits that are extremely exciting technology demonstrators, but too small and too noisy to fulfill any of the promises of quantum computing from the 90s and 00s.

1001101 5y ago
I would imagine Shor's algorithm adds some friction to scaling, and who can be permitted to use scaled systems in the near term.
- krastanov 5y ago
  I do not see how Shor's algorithm, just one among my quantum algorithms, brings anything to the question of scaling.
  - krastanov 5y ago
    Pardon the typo: "many", not "my"
  - 1001101 5y ago
    No worries on the typo, and sorry if I said something dumb :) - I know I have something to learn here. I'm a layperson whose knowledge is basically the first chapter of an introductory undergrad text (although when I have time, I will dig into more). Perhaps I didn't get the terminology right, but I was thinking in terms of their total addressable market, depending on the quantity and quality of the qubits, they wont' be able to hand the keys to the castle of public key cryptography (so Shor's algorithm) to just anyone. Maybe that's baked into their business plan. Then I thought of some ways one might still probably get a bit done in this respect with a smaller number of qubits if SAT is in BQP (very half-baked idea), but from what I googled, it sounds like that's an open question.
    - krastanov 5y ago
      I can not know for sure, but I doubt that this worry is on their mind. By the time these machines are reliable enough to do prime factorization, we would have had plenty of opportunities to move to "post quantum" public cryptography schemes that are as resilient to quantum attacks as they are to classical (cloudflare and google chrome have already implemented experimental HTTPS variants that are resilient). I doubt that Shor's algorithm would ever be a "killer app", because only a small fraction of encryption algorithms are susceptible to such an attack (it is not too far fetched to say that it is just a historical happenstance that our standard public encryption schemes happen to be susceptible).
      - 1001101 5y ago

cblconfederate 5y ago

they are going with the flow. I bet QuantumOps generalists will be a thing too.