John Smith

@JSmith_Crypto

computer science, quantum computing, and cryptocurrency! The future of crypto is

Vrijeme pridruživanja: prosinac 2018.

Tweetovi

Blokirali ste korisnika/cu @JSmith_Crypto

Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @JSmith_Crypto

  1. Prikvačeni tweet
    7. kol 2019.

    1/ The model for projecting when quantum computers will break 's 256-bit ECDSA is complete. You can find the code at . I have attached some results for assorted parameter settings. You will need Matlab to run it for now, but a web app is coming!

    Prikaži ovu nit
    Poništi
  2. 27. kol 2019.
    Poništi
  3. 18. kol 2019.
    Prikaži ovu nit
    Poništi
  4. 18. kol 2019.

    Idk guys I think maybe we can pull it off!!

    Prikaži ovu nit
    Poništi
  5. 18. kol 2019.

    The great once decreed the "non-BS" timeline for quantum comps becoming a security threat to be 50-100yrs. Well, I checked and... For lattice surgery, that means we'll need ~5% more qubits ~1% lower gate error ~3% better algorithm /year!

    Prikaži ovu nit
    Poništi
  6. proslijedio/la je Tweet
    11. kol 2019.
    Poništi
  7. 12. kol 2019.

    The legend just hit 1,000 followers on Twitter after joining 2 months ago. I predict a quadratic speedup on his way to 2,000

    Poništi
  8. 7. kol 2019.

    9/ That's all for now. Happy to answer any methodological questions.

    Prikaži ovu nit
    Poništi
  9. 7. kol 2019.

    8/ Some notes on the parameters: 100% increase per year = doubling 50% decrease per year = halving. algorithmic improvement is a discounting term, so think of 50% as halving the cost of the calculation per year. Uncertainty represents 1 Gaussian standard deviation

    Prikaži ovu nit
    Poništi
  10. 7. kol 2019.

    7/ Reply to this tweet with the values of: >% increase in # qubits per year (>= 0) >% decrease in physical gate error per year (>= 0, < 100) >% algorithmic improvement per year (>= 0, < 100) >% uncertainty about the above estimates (>= 0) & I will reply with the model's results

    Prikaži ovu nit
    Poništi
  11. 7. kol 2019.

    6/ Finally, since the web app is not up yet, and many of you likely do not have Matlab, if you reply to the next tweet with a requested combination of parameters, I will run the model for you and attach the resulting graphs, when I have time, for ~1wk after posting.

    Prikaži ovu nit
    Poništi
  12. 7. kol 2019.

    5/ One interesting takeaway from this exercise for me was that QCs actually do not need to double in power every year per se to threaten ECDSA in as early as 10-20 years from now.

    Prikaži ovu nit
    Poništi
  13. 7. kol 2019.

    4/ This modeling tool is still quite useful for those interested in cryptocurrency, for obvious reasons...

    Prikaži ovu nit
    Poništi
  14. 7. kol 2019.

    3/ I decided NOT to include any additional Bitcoin risk-assessment component to this. The details of Bitcoin's migration to QR sigs remain unclear, & modeling it at this stage would be highly imprecise. I also thought it would be better to just have one main purpose per script.

    Prikaži ovu nit
    Poništi
  15. 7. kol 2019.

    2/ This model takes your guesses for how QCs will scale (# of qubits, gate error, algorithmic improvements, & uncertainty), & applies them to the surface-code approach by Gidney & Ekera (2019) to generate a probability distribution for when ECDSA will be broken

    Prikaži ovu nit
    Poništi
  16. 5. kol 2019.

    What are the chances that a large-scale Quantum Computer already exists?

    Poništi
  17. proslijedio/la je Tweet
    3. kol 2019.

    I agree with everything in this! Would also suggest the article title is analogous to "Someone proved P=NP. It won't change the world". Q supremacy will be a massive scientific achievement, which changes the world in itself, even w/out practical applications (that may come later)

    Poništi
  18. 31. srp 2019.

    14/ Anyways, this model can be updated as time goes by, to generate predictions with tighter confidence intervals. And the second part is coming soon. That's all for now! [end]

    Prikaži ovu nit
    Poništi
  19. 31. srp 2019.

    13/ One interesting thing I've noticed from this exercise so far, though, is that it's not that easy to see threatening QCs coming from more than a couple of years out if all you do is set a size threshold. The risk tends to go, metaphorically, from "0 to 100 real quick”...

    Prikaži ovu nit
    Poništi
  20. 31. srp 2019.

    12/ Not sure what I'll find once I model Bitcoin’s adaptation process. Maybe the most realistic parameters I can come up with will actually give a 0% chance of things going south. We’ll see. You can audit for yourself (& plug in your own assumptions) once the code is up.

    Prikaži ovu nit
    Poništi

Čini se da učitavanje traje već neko vrijeme.

Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.

    Možda bi vam se svidjelo i ovo:

    ·