quantumearl.bsky.social
Arch mage of quantumancy @ riverlane / Prof @ Sheffield uni / quantum error correction and algorithms / poster of posts / reader of sci-fi / player of games / father of daughters / cronic typer of typoes & requester of edit buttons
117 posts
625 followers
162 following
Regular Contributor
Active Commenter
comment in response to
post
Pray that @preskill.bsky.social does not release the teraquop
comment in response to
post
A MegaBot?
comment in response to
post
Nope, do you recommend?
comment in response to
post
I think we’ll stop posting about it within a few days because something else will happen that distracts everyone.
So yeah, it can go on a long time.
comment in response to
post
Chetan Nayak, in a comment on Scott Aaronson's blog, clarifies that this is the case and gives some details that I don't think were previously shared (that they see Z flips on a 10 ms timescale and X flips on a 5 us timescale). His March meeting talk should be interesting. scottaaronson.blog?p=8669
comment in response to
post
I am going to sleep now before you find any more typos.
comment in response to
post
Gah.,,, where is my edit button.
comment in response to
post
Fair point. I guess the pattern at play is that journalists have a few trusted experts and don’t realise that should broaden their menu and/or just read Bluesky. lol
comment in response to
post
It is hard to respond in real time on these things.
Every week there is at least one big claim that people ask for comment on. And if you read the paper you need to drop everything else in your day.
I can imagine that being a real time scientific commentator, and very accurate, is full time.
comment in response to
post
I can see you mention the PR in the article, but was this a quickly written story in response.
Or do these sorts of exchanges get co-ordinated days in advance?
comment in response to
post
This is a great news piece. Thanks for writing.
I am curious. Did you and Nature have prior access to the Microsoft press release?
comment in response to
post
I am with Davide in the comments. Scott is usually a good commentary but misses the punch line here.
Paper with 0 qubits; and MS PR claiming 8 qubits.
Even worse, a single Majorana pair is really 1/2 of a qubit, and there is no peer reviewed Majorana pair data.
comment in response to
post
…
4. A 1 pager on geopolitics of QC, and I agree the emerging nationalistic situation is unfortunate for science, but really a reflection of the wider world.
5. A 1 page interview with @preskill.bsky.social on the Megaquop machine, which resonances with @riverlanes.bsky.social mission.
comment in response to
post
1. Editorial about the need for companies to publish progress to moderate hype in the field.
2. A good 2 page overview of the pros and cons of quantum types like NA, TI and SC.
3. A 2 pager on PsiQuantum that ironically goes against the message in 1 and 2
….
comment in response to
post
Dear blue sky. Please add an edit button. Thanks.
comment in response to
post
www.nature.com/articles/d41...
comment in response to
post
www.nature.com/articles/d41...
comment in response to
post
Nice
comment in response to
post
Ah, someone with a keen eye. Care for a jelly baby ;)
comment in response to
post
Thanks. After growing my hair long my old hats are too small, so I had to switch up for something bigger and more eccentric.
comment in response to
post
This year is a bit different.
I have “prof” in my title;
the physics dept has merged with maths and chemistry to form a “school”;
Yingkai Ouyang will teach a few lectures in the module;
and I am doing 6 weeks in a row without taking a break to attend APS march meeting.
comment in response to
post
Nature Electronics is new to me and quite an interesting journal, with the cover art and editorial article also on the theme of quantum computing.
I also also learning that many of the hardest scaling challenges in QC are "just classical" problems, including electronics, calibration and software.
comment in response to
post
A conference on a Saturday!
comment in response to
post
There is that bit for break even, but further down it says 100x for a reasonable overhead (I.e. “well below threshold).
Caveat: I still haven’t read it cover to cover
comment in response to
post
Phys rev used to have a healthy discussion via comments but it seems to have dried up:
journals.aps.org/prl/authors/....
comment in response to
post
That’s what it says!
comment in response to
post
Right now, my main takeaway from this paper is that this technology does not look viable! (Despite the heroic efforts involved in the work)
The required improvement seem enormous. And even then we are talking a colossal number of server racks (millions?) for a full QC.
Is this a valid takeaway?
comment in response to
post
But the conclusions are telling me component need losses reduced by 100x on a decibel scale. If I look at Wikipedia to convert this to say a linear power scale then this is a 10 billion x reduction.
comment in response to
post
I’m back! Now having read the conclusions.
I can see there is a fully integrated system now. And something i’ve learned over the years is that integration is complex and hard. The armchair theortist’s mindset that “now you just plug bits together” is very flawed.
So congrats on that front.
comment in response to
post
Thanks for clarifications. What was the intention of the words “scale model” that threw me?
I may return in the morning with more questions :)
comment in response to
post
Reading the abstract on my phone. It sounds like a full demonstration but then they say “scale model” and showing “feasibility”. Which sounds like a simulation. Unclear from abstract what they have done…. For further reading.
comment in response to
post
Not quite ready for that
comment in response to
post
I just got my phone out and on auto pilot I opened the blue sky app like I used to open X on auto pilot