Who is excited for Jensen to host this panel? Using this space to discuss the conversation.
#Quantum Computing: Where We Stand [S74495]
sharp fulcrum
timber pumice
I am excited!!
sharp fulcrum
Quantum Computing: Where We Stand [S74495]
timber pumice
I am excite, unfortunately I am not able to come in person
dreamy pelican
It will be live streamed, so come her and join us for a watch party in less than half an hour!
dreamy pelican
Coming up in just a few minutes:
https://register.nvidia.com/flow/nvidia/gtcs25/vap/page/vsessioncatalog/session/1736991707221001VCoU
dreamy pelican
If you know Jensen a bit, then you just have to believe that he actually meant it that he wasn't aware that (some) of the QC companies were public already.
So that is a really nice move to invite them here and take the heat 🙂
https://developer.nvidia.com/cuquantum-sdk and don't forget to check out #1080016198227468378
Ouch, he should have seen that one chair was still empty.
I don't pretend to understand how QC Hardware is set up, but it seems that even within this small selection of companies there are already quite a few different approaches to the technology. Even if the differences are small.
Oh Wow! i did not know that "trapped Ions" was already done in 1995!
I wonder whether he will ask about Quantum Supremacy at some point. But "Usefulness" is a good one as well.
What does it mean if they say "the next few years"? 5? 10? 20?
That is a nice line of discussion. Trying to differentiate between classical computing and QC in terms of applications is an important step for more people to understand the merit of QC.
2^28 increase in performance by the end of the year? Wh00t? I understand why people are sceptical.
Found a blog post about LS-DYNA from November, as context to what the IonQ chair said. https://developer.nvidia.com/blog/advancing-ansys-workloads-with-nvidia-grace-and-nvidia-grace-hopper/
Well, QC helping AC instead of the other way round sounds like quite the challenge. And closing in on AGI with the help of QC within a year?
I like the answer that one simply hopes for less scepticism. That should work out.
Nice touch that they now pull out actual Hardware from their pockets just like Jensen likes doing.
I like how Seeqc tries to bridge his tech presentation to things from Jensen's keynote. That way he makes things much more approachable to newbies to this field.
100 logical (error corrected) Qubits, meaning at least 10,000 physical Qubits or more. Wow! Seeqc really has a good explanation of what an important goal should be, integrating so that they reduce the amount of engineering problems to solve.
I really like the Seeqc guy. Making a reference to ENIAC which maybe only half of the audience even knows, nice touch. 🙂
Requirement for fast enough global error correction by GPUs: 500-800 nanoseconds latency! Insane!
Pete Shadbolt has a good point. If the same effort is put into QC development like was put into semiconductor development to scale, shrink and make production more efficient, it should really not be a multi-decade problem to scale QC up by 4 or 5 orders of magnitude.
Seems really like they are all in line with expectations. Rapidly scale up error correction and logical Qubits while still trying to create revenue by finding applications for QC.
Jensen's appreciation of the insane complexity of QC is really a nice touch.
normal fjord
Jensen's appreciation of the insane complexity of QC is really a nice touch.
Passion is key. The rest follows easily. That's why Blockchain and DeFi haven't been mass adopted yet. People gen pop/NPCs just see it as an investment which is truly sad. Do keep in mind that NVIDIA has a monopoly though. I'm fine with it. NES and NDLI take care of my needs as I progress.
dreamy pelican
The passion was really tangible during the first two parts of the panel. The engagement feels even more "electrified" than with LLMs right now.
normal fjord
The passion was really tangible during the first two parts of the panel. The eng...
There's a lot of sensationalism as well but it truly is on par with the Internet and dare I say the IR thinking of this long-term. Where was this in Grad School: https://developer.nvidia.com/amgx?ncid=em-nurt-245273-vt33
dreamy pelican
Is the expectation to outperform classical compute solutions with 100 logical Qubits based simply on hypothetical extrapolation or something more specific?
dreamy pelican
There's a lot of sensationalism as well but it truly is on par with the Internet...
IR thinking?
normal fjord
I'm not entirely sure I can answer that with certainty, sorry. I'm currently facing a dilemma between a dedicated workstation or just dual birfurcated 5090s. The Dyson BP04 is a godsend. I'm also not too familiar with AMD chipsets but I'll try it (outside of gaming).
dreamy pelican
"How many Qubits do you have?" Good one.
Great panel! A lot of interesting and helpful information shared by all involved.
small gate
I missed watching this session. 
Hopefully, there's a replay soon. So far, nothing on the GTC website yet.
high kindle
It seems that #quantum-computing channel disappeared. Is it true?
dreamy pelican
It seems that #quantum-computing channel disappeared. Is it true?
Yep, we archived it as part of the general clean-up we did a short while ago. #announcements message