Pinned Tweet
B300s Available in Q4.
128 to 256 nodes (1,024 to 2,048 GPUs). Liquid or air.
RFS December. 3-5 year offtake.
DM for rates.
every company ive spoken with this month has answered compute when asked about their biggest stressor and what keeps them up at nights
sourcing compute, cost of compute, compute compute compute
2
1
12
4,779
McDavid retweeted
Today I call upon the blockchain industry to calmly begin planning for "bunker mode". My personal recommendation is to set in motion a controlled mass migration of assets to fresh addresses, i.e. addresses whose pubkeys remain hidden behind a hash.
Holders, starting with large and sophisticated ones, should consider moving the bulk of their funds to addresses that have never signed a transaction. And when they do sign one, they should also move remaining funds to a new address (possibly generated from the same seed phrase).
Don't rush. While I believe there is cause for action a rushed migration would do more harm than good. Don't panic either. Moving assets to protected addresses is a simple, preventative step which does not require new cryptography or new wallets.
IMO it is now reasonable to brace for the possibility that ECDSA breaks before qday, in the worst case in months not years. By "break" I mean fast private key recovery (e.g. in one week) on available hardware (e.g. a large GPU cluster).
Recent days have been humbling for human mathematical intuition. Long-held, unquestioned hypotheses have fallen. This includes the n log(n) bound for integer multiplication and the 3SUM conjecture. In hindsight, May's unexpected disproof of the Erdős unit distance conjecture was our warning shot.
Yesterday's OpenAI drop made it clear that mathematical superintelligence is upon us. They say there are weeks where decades happen. We are about to live through weeks where centuries of mathematical progress happen. Could our magic 64-byte ECDSA signatures be too good to be true? Was it just security through obscurity all this time?
Elliptic curves feel especially vulnerable to superintelligence. Curves carry rich structure, with room for fancy tricks like Schoof, Frobenius, pairings. (By contrast, hashes are designed to minimise algebraic structure.)
Separately, as Ewin Tang can attest, an efficient quantum algorithm sometimes foreshadows an efficient classical one. We should be open to the possibility of a classical counterpart to Shor that breaks elliptic curves and RSA at once.
Also noteworthy is the striking under-representation of cryptographic breakthroughs among the 722 mathematical results OpenAI published. I've witnessed first-hand the US government censoring academic quantum cryptanalysis results. Backroom interventionism is my base case.
I urge large, sophisticated actors to lead by example. Project11's "risq list" (bitcoin-risq-list.projecteleven[.]com) is a great tracker of exposed BTC pubkeys. Binance, Bitbank, Robinhood, Bitfinex, and Tether have an opportunity to harden their cold storage. Next month I'll address institutions in London in a live Q&A (forum.ethereuminstitutional[.]org/london-2026).
Again, please do not rush. Wallets holding under 50 BTC enjoy partial cover from "Satoshi's shield", i.e. his 20K exposed addresses that hold 50 BTC each. Load-bearing signers like oracles and L2 security councils should consider rotating ECDSA pubkeys with every signed message and/or multi-signing with a hash-based schemes like SPHINCS.
Exiting bunker mode safely will require post-AI cryptography. My inclination is to go all-in on hash-based cryptography and avoid structured mathematical assumptions entirely, whether from curves, lattices, or isogenies. A single battle-tested hash (e.g. from the SHA or BLAKE families) yields plausible post-AI security.
The Ethereum roadmap on strawmap[.]org fully embraces hash-based cryptography with end-to-end formal verification as a response to the quantum threat. Those timelines must now be revisited and accelerated in light of mathematical superintelligence. I'll be pushing for maximum defensive acceleration.
757
1,284
9,392
2,055,300
McDavid retweeted
Data center lead times getting beyond control
As shared by Cushman & Wakefield / Hyundai Securities, it’s not down to GPUs or even memory.
- Copper, aluminum: ~4–10 weeks
- Most HVAC / chillers: ~15–30 weeks
- UPS, LV/MV switchgear: ~35–60 weeks
- Generators: up to ~100 weeks
- Pad-mounted transformers: up to ~115 weeks
One can construct building faster than getting the transformer that powers it.
#DataCenters #AIInfrastructure #SupplyChain
ALT Made with AI
28
116
508
74,459
neoclouds differentiate heavily along these 3 pts:
1. finding and navigating the risks of new power supply (permits, lead times, etc)
2. sourcing equity capital (prepay isnt enough)
3. managing SLAs
contrary to popular belief, there's lots of compute supply in the market. what's interesting is that since the buyers are providing all the credit (and most of the time, a significant portion of upfront capital as well), there's little differentiation among the dozens and dozens of neoclouds and new operators entering the market. what are they even providing that their next competitor can't, if not the risk? which makes the reflexivity and rocketing prices (for 3yr+ terms) all the more puzzling. i don't think this will last
more buyers are taking matters into their own hands: the ones that can acquire cheaper cost of capital than new neocloud operators will seek to build their own capacity. many of them even enjoy a far shorter payback period than the operator ever would!
169
Congrats team!!
I’m excited to share that today @liquidcompute emerges from stealth with a $15 million seed round. We are announcing pending applications before the CFTC for Designated Contract Market (DCM) and Derivatives Clearing Organization (DCO) status to build the first regulated orderbook to trade both cash and physically settled contracts on compute.
The reality of compute is that it is heterogeneous and cannot be stored, functioning much like electricity rather than oil. Our approach focuses on developing a highly efficient short-term market to underpin a cash-settled derivatives layer, operating similarly to PJM or ERCOT sitting below liquid derivatives markets.
Over recent months, I’m also happy to announce that Liquid Compute has collaborated with leading financial firms like Susquehanna International Group, @BGCGroupInc and @wintermute_t to facilitate liquid OTC trading for AI startups, neocloud providers, and lenders.
This round was co-led by @chemistry and @FirstMarkCap , with participation from K8 Cap, Night Capital, @ufo_holdings, @TrueBridgeCP, Brainchild Holdings (@kalvepuri), @ycombinator, and angels such as Dmitry Balyasny, @taro_f , @JdotJdotF, Ethan Lin and others from companies like CoreWeave, Jane Street, OpenAI
4
1,403
why aren't more well capitalized players in need of compute putting the GPUs and debt on their own balance sheet? hyperscalers are doing this but the Series A that raised $100M+ isn't.
why is this?
if you need 256 nodes by EOY at a particular price.. you should hire an operator at $0.15 a GPU/hr and go and source a BOM and colo. it's not hard.
this is the best path to get the fastest RFS and the cheapest GPU/hr pricing.
4
208
McDavid retweeted
Claude:
🌟discombobulated for 45m15s
Two things, and the second is genuinely important.
Answering your question first — yes, things were broken, not working. But the real question is the reason why.
Before we go on, you have a call to make, and it’s genuinely yours: do you want me to do a handstand while juggling flaming bowling pins, or prepare a 7-course meal for a Nigerian family in Islamabad?
Codex:
52
225
6,613
361,244
We're brokering offtake on B300s (16-128 nodes). DM for details!
The Compute Desk Nvidia B300 GPU-hour index is at its all-time high. As margins on training grow thinner, neoclouds are collectively shifting focus to inference. High-priced B300s are generating ROI by lowering the marginal cost per token generated through batch inference.
2
8
851
congrats team
Itô (@itomarkets) helps companies buy compute at fixed GPU-hour rates and lets suppliers sell excess capacity to hedge tail risk.
$350k notional traded in a few weeks. 10+ compute providers. 3 hedge fund partners. Compute is becoming a commodity you can trade, not just a bill you pay.
Congrats on the launch @affaan and @absurdistphil.
alliance.xyz/launches/ito
2
230
I had a similar phase. Luckily I was 28. But I wanted it so bad. It would haunt my dreams. Back squats finally led to me getting this lone off the backboard jam on video.
i’m training to dunk a basketball
the odds don’t look good
48 yr old
6 ft
180 lbs
standing reach 91”
needed reach: 130”
that’s a 39” vertical
which exceeds ave nba player (35.5”)
here are the thresholds I need to meet:
> takeoff velocity 14.5 ft/s
> ground force: 585 lbs
> time to take off: 0.2 sec
> peak power: 14.8 hp (11 kW)
landing is even more intimidating
> creates load equal to 6-10x body weight
> 1,080 to 1,800 lbs of ground reaction force
injury risk
> 3-5x rupture risk for men my age
> 1,080-1,800 lbs funneled via centimeter thick tendon
> knees absorbing 7-9x body weight
age disadvantage
power+rate of force down 25-40% at age 48
what is your assessment, will I be able to do it?
2
8
1,953
McDavid retweeted
Turns out that if you do well in school and are at least moderately attractive, they just let you raw dog ADHD and autism until your thirties.
357
6,344
82,582
1,937,310
VVV has value does not mean token holders aren't being screwed and misled.
so VVV is like a physically settled compute future for compute on Venice and therefore has value.
ok! makes sense. VVV does have value.
however... Hasseeb is glossing over the fact that revenue, brand, and all the truly interesting value creation is going to accrue to equity holders and token holders are going to be left out on this ride. VVV should thus get rerated to what it actually is (a receipt for compute). and thus probably trade at a massive discount to its current value. inference costs will go down over time and be commoditized. VVV should thus trade at an increasing discount as time goes on. meanwhile equity holders will probably continue to see number go up.
14
13
4,431