Having Enough Compute Doesn't Guarantee a Top-Tier Model
teortaxesTex · x · 2026-07-14
This discussion suggests that players like DeepSeek don't find it hard to secure the compute needed for a "10T" scale pretraining model, as the total volume required isn't as massive as imagined.
The real challenge isn't "can it be trained," but rather:
- What kind of model is actually worth such compute;
- How much it will cost to figure out an Anthropic-level recipe;
- Essentially, it's about everything besides compute: training recipes, data, and engineering experience.
Related event: Post-Training, Not Compute, is the Real Bottleneck for Top AI Models(2 posts)→
More from Infra
- Nothing phone mockup turns a film joke into a modular design meme — ZeYanjie · 2026-07-22
- Actual Computer says its inference stack is tuned for Nvidia’s consumer Blackwell lineup — markjeffrey · 2026-07-22
- Ben Bajarin says CPU demand is still being badly underestimated — BenBajarin · 2026-07-22
- An energy model says the U.S. could run short of natural gas starting in 2028 — churchkey · 2026-07-22
- Arbitrum fee simulation shows higher gas capacity but lower L2 revenue under ArbOS61 — tomwanhh · 2026-07-22
- NVIDIA pushes OpenUSD as the common layer for simulation and physical AI — MonaJalal_ · 2026-07-22