What next for the compute crunch?
2 days ago
- AI compute demand is exploding, with GitHub commits up ~14x annualized, indicating massive inference growth.
- AI companies are compute-starved, leading to usage limits, outages, and a domino effect as users switch between services.
- Supply constraints include delayed GB200 rollout due to liquid cooling challenges, labor shortages, and geopolitical tensions.
- DRAM fabrication is a hard constraint, though innovations like TurboQuant provide temporary relief.
- The next 18-24 months will see persistent compute shortages, likely forcing price increases despite market share concerns.
- Users may become less price sensitive as models improve, and eventually supply will catch up, but short-term pain is expected.