AI buildout is now bottlenecked by DRAM, GPUs and packaging
tengyanAI · x · 2026-07-20
Tessara’s weekly AI buildout update says the physical supply chain is now the real bottleneck, not the models themselves.
- 35 of 54 tracked constraints are rated tight or worse, or 65% supply stress overall.
- The single most binding constraint is DRAM DDR5, scored at 86 and marked critical.
- Other critical pressure points include Blackwell GPUs, OSAT packaging, power transformers, FC-BGA substrates, NAND flash, and HBM3e.
- Two additional tight areas called out are indium phosphide and Chinese AI accelerators.
More from Infra
- Open-source Latch compiles PyTorch into training megakernels, 2.75x faster on H100 — sguada · 2026-10-27
- NVIDIA Claims HGX Rubin NVL8 Delivers 10x More Tokens per Megawatt Than B200 — thursdai_pod · 2026-10-27
- Commoditization dilemma: either human labor or GPUs become the commodity — _AustinCalvert_ · 2026-10-27
- DeepSpeed shows AutoEP/AutoTP/AutoSP parallelism gains, hosts office hours — PyTorch · 2026-10-27
- Running a local 9B as a coding agent: brief the big model to plan, let the small model execute — Electrical_Ninja3805 · 2026-10-27
- Cloudflare Durable Objects validated as top architecture for AI agent WebSocket servers — tristanbob · 2026-10-27