In-House Silicon

The custom-silicon revolt (TPU/Trainium/Maia/MTIA) — real but INFERENCE- only, anchored by Anthropic running mostly off Nvidia on Trainium + TPU. Nvidia's absolute demand still grows; the clearest winners are the co- design houses (Broadcom, Marvell). Watch whether displacement reaches frontier training.

STATUS · OPEN OPENED · 2026-07-24 LAST SEEN · 2026-07-25

Google Amazon AWS Microsoft Meta AI OpenAI Broadcom

2026-07-27 — Four programs, one wall ⟨crawl 2026-07-27⟩

TPU v7 Ironwood (Broadcom/MediaTek/TSMC co-design) · Maia 200 (dual-purpose, volumes undisclosed) · Trainium 1.4M deployed · MTIA “millions” H100-eq, Iris Sept-26. All four decline to state the own-vs-Nvidia split — the wall every capex crawl hit.

2026-07-24 — Opened + crawled

Thread opened (ben-steer). ⟨steer 2026-07-24⟩
Verdict: diversification at the margin and on inference, not a rout. The CUDA-vs-XLA software moat keeps Nvidia the default for anyone without a captive anchor customer; the “in-house” silicon is largely outsourced to Broadcom/Marvell. ⚠ WebSearch budget exhausted (fan-out) — primary-doc fetch path; Maia 200 + some codenames single-sourced, flagged. ⟨crawl 2026-07-24⟩

← Backstory (crawl 2026-07-24 → artifacts/findings/inhouse-silicon-2026-07-24.md)

2026-06 — Broadcom is the real winner

Broadcom’s custom-AI revenue hit $10.8B in a single quarter (+143% YoY), guiding $29.4B next and reaffirming “>$100B AI revenue in 2027”; ~6 hyperscaler XPU customers. Marvell (Trainium/Maia co-design) projecting ~$11B AI-ASIC revenue in 2026. The revolt enriches the co-designers. (Reuters) ⟨crawl 2026-07-24⟩

2026-04 — Google Ironwood + the Anthropic anchor

Ironwood (TPU v7) hit broad Cloud GA (04-22); combined with the Oct 2025 Anthropic–Google deal (up to 1M TPUs / >1GW in 2026, first phase 400k Ironwood). SemiAnalysis: Ironwood ~44% lower TCO than a GB200 server; Anthropic ~52% lower TCO/PFLOP vs GB300 — the economics behind the revolt. (SemiAnalysis) ⟨crawl 2026-07-24⟩

2026-03 — Meta MTIA: inference, not frontier

Meta’s MTIA 300 in production (ranking/rec), 400/450/500 on a 6-month cadence, “hundreds of thousands” deployed for inference — but explicitly NOT frontier-LLM training (Meta still buys Nvidia/AMD for that). The workload-specific pattern. (Meta) ⟨crawl 2026-07-24⟩

2026-01 — Microsoft the laggard

Maia 200 launched — inference-only (TSMC 3nm, claims 30% better perf/$); next-gen “Braga” reportedly delayed into 2026. Microsoft is furthest behind and still overwhelmingly Nvidia-reliant for training. ⚠ Maia 200 specs are a vendor claim, single-sourced — don’t treat as independent benchmark. (Microsoft) ⟨crawl 2026-07-24⟩

2025-10→12 — The Anthropic-on-Amazon anchor

Amazon Trainium3 UltraServers GA (re:Invent, up to 50% lower cost); Project Rainier live (~500k Trainium2 for Claude training, Indiana); AWS passed 1M+ Trainium deployed (+150% YoY); Anthropic committed >$100B to AWS/10yr, up to 5GW Trainium. A frontier lab running mostly off Nvidia is the strongest evidence the revolt is real. (Anthropic) ⟨crawl 2026-07-24⟩

The Projection