Large Power Transformer Lead Times Hit 2029, Prices Up 77% Since 2019, Choking AI Data Center Buildout
sahilypatel · x · 2026-09-20
The deepest bottleneck in the AI data center boom may be the transformer: delivery times for large power transformers now stretch to 2029, with standard units averaging 128 weeks and generator step-up units 144 weeks. Prices have risen about 77% since 2019.
The chain is simple: every AI data center needs power, every power plant needs a transformer, and every transformer needs a core made from grain-oriented electrical steel — rolled so all crystal grains align, otherwise magnetic domains don't line up optimally and the decades-long-lived machine runs less efficiently. Only one U.S. company currently makes this steel.
The supply structure is fragile: about 80% of U.S. large power transformers are imported, shipped by rail on specialized cars, and each is custom-designed for a single substation with no spares kept. As the author puts it, one of the largest capital buildouts in history is being bottlenecked by a 19th-century machine made of steel and copper.
More from Infra
- Athena engine runs DeepSeek V4 + Qwen3.8 at 262K context on one DGX Spark — solyarisoftware · 2026-09-20
- Dev benchmarks Bend 2 on M4 Max: parallel kernel 6.4x faster than NumPy — arthurcolle · 2026-09-20
- China's CXMT says new memory-chip platform enters mass production — johnnyApplePRNG · 2026-09-20
- Unverified claim: OpenAI is out of compute — ns123abc · 2026-09-20
- awesome-local-ai: one-command local AI stacks with coding agents, benchmarked on real hardware — julianharris · 2026-09-20
- Google open-sources AX, a Kubernetes-style orchestrator built for agentic workloads — rakyll · 2026-09-20