AI & Compute
Model infrastructure, inference, accelerators, and IO bottlenecks.
We develop proprietary theses to guide internal investment decisions.
Model infrastructure, inference, accelerators, and IO bottlenecks.
GPU supply chain, packaging, memory, and equipment.
Economic security networks and modular designs.
Streaming, vector search, and privacy‑preserving compute.
Cooling, power delivery, and site optimization.
Developer ergonomics, orchestration, and observability.