Caching Gpu Jobs in United States

The Caching Infrastructure team is responsible for building a caching layer that powers many critical use cases at OpenAI.

Own verification and validation of GPU software, including unit tests, system-level tests, and hardware-in-the-loop tests for GPU software.

can deliver frontier inference performance on AMD GPUs.

Deep architectural knowledge of the modern GPU/TPU ecosystem.

Own GPU operational readiness.

, KV-cache efficiency, and tensor/expert parallelism.

Building performance-modeling and observability frameworks for GPU utilization and cost.

flow for GPUs, detects bugs early, and generates automated alerts to maintain the GPU stack's functionality and performance Create profiler and debugger

, and distributed GPU workloads.

in GPU programming and parallel computing.

in GPU programming and parallel computing.

Solve for GPU scarcity.

the default for Anthropic engineers Design and operate CDC-driven cache invalidation that keeps cached data consistent with source-of-truth databases

for the physical and server infrastructure that powers neural recording training, data processing, and engineering workloads: the data center floor itself, the GPU