Description
Annapurna Labs is hiring a Post-Silicon Validation Engineer to measure and qualify the performance of AWS custom ML training chips against architectural targets. The role involves benchmarking micro-architectures and full model-training workloads, analyzing throughput, memory bandwidth, interconnect latency, and other performance metrics, profiling real transformer, LLM, and vision models, identifying bottlenecks, optimizing with architecture teams, and building automated regression dashboards. The position requires substantial software development, systems design, machine learning, computer architecture, hardware profiling, and statistical analysis experience, and is based in Austin, Texas, with a base salary of 143,700 to 194,400 USD annually.
