Wafer

Member of Technical Staff

Member of Technical Staff Inference performance depends on how efficiently models use the underlying hardware. Techniques across kernels, compilers, runtimes, and serving systems can dramatically improve latency and…