Software Engineering Manager, AI Kernels & Performance Optimization — MTIA Software

Meta designs and deploys its own AI systems. MTIA — the Meta Training and Inference Accelerator — is Meta's family of in-house AI accelerator ASICs, running recommendation and ranking workloads in production across Meta's data centers today and expanding into generative AI inference and training as successive silicon generations land. The MTIA Software team is part of the AI & Compute Foundation (ACF) organization within Meta Infrastructure.

We build the entire stack a chip vendor would normally supply — compiler and LLVM toolchain, runtime, kernel authoring frameworks and libraries, developer tooling, and deep PyTorch integration — co-designing it with the silicon teams generation over generation. Within that stack, the AI Kernels and Performance Optimization team (KORE) owns the layer where silicon architecture meets arithmetic. Our mission is roofline-level performance and programmability at scale: we design and implement high-performance compute and communication kernels (GEMM, attention variants, normalization, collectives, elementwise and reduction fusions, sparse and quantized paths) against novel architectural features including matrix engines, on-chip reduction fabrics, and software-managed memory hierarchies.