Manager, Software Engineering, MTIA Software

The MTIA (Meta Training & Inference Accelerator) Software team is part of AI and Compute Foundation organization. The team’s mission is to explore, develop and help productize high-performance software and hardware technologies for AI at datacenter scale. The team co-optimizes both SW (e.g., algorithms and models) and HW (e.g., platform and network) to come up with balanced system design. To develop new systems, requires understanding performance bottlenecks on existing systems. As a result, the team invests significantly into optimizing AI production models on existing systems. This has resulted in TCO wins for all key AI services. Team has been developing AI compilers to accelerate Meta’s DL/ML workloads on the specialized MTIA AI accelerator hardware in a highly performant and flexible way. As part of the AI acceleration software stack, we develop compiler optimizations exploiting various hardware architectural features, achieving high performance for our inference and training workloads.