ML Compiler Engineer for AI Hardware Accelerators

TetraMem - Accelerate The World; San Jose (CA); On-site; USD 160,000 - 300,000; Full time

Job summary

TetraMem — Accelerate The World is seeking engineers to design and maintain compiler toolchains that translate ML models into optimized workloads for our analog in-memory computing hardware. You will develop runtimes, libraries, and SDKs to deploy AI across our accelerators and implement advanced compiler optimizations to maximize performance and energy efficiency.

Collaborate with ML engineers on model conversion, validation, benchmarking, and deployment, and work with hardware teams to

Responsibilities

• Design, develop, and maintain compiler toolchains for translating ML models to optimized workloads on TetraMem hardware.
• Develop runtime systems, libraries, and SDK components for AI deployment on accelerators.
• Implement compiler optimizations: graph transformations, fusion, memory optimization, scheduling, code generation.
• Research techniques to improve inference speed, latency, throughput, and power across AI workloads.
• Collaborate with ML engineers for model conversion, validation, benchmarking, and deployment.
• Co-design software and hardware features with architects to improve performance and usability.

Job description

TetraMem — Accelerate The World is seeking engineers to design and maintain compiler toolchains that translate ML models into optimized workloads for our analog in-memory computing hardware. You will develop runtimes, libraries, and SDKs to deploy AI across our accelerators and implement advanced compiler optimizations to maximize performance and energy efficiency.

Collaborate with ML engineers on model conversion, validation, benchmarking, and deployment, and work with hardware teams to