Responsibilities: - Design, develop, and maintain compiler toolchains that translate machine learning models from industry-standard frameworks into optimized workloads for TetraMem's analog in-memory computing hardware.
- Develop runtime systems, software libraries, and SDK components that enable efficient deployment, execution, and management of AI applications on TetraMem accelerators.
- Implement compiler optimizations, including graph transformations, operator fusion, memory optimization, scheduling, and code generation to maximize performance and energy efficiency.
- Research and develop innovative techniques to improve machine learning inference speed, latency, throughput, and power consumption across a wide range of AI workloads.
- Collaborate closely with machine learning engineers to support model conversion, validation, optimization, benchmarking, and deployment.
- Partner with hardware architects and silicon engineering teams to co-design software and hardware features that improve system performance, programmability, and usability.
- Develop performance analysis, profiling, debugging, and benchmarking tools to evaluate and optimize AI workloads on current and future TetraMem platforms.
- Integrate and support industry-standard machine learning frameworks and model formats, including PyTorch, TensorFlow, ONNX, and other emerging AI ecosystems.
- Lead technical design reviews, contribute to software architecture decisions, and establish best practices for scalable, maintainable, and high-quality software development.
- Mentor junior engineers, contribute to technical documentation, and help define the long-term roadmap for TetraMem's compiler, runtime, and SDK technologies.
Requirements: - MS or PhD in Computer Engineering/CS/EE
- 5+ years industry experience as a compiler engineer or developer
- Experience developing compilers for GPU, dataflow compilers, or ML compilers
- Startup mindset/experience
Experience in one or more of the following areas considered a strong plus: - Experience in RISC-V CPU/VPU kernel development and optimization
- Experience providing technical leadership and/or guidance to other engineers
- Knowledge of popular CPU/GPU compilers such as GCC, Clang
- Knowledge of ML compilers such as MLIR
- Experience with LLVM and other open-source compiler libraries and tools
- Publications on compilation of ML or dataflow programs for HW acceleration
Salary Range: $160,000 - $300,000 / year
