Normal Computing
Compiler Frontend Engineer
Mountain View +1 · On-site · Full-time · Posted Oct 22, 2026
- Location
- Mountain View
New York
About the job
Extend Normal Computing’s compiler frontend and mid-level infrastructure to bring new LLM architectures to its pre-silicon hardware, validating model support with simulators and FPGA prototypes.
What you'll do
- Add compiler IR, operations, and lowering passes for new LLM architectures, quantization schemes, and numeric formats, with tests to ensure correctness.
- Track emerging LLM architectures and recommend which models the team should support next.
- Coordinate with backend compiler and LLM serving engineers to move models from compiler support to serving.
What they ask for
- Hands-on ML compiler experience, including writing and shipping compiler passes.
- Deep familiarity with PyTorch internals; compilation-stack experience with TorchDynamo, FX, or torch.export is preferred within this requirement.
- Working knowledge of modern LLM architectures, including attention, MoE routing, and quantization.
- Fluency in Python and familiarity with C++ or Rust.
Nice to have
- Experience with other ML compilers such as IREE, XLA, or TVM; or frameworks and graph formats such as JAX or ONNX.
- Experience with model parallelism, hardware-software co-design, non-standard or in-memory hardware, or open-source compiler/ML infrastructure contributions.
Tools
PyTorch, TorchDynamo, FX, torch.export, Python, C++, Rust, IREE
AI focus
Compiler and model bring-up for LLM architectures, including attention variants, MoE routing, and quantization; the role also tracks LLM research and calls for publishing on behalf of Normal.
Model-written summary · the original posting is authoritative