Skip to main content
Normal Computing

Compiler Frontend Engineer

Normal Computing Mountain View, CA, United States Full-time 28 days ago
Embedded & Systems Software AI/ML Hardware

Extend IR, ops, and lowering passes for in-memory compute hardware targeting modern LLM architectures.

Originally posted October 22, 2026 by the employer.

About the role

As a Compiler Frontend Engineer, you will extend IR, ops, and lowering passes to support new LLM architectures for pre-silicon hardware that executes memory-bound inference inside memory. You will validate bring-up against simulators and FPGA prototypes to prepare models for silicon arrival.

What you'll do

  • Bring up frontend and mid-level support for new LLM architectures in the compiler, including attention variants and MoE routing schemes.
  • Extend IR, ops, and lowering passes to represent architectural primitives faithfully with comprehensive test coverage.
  • Implement quantization schemes and numeric formats used by new models to ensure lowering preserves model accuracy.
  • Track the LLM architecture landscape to recommend priorities for upcoming model bring-up.
  • Hand lowered programs to backend compiler engineers mapping onto hardware, and collaborate with LLM serving engineers to take models to production serving.

What you'll need

  • Hands-on ML compiler experience writing and shipping compiler passes.
  • Deep familiarity with PyTorch internals, ideally including TorchDynamo, FX, or torch.export.
  • Working knowledge of LLM architectures, covering attention mechanisms, MoE routing, and quantization schemes.
  • Fluency in Python, along with familiarity with C++ or Rust.
  • Experience building software for evolving hardware environments, spanning simulation through hardware bring-up.

Nice to have

  • Experience with other ML compilers such as IREE, XLA, or TVM.
  • Familiarity with ML frameworks and graph formats such as JAX or ONNX.
  • Experience partitioning models across devices using tensor, pipeline, or expert parallelism.
  • Experience with hardware-software co-design or non-standard/in-memory hardware targets.
  • Contributions to open-source compiler or ML infrastructure projects.

Skills: ML compiler, PyTorch, IR lowering, LLM architectures, quantization

More semiconductor jobs: Mountain View jobs

Market context

Tracked in our index since 23 Oct 2026. This is our observation date, not the employer posting date. Specialty: Infrastructure/Platform Software (AI ML Hardware). 109 active roles in this specialty at 29 employers in our index, as of 18 Nov 2026.

How we counted: postings read from employer career sites. How we count

Normal Computing

Normal Computing

Apply now
Mountain View, CA, United States
On-site
Full-time
28 days ago

Share this job