Compiler Frontend Engineer

Normal Computing Corporation • Palo Alto, CA

16h

About The Position

Normal Computing is an applied AI company building foundational hardware and software for the semiconductor industry, critical AI infrastructure, and systems that power the world. The company works across New York City, Silicon Valley (Mountain View), London, Copenhagen, and Seoul.

Normal's pre-silicon hardware runs memory-bound large-model inference inside memory. Its compiler makes that hardware programmable and must support new LLM architectures, attention variants, MoE routing schemes, and quantization approaches. The Compiler Frontend Engineer will focus on frontend and model bring-up, extending the compiler's IR, operations, and lowering passes so new architectures are represented faithfully and supported before customer demand. This research-adjacent role involves closely following relevant literature and publishing on behalf of Normal. Model bring-up will be validated using simulators and FPGA prototypes before silicon is available.

Requirements

• Hands-on ML compiler experience, including writing and shipping compiler passes.
• Deep familiarity with PyTorch internals, ideally including its compilation stack: TorchDynamo, FX, and torch.export.
• Working knowledge of modern LLM architectures, including attention mechanisms, MoE routing, and quantization schemes.
• Patience to take new architectures from initial execution through correctness and performance.
• Comfort designing compiler infrastructure in a young codebase and extending an established one.
• Comfort building software for evolving hardware, from simulation through bring-up.
• Fluency in Python.
• Familiarity with C++ or Rust.

Nice To Haves

• Experience with other ML compilers, such as IREE, XLA, or TVM.
• Familiarity with other 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.
• Experience with non-standard or in-memory hardware targets.
• Contributions to open-source compiler or ML infrastructure projects.

Responsibilities

• Bring up frontend and mid-level compiler support for new LLM architectures, including attention variants and MoE routing schemes.
• Extend IR, operations, and lowering passes to faithfully represent new architectural primitives, with test coverage to maintain correctness as the compiler evolves.
• Represent new models' quantization schemes and numeric formats so lowering preserves model accuracy.
• Track developments in LLM architectures and recommend which architectures the team should bring up next.
• Hand lowered programs to backend compiler engineers for hardware mapping and collaborate with the LLM serving engineer to move new models from compiler support to serving.
• Follow relevant research literature and publish on behalf of Normal.
• Validate model bring-up using simulators and FPGA prototypes ahead of silicon availability.

What This Job Offers

• Job Type: Full-time
• Education Level: No Education Listed