Senior Runtime Engineer, AI Hardware

Normal Computing Corporation; Mountain View (CA); On-site; USD 180,000 - 320,000; Full time

Job summary

Normal Computing is seeking a senior Runtime Engineer to own the design and implementation of the runtime that powers our pre-silicon hardware platform. You will work on device management, memory allocation, executable loading, and command submission, building a stable interface for future silicon.

This role is hands-on and cross-functional, requiring deep systems programming in C++ or Rust, experience with accelerators, and a track record of delivering complex components end-to-end.

Qualifications

• Experience building runtimes, drivers, or hardware abstraction layers for accelerators or non-standard hardware.
• Strong systems programming skills in C++ or Rust with memory management and concurrency.
• Track record owning a complex systems component end to end across teams.
• Understanding of LLM inference workloads and runtime needs.
• Comfort building software for evolving hardware from simulation to bring-up.

Responsibilities

• Own runtime architecture design including device management and memory allocation.
• Hide hardware differences behind a stable interface.
• Manage device state like resident model weights.
• Keep runtime overhead off the inference path for latency and throughput.
• Define boundaries with compiler above and driver below; collaborate with hardware team.
• Provide technical leadership and review designs with cross-team engineers.

Skills

Runtime architecture

Systems programming

C++

Rust

LLM inference awareness

Tools

C++

Rust

Job description

Normal Computing is seeking a senior Runtime Engineer to own the design and implementation of the runtime that powers our pre-silicon hardware platform. You will work on device management, memory allocation, executable loading, and command submission, building a stable interface for future silicon.

This role is hands-on and cross-functional, requiring deep systems programming in C++ or Rust, experience with accelerators, and a track record of delivering complex components end-to-end.