Runtime Engineer
Normal Computing
- New York, United States
- Mid Level
- Posted
Skills
Required
- inference
- LLM
- C++ / Rust
- memory management
- concurrency
- building software for hardware
- integrating a runtime
- vLLM
- in-memory compute
Preferred
- open-source runtime / systems infrastructure
Join our team as a Runtime Engineer, where you will be responsible for designing and implementing the runtime layer for our innovative hardware. This is a senior, hands-on role that requires experience in building runtimes, drivers, or hardware abstraction layers for accelerators or other non-standard hardware. You will work closely with cross-functional teams and set the technical direction for the runtime.
About Normal Computing
USD 200,000 - 300,000
Responsibilities
- Conception et mise en œuvre de l'architecture du runtime, y compris la gestion des dispositifs, l'allocation de mémoire et le chargement des exécutables.
- Gestion de l'état des dispositifs sur le matériel, y compris les poids de modèle résidents, pour garantir la fiabilité de la couche de service.
- Définition des interfaces avec le compilateur au-dessus du runtime et avec le pilote et le firmware en dessous, en assurant une intégration fluide.
- Experience building runtimes, drivers, or hardware abstraction layers for accelerators or other non-standard hardware
- Contributions to open-source runtime or systems infrastructure
Requirements
- Understanding of LLM inference workloads and what they demand of a runtime
- Track record of owning a complex systems component end to end, from design through production, and driving technical decisions across teams
- Strong systems programming in C++ or Rust: memory management, concurrency, and low-level performance
- Comfort building software for hardware that is still evolving, from simulation through bring-up
- Experience integrating a runtime with an inference server such as vLLM
- Experience with in-memory compute or other non-von-Neumann hardware
- Experience with open runtime designs such as IREE's HAL, or with CUDA or ROCm runtime internals
jobapplier