Назад
обновлено 5 дней назад

Software Engineer (GPU Networking)

165 000 - 330 000$
Формат работы
hybrid
Тип работы
fulltime
Грейд
senior
Английский
b2
Страна
US/Canada
Вакансия из списка Hirify.GlobalВакансия из Hirify Global, списка международных tech-компаний
Для мэтча и отклика нужен Plus

Мэтч & Сопровод

Для мэтча с этой вакансией нужен Plus

Описание вакансии

Текст:
/
TL;DR
Software Engineer (GPU Networking): Building and optimizing high-performance GPU networking and distributed systems for AI inference with an accent on integrating RDMA capabilities and co-optimizing communication alongside computation. Focus on architecting the software fabric that unifies thousands of GPUs, enabling serverless-grade startup speeds for LLMs, and deep-diving into bleeding-edge hardware performance.

Location: Hybrid in San Francisco, Montreal, New York, Seattle, or Toronto

Salary: $165K–$330K annually, plus equity

Company

An AI infrastructure company building inference platforms, developer tooling, and distributed hardware systems for mission-critical AI applications.

What you will do

  • Integrate RDMA, RoCE, and InfiniBand capabilities into the inference stack.
  • Build and tune networking layers for disaggregated KV cache offload and Wide Expert Parallelism.
  • Develop checkpointing and storage mechanisms that enable sub-10-second startup for trillion-parameter models.
  • Characterize networking performance on H100, H200, B200, B300, GB200, and GB300 NVL72 clusters and create hardware acceptance tests.
  • Design observability tools for packet flow, congestion, and effective bandwidth across GPU interconnects.
  • Optimize NCCL and NVSHMEM communication and develop custom kernels to overlap computation with data transfer.

Requirements

  • Deep experience with high-performance networking protocols, including InfiniBand and RoCE v2.
  • Fluency in C++ or Python and the ability to work across high-level software and hardware.
  • Strong understanding of memory hierarchies in modern NVIDIA architectures, including H100 and Blackwell GPUs.
  • Experience debugging low-level systems such as TensorRT-LLM, custom C++ or Python bindings, or NVLink topologies.
  • Ability to evaluate when to use existing solutions and when to build custom networking infrastructure.

Nice to have

  • Knowledge of NCCL, NVSHMEM, and UCX.
  • Rust experience in systems-level or performance-critical networking.
  • Experience with GPUDirect Storage or high-performance filesystems such as Weka or 3FS.
  • Familiarity with TensorRT-LLM, vLLM, or SGLang.
  • Experience benchmarking and qualifying new hardware clusters.

Culture & Benefits

  • Work with bleeding-edge Blackwell and upcoming Rubin GPU architectures.
  • Operate across hardware interconnects, communication kernels, and distributed inference strategies.
  • Competitive compensation with meaningful equity.
  • U.S. employees receive full medical, dental, and vision coverage, plus access to a company-facilitated 401(k).
  • Flexible PTO, company-wide winter break, paid parental leave, and a fertility and family-building stipend.

Будьте осторожны: если работодатель просит войти в их систему, используя iCloud/Google, прислать код/пароль, запустить код/ПО, не делайте этого - это мошенники. Обязательно жмите "Пожаловаться" или пишите в поддержку. Подробнее в гайде →