Tower Research Capital
GPU Systems Engineer
$200K–$300K
New York +1 · Full-time · Posted Aug 10, 2026
- Location
- New York
New York, NY, United States - Experience
- 5+ years.
About the job
Designs, operates, and improves large-scale distributed GPU infrastructure that supports the firm’s trading and research workloads.
What you'll do
- Design, deploy, and scale GPU clusters, including hardware selection, network topology, and production operations.
- Profile GPU workloads and diagnose performance issues across compute, storage, networking, and hardware/software layers.
- Build automation for provisioning, monitoring, diagnostics, and self-healing across large infrastructure fleets.
What they ask for
- 5+ years engineering large-scale Linux systems in HPC, AI, or distributed-infrastructure environments.
- Deep Linux fundamentals, including installation, performance tuning, and debugging.
- Hands-on troubleshooting of distributed GPU workloads and working experience with GPUDirect RDMA.
- Solid Python plus CUDA or C/C++ experience; familiarity with Salt, Ansible, Puppet, or Chef.
Nice to have
- Experience with the rest of the NVIDIA stack, such as NCCL and NVLink.
Pay
$200,000-$300,000 USD base salary
Tools
Linux, GPUDirect RDMA, Python, CUDA, C/C++, Salt, Ansible, NVIDIA
AI focus
Supports AI infrastructure by designing and operating GPU clusters, profiling and benchmarking GPU workloads, and improving their performance.
Model-written summary · the original posting is authoritative