GPU Systems Engineer

Iceberg; New York (NY); On-site; USD 200,000 - 300,000; Full time

Job summary

Iceberg is seeking a GPU Systems Engineer in New York to design and scale GPU infrastructure across thousands of nodes and petabytes of storage. You will profile workloads, identify bottlenecks, and build automation to keep a massive fleet running with minimal human intervention.

You'll work across Linux, GPU infra, GPUDirect RDMA, CUDA, C/C++, Python, networking and distributed systems, with exposure to hardware selection and architecture from design to production.

Qualifications

• 5+ years of experience with large-scale Linux systems in HPC/AI or distributed infrastructure environments.
• Hands-on experience with GPU infrastructure design and performance profiling.
• Strong fundamentals in CUDA, C/C++, and Python.

Responsibilities

• Design and scale GPU clusters and storage infrastructure.
• Profile workloads and optimize performance across the stack.
• Develop automation to minimize manual intervention and outages.
• Collaborate across hardware, software, and networking teams.

Skills

Linux systems

GPU infrastructure

CUDA

C/C++

Python

Profiling & optimization

Distributed systems

Networking

Performance tuning

Automation

Tools

GPUDirect RDMA

NVIDIA CUDA

NCCL

NVLink

Profiling tools

Job description

New York | $200,000-$300,000 Base + Bonus

I'm working with one of the world's leading trading firms on an interesting GPU Systems Engineer position.

This is a hands-on infrastructure role working at scale, with large CPU and GPU clusters spanning thousands of nodes and hundreds of petabytes of storage.

You'll be designing and scaling GPU infrastructure, profiling workloads, finding performance bottlenecks and building the automation needed to keep a huge fleet running with minimal human intervention.

You'll be working across Linux, GPU infrastructure, GPUDirect RDMA, CUDA, C/C++, Python, networking and distributed systems, with the opportunity to get involved from hardware selection and architecture all the way through to production.

I'm looking for someone with 5+ years working with large-scale Linux systems in HPC, AI or distributed infrastructure environments who genuinely enjoys getting into the weeds when something isn't performing as it should.

Experience with NVIDIA, NCCL, NVLink or large-scale AI infrastructure would be a big plus.

$200k-$300k base + bonus.

If you're a systems engineer who likes working at scale and solving problems across hardware, software and networking, I'd love to speak.