Senior HPC Engineer (GPU)
Cosmic Labs - New York
Work at Cosmic Labs
Overview
- View job
Overview
Get AI-powered advice on this job and more exclusive features.
We're looking for an engineer who can solve real-world infrastructure problems — someone who’s tuned GPU workloads at scale, debugged fabric congestion under pressure, and understands the nuances of RDMA and packet-level performance.
If you've built or supported high-performance environments at organizations like AWS, Azure, GCP, Equinix, Digital Realty, or a national lab, we want to talk. No visa sponsorship.
What You’ll Do
- Triage and solve performance issues across GPU clusters, multi-node networks, and scheduler queues
- Tune traffic flows using RDMA, TCP/IP, custom congestion control
- Improve workload orchestration and maximize GPU utilization in volatile, high-throughput environments
- Build tooling and telemetry for packet routing, pacing, jitter analysis, and node-level health
- Work cross-functionally to help our protocol engine make smart, real-time decisions across cloud and bare metal
What We’re Looking For
- Experience running or tuning infrastructure at scale (bonus if from AWS, Azure, GCP, Equinix, Digital Realty, national labs, or HPC clusters)
- Strong understanding of GPU cluster orchestration (e.g., Slurm, Kubernetes, MIG)
- Deep knowledge of networking for AI: RDMA, NCCL, TCP/UDP tuning
- Comfort with low-level debugging: tcpdump, perf, custom scripts, and cluster logs
- Background in HPC, AI infrastructure, trading, or large-scale distributed systems is a MUST
- Excited to help build something fundamentally better than existing networking models
Our Stack Includes
- Linux (bare metal and cloud), NVIDIA and AMD GPUs
- No reliance on SmartNICs, custom silicon, or proprietary hardware
About Us
We’re building a new software layer that makes networks intelligent—capable of real-time path selection, pacing, and failover without hardware changes. Our early deployments focus on AI infrastructure, trading systems, and national security environments. We’re backed by leading investors and actively expanding our engineering team.
If you’re ready to help make compute faster, smarter, and more deterministic, we’d love to hear from you.
Seniority level
Seniority level
Mid-Senior level
Employment type
Employment type
Full-time
Job function
Job function
Engineering and Information TechnologyIndustries
Software Development
Referrals increase your chances of interviewing at Cosmic Labs by 2x
Sign in to set job alerts for “Senior Engineer” roles.
New York, NY $220,000.00-$260,000.00 4 days ago
Brooklyn, NY $150,000.00-$200,000.00 3 months ago
New York, NY $70,000.00-$150,000.00 1 week ago
New York, NY $130,000.00-$180,000.00 6 months ago
New York, NY $99,500.00-$200,000.00 1 day ago
New York, NY $140,000.00-$200,000.00 1 week ago
New York, NY $120,000.00-$180,000.00 5 months ago
Want to work with us, but don't see the right job listed?
Full Stack Software Engineer (All Levels)
New York, NY $140,000.00-$140,000.00 1 month ago
New York, NY $163,200.00-$223,200.00 1 week ago
New York, NY $235,000.00-$255,000.00 1 week ago
New York, NY $110,000.00-$150,000.00 1 month ago
New York, NY $99,500.00-$200,000.00 1 day ago
New York, NY $70,000.00-$150,000.00 2 days ago
New York, NY $140,000.00-$170,000.00 2 months ago
New York, NY $145,000.00-$260,000.00 8 months ago
Backend Software Engineer, CloudKitchens - New York City
Don't see the right opportunity? Apply here!
New York, NY $141,000.00-$202,000.00 2 days ago
New York, NY $120,000.00-$140,000.00 1 week ago
New York, NY $165,000.00-$165,000.00 1 year ago
New York, NY $140,000.00-$200,000.00 1 month ago
New York, NY $120,000.00-$220,000.00 1 month ago
Software Engineer - Frontend / Fullstack
New York, NY $140,000.00-$200,000.00 1 month ago
Backend Engineer, Real-time supply management
New York, NY $128,000.00-$160,000.00 2 weeks ago
New York, NY $100,000.00-$200,000.00 6 months ago
New York, NY $100,500.00-$173,250.00 1 month ago
New York, NY $140,000.00-$230,000.00 3 months ago
New York, NY $140,000.00-$185,000.00 1 week ago
We’re unlocking community knowledge in a new way. Experts add insights directly into each article, started with the help of AI.
#J-18808-Ljbffr