← All roles

HPC Performance Engineer

CoreWeave ·US ·Remote 75d ago
HPCKubernetesPythonGo

About the role

CoreWeave seeks an HPC Performance Engineer to join the HAVOCK Team, responsible for designing and optimizing bare-metal systems from POST to Kubernetes. The role involves maintaining custom Linux kernels, OS images, virtualization (kubevirt/qemu/vfio), and container runtime (containerd/nydus/kubelet). Collaborate with cross-functional teams to ensure low-level software stack performance, develop performance baselines, regression testing, and telemetry for distributed clusters.

Requirements

5+ years in Systems/HPC Performance Engineering, Bachelor's in Computer Engineering or related. Strong experience with MPI workloads, distributed system performance, RoCE, InfiniBand, GPUDirect, NUMA. Hands-on with HPC benchmarks (HPCC, HPL, OSU, MLPerf-HPC, STREAM, IO500). Deep Linux internals, programming in Python, robust code, automation testing, communication, documentation. Preferred: Golang, cloud, open-source, ML.

About the company

CoreWeave

CoreWeave is a cloud provider specializing in an AI-native platform built to power complex AI workloads. They offer GPU and CPU compute, storage, and infrastructure control solutions, positioning themselves as the essential cloud for AI innovation and significantly reducing total cost of ownership.

View company page →
$165k–242k
Remote New York, NY / Sunnyvale, CA / Bellevue, WANew York, USSunnyvale, USBellevue, US
Posted 75d ago
Apply for this role
Opens coreweave.com ↗
About the company
CoreWeave
https://coreweave.com

CoreWeave is a cloud provider specializing in an AI-native platform built to power complex AI workloads. They offer GPU and CPU compute, storage, and infrastructure control solutions, positioning themselves as the essential cloud for AI innovation and significantly reducing total cost of ownership.

View company page →