ML Performance Engineer: GPU & Systems Optimisation

Trading Interview; Greater London; Hybrid; GBP 120,000 - 180,000; Full time

Job summary

Jane Street in London is seeking an engineer with experience in low-level systems programming and optimisation to join our ML team. Your focus will be on optimizing the performance of our models for training and inference.

We reward curious minds and practical problem solving in a fast-moving environment. We value cross-disciplinary thinking across storage, networking and host- and GPU-level considerations to squeeze throughput and reduce latency in a modern trading platform.

Qualifications

• Understanding of modern ML techniques and toolsets.

Responsibilities

• Optimise model training performance end-to-end.
• Improve real-time inference latency and throughput.
• Architect platform optimisations across storage, networking and GPU subsystems.
• Assess throughput vs goodput at the vector load level.

Skills

ML techniques

Performance debugging

Low-level GPU knowledge

Distributed GPU training

Parallel programming

Networking & storage

Tools

CUDA GDB

NSight Systems

NSight Compute

Triton

CUTLASS

CUDNN

cuBLAS

NVLink

Infiniband

Job description

Jane Street in London is seeking an engineer with experience in low-level systems programming and optimisation to join our ML team. Your focus will be on optimizing the performance of our models for training and inference.

We reward curious minds and practical problem solving in a fast-moving environment. We value cross-disciplinary thinking across storage, networking and host- and GPU-level considerations to squeeze throughput and reduce latency in a modern trading platform.