Machine Learning Performance Engineer- World-Leading Prop Trading Fund | London, UK

Oxford Knight; London; On-site; GBP 70,000 - 100,000; Full time

Job summary

A leading proprietary trading fund is looking for a Machine Learning Performance Engineer to enhance the performance of their models. The role involves optimizing training and inference processes, requiring expertise in low-level systems programming and GPU technologies. The ideal candidate will be a curious engineer eager to tackle complex challenges in a tech-centric environment.

Benefits

Market-leading salaries

Generous benefits

Health and mental health support

Holiday entitlement

Parental leave

Retirement plans

Private gym

Tuition reimbursement

Recreation spaces with meals and snacks

Qualifications

• Experience with low-level GPU knowledge and CUDA.
• Understanding of performance optimization in ML.

Responsibilities

• Optimize performance of ML models including training and inference.
• Debug end-to-end training performance issues.

Skills

Machine Learning Techniques

Systems Knowledge

GPU Programming

Debugging

Optimization

Tools

CUDA

CUDA GDB

NSight Systems

NSight Compute

cuDNN

cuBLAS

Job description

Machine Learning Performance Engineer - World-Leading Prop Trading Fund

Summary:
An exciting opportunity to work at a tech-centric proprietary trading fund that trades a wide range of financial products, with offices across the globe. We are seeking an experienced engineer with expertise in low-level systems programming and optimization to join our growing ML team.

Machine learning is central to our operations. Your focus will be to optimize the performance of our models, including training and inference. This involves efficient large-scale training, low-latency real-time inference, and high-throughput research inference. The role requires a holistic approach, including improvements in CUDA, storage systems, networking, and host- and GPU-level considerations.

The ideal candidate is a smart, curious software engineer who enjoys solving complex problems and has a strong desire to learn new skills.

Requirements:

1. Understanding of modern ML techniques and tools, with a focus on performance.
2. Systems knowledge and experience to debug end-to-end training performance issues.
3. Low-level GPU knowledge, including CUDA or similar GPU programming (PTX, SASS, warps, cooperative groups, Tensor Cores, memory hierarchy).
4. Experience with debugging and optimization tools such as CUDA GDB, NSight Systems, NSight Compute.
5. Familiarity with libraries like Triton, CUTLASS, CUB, Thrust, cuDNN, and cuBLAS.

Benefits:

• Market-leading salaries
• Generous benefits including health, mental health, holiday entitlement, parental leave, retirement plans, private gym
• Focus on learning and development, including tuition reimbursement
• Recreation spaces with meals and snacks

We review all applications carefully. Due to high volume, we may not be able to respond to unsuccessful applicants.

Contact:

If you are a good fit, please reach out:

Dan Hampton

Email: [e-mail]

Phone: [phone]

LinkedIn: linkedin.com/in/dan-hampton-ab029392