GPU Systems Engineer – HPC / Parallel Computing

Vast.ai; San Francisco (CA); On-site; USD 140,000 - 220,000; Full time

Benefits offered by this job

Health insurance

Dental insurance

Vision insurance

Life insurance

401(k) with company match

Meaningful equity

Onsite meals

Job summary

Vast.ai is seeking a systems engineer to scale AI inference and optimize GPU performance at our San Francisco or Los Angeles offices. You will leverage your HPC background to push the bleeding edge of AI, working with CUDA/C++ and a modern tech stack.

Ideal candidates have advanced C++, experience with parallel frameworks, and a strong track record in high-performance systems. This role is full-time and on-site, offering equity and a fast-paced startup environment.

Qualifications

• Advanced C++ experience with modern standards (C++17/20).
• Strong knowledge of parallel programming frameworks (CUDA, HIP, SYCL, OpenCL, OpenACC).
• Solid background in systems optimization and HPC performance tooling.
• Experience with distributed training/inference is a plus.

Responsibilities

• Design and optimize GPU kernels and tensor libraries.
• Translate HPC techniques into scalable AI inference solutions.
• Evaluate emerging architectures and resource management approaches.
• Collaborate with technical leadership to improve GPU infrastructure efficiency.

Skills

Advanced C++ (C++17/20)

Parallel programming (CUDA, HIP, SYCL,

Systems optimization & HPC tooling

Distributed training/inference (bonus)

Tools

CUDA

HIP

SYCL

OpenCL

OpenACC

GPGPU

Tensor libraries

Job description

About Us

Vast.ai’s cloud powers AI projects and businesses all over the world. We are democratizing and decentralizing AI computing—reshaping our future for the benefit of humanity.

We are a growing and highly motivated team dedicated to an ambitious technical plan. Our structure is flat, our ambitions are out‑sized, and leadership is earned by shipping excellence.

We seek engineers with strong intrinsic drive, a true passion for advancing the state of the art, and a mix of architecture, coding, and communication skills.

LOCATION: On-site at our office in San Francisco or Westwood, Los Angeles.

About the Role

We’re looking for a systems engineer with HPC or parallel programming experience to help scale AI inference. You’ll leverage your knowledge of high-performance systems to optimize GPU performance at the bleeding edge of AI.

• Full-Time
• On-site at either our SF or LA offices

Tech Stack

CUDA/C++, GPGPU, Python, Linux

Key Responsibilities

• Design and optimize GPU kernels and tensor libraries
• Translate HPC techniques into scalable AI inference solutions
• Evaluate emerging architectures and resource management approaches
• Collaborate with technical leadership to improve GPU infrastructure efficiency

Ideal Experience

• Advanced C++ (C++17/20 preferred)
• Expertise with at least one parallel framework (CUDA, HIP, SYCL, OpenCL, OpenACC, or similar)
• Strong background in systems optimization and HPC performance tooling
• Familiarity with distributed training/inference frameworks (bonus)

Interview Process

After submitting your application, our technical team reviews your credentials. If selected, you'll proceed through the following stages:

• Initial screening (virtual, 15 minutes)
• Quick dive into Vast, systems and architectures (virtual, 30 minutes)
• LLM-assisted coding assessment (virtual, 1 hour)
• Meet and greet with coding assessment (on-site, 2 hours)

Our goal is to complete the interview process in two weeks.

Benefits

• Comprehensive health, dental, vision, and life insurance
• 401(k) with company match
• Meaningful early-stage equity
• Onsite meals, snacks, and close collaboration with founders/tech leaders
• Ambitious, fast-paced startup culture where initiative is rewarded