Latest AI JobsBrowse Jobs
S

GPU Kernel Engineer

SciforiumSan Francisco, CAEngineering
Apply on company site
Salary
$190K – $250K
Job type
Full-time
Experience
Senior · 5+ yrs
Posted
7 Oct 2026
Status
Open · checked 7d ago

About this role

Sciforium is hiring a GPU Kernel Engineer in San Francisco, California, to design and optimize custom low-level GPU code for artificial intelligence systems. This full-time position suits professionals with a strong background in GPU programming and high-performance computing who want to improve the efficiency of large-scale machine learning platforms.

Role Overview

Sciforium builds multimodal artificial intelligence models and a serving platform, operating in San Francisco, CA. The GPU Kernel Engineer will focus on creating custom GPU kernels and integrating them into machine learning frameworks for training and inference operations.

Key Responsibilities

Design, implement, and optimize custom GPU kernels using C++, PTX, CUDA, ROCm, Triton, and JAX Pallas.
Profile and optimize end-to-end performance of machine learning operations for large-scale language model training and inference.
Integrate low-level GPU kernels into PyTorch, JAX, and internal runtimes.
Develop performance models, resolve bottlenecks, and accelerate artificial intelligence workloads.
Partner with researchers, distributed systems engineers, and model-serving teams.
Coordinate with hardware vendors regarding architecture capabilities and toolchain improvements.
Build tooling, documentation, benchmarking suites, and testing frameworks.

Requirements

Minimum of 5 years of industry or research experience in GPU kernel development or high-performance computing.
Degree in Computer Science, Computer Engineering, Electrical Engineering, Applied Mathematics, or a related field.
Programming proficiency in C++, Python, and familiarity with machine learning frameworks.
Expertise in CUDA or ROCm, GPU memory models, and performance optimization strategies.
Practical experience with Triton or JAX Pallas for custom kernel development.
Knowledge of PTX, GPU assembly, and low-level GPU execution.
Experience writing and optimizing custom GPU kernels in C++ and PTX.
Background working with large-scale language model training or inference workloads.

Nice to Have

Experience with AMD GPUs and ROCm optimization.
Familiarity with JAX FFI and custom machine learning operator development.
Experience with model serving frameworks such as vLLM or TensorRT.
Exposure to TPUs, XLA, or similar accelerator programming environments.
Contributions to open-source machine learning systems, compilers, or GPU kernels.

Skills and Tools

C++
Python
CUDA
ROCm
PTX
Triton
JAX Pallas
PyTorch
JAX

Salary and Benefits

Salary range between $190K and $250K with equity offers.
Medical, dental, and vision insurance.
401(k) plan.
Daily lunch, snacks, and beverages.
Flexible time off.

About the Company

Sciforium is an artificial intelligence infrastructure organization developing multimodal models and a serving platform. The team is backed by funding and direct sponsorship from AMD, with engineering support to scale infrastructure for frontier models and real-time applications.

How to Apply

Interested candidates should visit the official Sciforium careers page to submit an application for the GPU Kernel Engineer position.

Skills and tech stack

C++PythonCUDAROCmPTXTritonJAX PallasPyTorchJAXvLLMTensorRTXLALLMDistributed Systems

Questions candidates ask

What is the salary for the GPU Kernel Engineer role?

The salary ranges from $190K to $250K, and the position includes equity offers.

Where is the job located?

The position is located in San Francisco, CA.

How many years of experience are required?

Candidates must have 5+ years of industry or research experience in GPU kernel development or high-performance computing.

What benefits are provided?

Benefits include medical, dental, and vision insurance, a 401(k) plan, daily lunch, snacks, and beverages, flexible time off, and equity.

What specific technical proficiencies are needed?

Applicants need strong skills in C++, Python, CUDA or ROCm, Triton or JAX Pallas, PTX, and familiarity with machine learning frameworks like PyTorch and JAX.

Ready to apply?

The application is on the official Sciforium careers page.

Apply on company site