Senior Software Engineer, CUTLASS Kernels

at Nvidia

📍 Santa Clara, United States

USD 152,000-287,500 per year

SENIOR

✅ On-site

Tech Stack
Tag name is followed by "@" symbol and proficiency level value. About proficiency levels:

1-2 — basic awareness. Minimal hands-on experience, and a rudimentary understanding of the technology's purpose;

3-6 — daily use. Comfortable and regular usage, capable of handling common tasks and challenges related to the technology;

7-9 — you are an expert, you can teach others, you know all the pitfalls and tricks;

10 — exceptional knowledge, comprehensive understanding, and adeptness in all aspects of the technology, including advanced problem-solving. Think twice before claiming or demanding such level.

AI CUDA @ 4 Debugging @ 7 Deep Learning GPU HPC OpenCL @ 4 Python

Details

NVIDIA's high-performance computing platforms are powering the AI revolution across many applications and industries. Within NVIDIA's software stack, CUTLASS is an open-source ecosystem dedicated to high-performance linear algebra and Tensor Core primitives. Since 2017, it has provided C++ and Python abstractions for implementing custom matrix multiply (GEMM) and related mathematical and deep learning computations on NVIDIA GPUs.

The CUTLASS team develops and optimizes math kernels to extract the highest performance from NVIDIA hardware architectures.

Responsibilities

Write Tensor Core-based deep learning kernels, including grouped-GEMM, attention, and convolution, using CUTLASS CUDA C++ and Python DSL for Blackwell, Rubin, and future architectures.
Optimize kernels for peak throughput on both silicon and software performance simulators.
Collaborate with NVIDIA teams across GPU architecture, NVVM/PTX compiler, CUDA libraries, and deep learning frameworks to ensure fast, functional, and timely kernel delivery to customers.

Requirements

Master's or PhD degree in Computer Science, Computer Engineering, or a related field, or equivalent experience.
3+ years of relevant industry experience.
Strong proficiency in C++ programming and software design, including debugging, performance evaluation, and testing.
Experience with CUDA, OpenCL, HIP, SYCL, Mojo, Pallas, Triton, Mosaic, Halide, or another general-purpose or domain-specific programming language targeting highly parallel accelerators.
Deep understanding of computer architecture and some experience working at the assembly level.

Preferred Qualifications

Experience writing code specifically targeting NVIDIA Tensor Cores, particularly through PTX or CUDA/cuTile.
Open-source contributions to math kernel libraries or frameworks.

Benefits

Equity and benefits are provided.
NVIDIA is committed to fostering an inclusive work environment and is an equal opportunity employer.

Applications for this job will be accepted at least until June 5, 2026. This posting is for an existing vacancy. NVIDIA uses AI tools in its recruiting processes.