GPU Programming Software Engineer | $60-$95/hr Remote
Overview
This remote contract role puts your GPU expertise to work on large language model training. You will build and refine GPU kernels and shaders with CUDA, WebGPU, or GLSL, then profile them until they hit the performance targets. C++ handles the host-side logic that ties everything together. The project is output-based, so you earn per completed task that passes specs, with rates listed at $60 to $95 per hour. Expect a quick start: roles often fill in two days and first tasks begin within 24 to 48 hours after onboarding.
What You'll Do5
- 1Build GPU kernels and shaders in CUDA, WebGPU, or GLSL to support LLM training tasks.
- 2Profile existing GPU code and adjust memory access, occupancy, or instruction flow to close performance gaps.
- 3Develop C++ host-side integration code that orchestrates kernel launches and data movement.
- 4Translate model architecture requirements into discrete GPU task definitions and execution plans.
- 5Test and refine GPU solutions against performance targets and project acceptance criteria.
Requirements4
- 1Advanced hands-on skill with a GPU language that runs on NVIDIA chips, such as CUDA.
- 2Track record in a GPU-focused domain: graphics programming, ML acceleration, scientific computing, high performance computing, or similar.
- 3Strong C++ ability for writing host-side logic and system integration.
- 4Comfort with profiling and tuning GPU kernels or shaders for speed and efficiency.
Who Should Apply
An ideal match has shipped GPU kernels or shaders on NVIDIA hardware and can point to measurable performance wins in graphics, HPC, or ML acceleration. You should be comfortable owning C++ host code and profiling your own work end to end. Contractors who enjoy output-based pay and can start first tasks within a day or two tend to thrive here. Engineers who prefer fixed hourly billing, a defined project scope, or a long ramp-up period will find this setup frustrating. Common rejections go to people who mention CUDA on paper but cannot show NVIDIA-specific optimization examples, or who hesitate to commit to the weekly minimum task count.
Salary Insight
The posting lists $60 to $95 per hour. The description adds that pay is output-based and you earn per task that meets the project specifications, so your real income depends on task volume and pace. Per-task rates are not broken out; expect to nail those down in the hiring conversation.
Location
Required Skills
Application Tip
Include two or three code samples that show kernel or shader optimizations on NVIDIA GPUs, with before-and-after profiling results. Describe the C++ host-side logic you wrote around each kernel.
See NearSkill jobs more often in your search
How your application is processed
1Application received
Your resume and details are logged the moment you apply.
2ATS + eligibility screening
We check your profile against the role’s skills, seniority, and requirements.
3Employer sees qualified profiles only
Only candidates who clear screening move forward.
Similar open positions
Explore active roles that match your skills and interests.
Micro1
VerifiedCUDA Engineering Expert
A remote contract role focused on optimizing GPU kernels with CUDA for a customer project in collaboration with a top AI lab. You’ll profile, tune, and refactor CUDA and C++ code to boost throughput on modern GPUs. Experience with GLSL and WebGPU helps you implement shader logic and graphics workflows within existing pipelines. Clear, actionable documentation and technical communication are essential as you contribute to design discussions and stay current on GPU programming advances.
Mercor
VerifiedGPU Kernel Expert
This remote contract role puts you inside the evaluation loop for GPU/accelerator kernel tasks generated for a frontier AI lab. You will review assignments built on CUDA, Triton, NKI, and Pallas, checking whether they are numerically sound, correctly scoped, and safe to run. Your written, rubric-based feedback helps decide which tasks are used to train and evaluate the lab's models.
Mercor
VerifiedCUDA Engineering Expert
Mercor is putting together a team of GPU kernel specialists for a project backed by a top-tier AI lab. You'll dig into kernel code, use profiler data to spot performance bottlenecks, and make targeted performance improvements across modern GPU hardware. The role is a short-term contract, and you don't need to be an expert in every underlying algorithm to contribute. The work centers on CUDA and C++17 skills, plus a practical eye for profiler metrics like occupancy and cache throughput.
Mercor
VerifiedSoftware Engineer, Full Stack (Python, Java, Rust, C#, C++)
This role puts you inside a leading AI lab's generative AI team, where you'll build production-grade full-stack applications that push frontier large language models to their limits. You'll integrate unreleased model APIs into working software, uncover edge-case failures, and iterate quickly in a fast-paced research environment. The position is a full-time (40 hours/week) remote W-2 engagement through Cincinnatus LLC, with pay at $50–65 per hour.
Mercor
VerifiedSenior Software Engineer, Full Stack (Python, Java, Rust, C#, C++)
This role places senior full-stack engineers inside a top-tier AI lab's generative AI team, where you'll build real applications and internal tools on top of frontier large language models before they're publicly released. You'll work directly with the lab's engineering manager, moving quickly and shipping working code from day one. It's a fully remote, 40-hour-per-week W-2 contract through Cincinnatus LLC, with a competitive hourly rate.
Micro1
VerifiedSenior Software Engineer
Building reinforcement learning environments sits at the center of this remote contractor role. You'll create reproducible workflows that mirror real DevOps and CI/CD debugging scenarios, using command-line tools like git, docker, gdb, and ffmpeg. Each environment you produce comes with a golden reference solution that marks what a correct result looks like. Your command of C++ or Python (or Java, GoLang, TypeScript, Rust) matters more than hands-on AI experience, though the output here shapes how AI models learn and reason. Expect to work about 15 hours a week with output-based pay.