GPU Kernel Expert
Evaluate the quality, correctness, and completeness of GPU/accelerator kernel development tasks used to train and evaluate a frontier AI lab's models.
- Pay
- Firm hourly pay: $70-$90 per hour
- Location
- Remote — United States
- Eligibility
- Remote; check the U.S. location wording
- Qualification difficulty
- Selective
How current is this information?
The public role and application path were checked. Details can still change; this is not an endorsement or guarantee.
- Platform
- Mercor
- Fit category
- Software engineering
- Listing/source checked
- Sep 1, 2026
- Inventory presence checked
- Sep 28, 2026
- Apply link checked
- Sep 1, 2026
Application
Continue to the current Mercor listing
Apply on MercorOpens the current Mercor page in a new tab. This may be a referral link, and Specialist AI Work may be paid if the platform credits it. That does not change the role's advertised pay or how roles are ordered here.
What this role involves
Evaluate the quality, correctness, and completeness of GPU/accelerator kernel development tasks used to train and evaluate a frontier AI lab's models. You'll assess numerical correctness, performance-benchmarking fairness, task scoping, and compilation/runtime validity across diverse kernel task types — and provide clear, rubric-based written feedback.
Basic Qualifications • 3+ years of hands-on experience developing, optimizing, or verifying GPU/accelerator kernels in at least two of: CUDA, Triton, NKI, or Pallas (JAX) • Strong understanding of numerical-correctness criteria for kernels (absolute/relative/ULP tolerances, reference-implementation selection) • Demonstrated experience with performance profiling and benchmarking (nsight, ncu, roofline analysis, or framework-native profilers) • Familiarity with common compilation and runtime failure modes (driver mismatches, OOM, launch-configuration errors, shape/stride mismatches, autotuning failures) • Experience with at least three kernel task types: generation from specification, translation/lowering across frameworks, migration between hardware targets, debugging, performance optimization, or operator fusion
Preferred Qualifications
• Experience across both NVIDIA GPU (CUDA/Triton) and custom-accelerator (NKI/Pallas/TPU) ecosystems • Background in compiler engineering, MLIR, or intermediate-representation lowering • Understanding of memory-hierarchy optimization (shared-memory tiling, register pressure, bank conflicts, coalescing patterns) • Contributions to kernel libraries (cuBLAS, cuDNN, Triton community kernels, JAX/XLA custom calls)
Before you apply
Review the main fit signals and unresolved details before opening the platform.
Why it may fit
- Professionals whose experience matches the current GPU Kernel Expert requirements.
- Applicants comfortable completing Mercor's role-specific assessment.
Check before applying
Reasons to pause
- You cannot meet the listing's stated remote or location eligibility.
- You need guaranteed acceptance, hours, or project duration.
Still to verify
- Review the official Mercor listing before applying. Requirements, screening, pay, hours, and project availability can change.
- The reviewed listing had limited public detail; check the current platform page for the full requirements.
What to prepare
- Complete Mercor's role-specific application or assessment.
- Review the current Mercor listing before applying.
Role tools
Where these choices are saved
Save, Not for me, Compare, and application tracking remain in this browser.
Application tips
- Complete Mercor's role-specific application or assessment carefully.
- Review the current Mercor listing and its eligibility details before applying.