GPU Kernel Expert
70 $ bis 90 $/Std.Richtwert laut Mercor
Veröffentlicht am 25. August 2026 · Bewerbung bis 15. November 2026
Nur: Vereinigte Staaten
Ich leite Sie zur offiziellen Seite von Mercor weiter. Die Bewerbung ist kostenlos und auf Englisch.
Ich erhalte von der Plattform eine Provision, wenn ein von mir empfohlener Kandidat eingestellt wird. Für Sie ändert sich nichts.
This role evaluates the quality, correctness, and completeness of GPU and accelerator kernel development tasks to train frontier AI models. The candidate must have at least three years of hands-on experience developing, optimizing, or verifying GPU or accelerator kernels in CUDA, Triton, NKI, or Pallas.
Beschreibung auf Englisch, wie von Mercor veröffentlicht.
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)
3 offene Stellen
Nur: Vereinigte Staaten
Über Mercor
Mercor ist ein US-Marktplatz, der Remote-Experten für KI- und Beratungsprojekte einstellt, vom Recht bis zum Ingenieurwesen. Die Originalanzeige steht auf deren Website.
Diese Stelle auf Mercor ansehenÄhnliche Stellen
Trainium (NKI) Kernel Expert
Evaluates the quality, correctness, and optimization of Neuron Kernel Interface development tasks for AWS Trainium hardware. Requires two years of hands-on experience developing or optimizing NKI kernels and strong knowledge of memory-hierarchy management.
Vor 51 Tagen veröffentlicht70 $ bis 90 $/Std.
CVE Vulnerability Expert
The CVE Vulnerability Expert evaluates vulnerability reproduction and remediation tasks for AI model training. The role requires at least three years of hands-on experience in application security, penetration testing, or vulnerability research, along with proficiency in Docker and Docker Compose.
Vor 51 Tagen veröffentlicht70 $ bis 90 $/Std.
SWE-Bench Task Auditor
This remote role involves evaluating and auditing software-engineering benchmark tasks, reference patches, and test harnesses for AI models. Candidates need at least three years of professional software engineering experience, open-source contribution history, and strong code-review skills.
Vor 48 Tagen veröffentlicht70 $ bis 90 $/Std.
Kubernetes Task Auditor
This remote role involves evaluating and grading Kubernetes tasks for AI model training by assessing cluster operations and troubleshooting production scenarios. The ideal candidate brings at least three years of hands-on production Kubernetes experience and deep knowledge of cluster internals and failure modes.
Vor 48 Tagen veröffentlicht70 $ bis 90 $/Std.