Remote Part-Time NKI Kernel Engineer for AWS Trainium

Anyone AI Inc.; Argentina; A distancia; ARS 167.954.000 - 314.914.000; A tiempo parcial

Descripción de la vacante

Anyone AI Inc. is seeking experienced engineers to work remotely on evaluating and improving kernel development tasks for AI workloads using AWS Trainium and Neuron Kernel Interface (NKI).

You will review NKI kernels, migrate CUDA to NKI, optimize performance, and assess memory management and DMA scheduling to ensure idiomatic Trainium usage. Strong feedback and cross-backend numerical understanding are essential.

Formación

• ..
• 2+ years of hands-on experience developing or optimizing kernels with the Neuron Kernel Interface (NKI).
• Experience working with AWS Trainium and/or Inferentia2.
• Strong understanding of SBUF / PSUM / HBM memory hierarchy.
• DMA orchestration.
• Ability to evaluate CUDA → NKI migrations.
• Experience profiling and optimizing workloads on Trainium.
• Understanding of numerical differences across GPU and Trainium backends.
• Strong ability to analyze complex technical implementations and provide clear written feedback.

Responsabilidades

• Review and evaluate NKI kernel implementations for correctness and idiomatic Trainium use.
• Migrate CUDA kernels to NKI where appropriate.
• Optimize Trainium kernel performance and benchmarking.
• Manage memory across SBUF, PSUM, and HBM.
• Handle tile-based computation and DMA scheduling.
• Ensure cross‑platform numerical correctness between CUDA/Triton and NKI.
• Identify Trainium-specific bottlenecks and optimization opportunities.
• Provide technical feedback and quality assessment of kernel implementations.

Conocimientos

NKI kernel experience

AWS Trainium / Inferentia2

Memory hierarchy (SBUF/PSUM/HBM)

DMA orchestration

CUDA→NKI migrations

Trainium performance profiling

Cross-backend numerical differences

Technical feedback & documentation

Herramientas

AWS Neuron SDK

Neuron Compiler

Trn1/Trn2 instances

Descripción del empleo

Anyone AI Inc. is seeking experienced engineers to work remotely on evaluating and improving kernel development tasks for AI workloads using AWS Trainium and Neuron Kernel Interface (NKI).

You will review NKI kernels, migrate CUDA to NKI, optimize performance, and assess memory management and DMA scheduling to ensure idiomatic Trainium usage. Strong feedback and cross-backend numerical understanding are essential.