Назад
Company hidden
2 месяца назад

AWS Trainium / Neuron Kernel Interface (NKI) Engineer

65$
Формат работы
remote (только Europe)
Тип работы
project
Грейд
middle
Английский
b2
Страна
Argentina/Chile/Spain +6 еще
Вакансия из списка Hirify.GlobalВакансия из Hirify Global, списка международных tech-компаний
Для мэтча и отклика нужен Plus

Мэтч & Сопровод

Для мэтча с этой вакансией нужен Plus

Описание вакансии

Текст:
/
TL;DR
AWS Trainium / Neuron Kernel Interface (NKI) Engineer (AI kernel engineering): Reviewing and improving NKI kernel implementations for AWS Trainium and Inferentia2 workloads with an accent on correctness, CUDA-to-NKI migration, memory hierarchy, and performance benchmarking. Focus on profiling Trainium workloads, optimizing tile-based computation and DMA scheduling, and assessing numerical equivalence across GPU and Trainium backends.

Location: Remote contract role available in Argentina, Brazil, Chile, Colombia, Ecuador, Mexico, Portugal, Spain, and Uruguay.

Compensation: $65 per hour.

Company

hirify.global is running a specialized project focused on evaluating and improving kernel development tasks for AI workloads.

What you will do

  • Review and evaluate NKI kernel correctness, Trainium-specific development patterns, and kernel implementation quality.
  • Assess CUDA-to-NKI kernel migrations and cross-platform numerical correctness between CUDA, Triton, and NKI.
  • Analyze and optimize Trainium performance, including benchmarking, bottleneck identification, and optimization opportunities.
  • Evaluate memory management across SBUF, PSUM, and HBM, along with tile-based computation and DMA scheduling.
  • Provide clear written technical feedback on whether implementations are idiomatic and optimized for Trainium hardware.

Requirements

  • At least 2 years of hands-on experience developing or optimizing kernels with the Neuron Kernel Interface (NKI).
  • Hands-on experience with AWS Trainium and/or Inferentia2.
  • Strong understanding of tile-based computation, SBUF / PSUM / HBM memory hierarchy, partition dimension constraints, DMA orchestration, and Trainium-specific optimization techniques.
  • Ability to evaluate CUDA-to-NKI migrations and 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.

Nice to have

  • Experience with the AWS Neuron SDK or Neuron Compiler.
  • CUDA or Triton kernel development experience.
  • Knowledge of NeuronCore-v2 architecture.
  • Experience with FP32, BF16, FP8, and INT8 workloads.
  • Experience benchmarking workloads on Trn1 or Trn2 instances.
  • Familiarity with nki.language, @nki.jit, or XLA custom calls.
  • Experience with technical evaluation, AI/ML data projects, RLHF, or rubric-based assessment.

Culture & Benefits

  • Fully remote engagement available in the listed countries.
  • Part-time, project-based consulting engagement.
  • Focus on AWS Trainium / NKI kernel engineering and technical evaluation.

Будьте осторожны

Если работодатель просит войти в их систему, используя iCloud/Google, прислать код/пароль, запустить код/ПО, не делайте этого. Это мошенники. Обязательно жмите «Пожаловаться» или пишите в поддержку. Подробнее в гайде →