24 sep
|
Anyone AI
|
Madrid
Anyone AI is seeking experienced engineers to evaluate and improve kernel development tasks for AI workloads on AWS Trainium and Inferentia2 hardware. You will work on NKI kernels, migrations from CUDA, and performance benchmarking to optimize memory usage and DMA scheduling.
You will review Trainium-specific patterns, optimize for SBUF/PSUM/HBM hierarchies, and ensure idiomatic, hardware-aware implementations rather than direct GPU translations. This role focuses on kernel-level excellence.
#J-18808-Ljbffr
📌 NKI Kernel Engineer – AWS Trainium Performance (Madrid)
🏢 Anyone AI
📍 Madrid