Performant C/CUDA inference engine for Qwen 3.6 35B on RTX 5090 / Blackwell

Software Development, AI & Machine Learning, DevOps & Infrastructure(github.com)view on HackerNews
q36inference engineCUDACGPUQwen 3.6 35BRTX 5090Blackwell

Author: ambuds

Date: 7/13/2026

Article Summary:
A hyper-optimized, zero-dependency C/CUDA inference engine for Qwen 3.6 35B on RTX 5090 / Blackwell.