AMD Strix Halo

2 articles tagged with this topic

LocalLLaMAllama.cpp

AI Inference Goes Specialist: New Engines Drop Generality to Max Out Hardware

A wave of narrow inference engines optimized for one model + one chip is emerging. AI deployment is shifting from "one engine for all" to "one engine

Oct 32 min read

Claimed 70 tokens/sec fails real test — AI benchmark inflation strikes again

Gufo claims 70 tokens/sec; real tests show ~38, 13% slower than Halogen. Benchmark task lets speculative decoding cheat. Another AI inflation case.

Oct 12 min read