llama.cpp Vulkan SCALE fix PR #32735 delivers 2.3-3.2x speedup on Radeon R9700
pmttyji · reddit · 2026-11-22
Revertron's llama.cpp PR #32735 (vulkan: support non-contiguous src in SCALE) yields massive gains. Benchmarked with Inkling-2-Mini-UD-Q8KXL.gguf on a Radeon AI PRO R9700: pp512 jumps from 507.3 to 1604.8 tok/s (3.2x), tg128 from 19.1 to 53.8 tok/s (2.8x), and at d16384 depth pp512/tg128 improve 2.3x/2.5x. Tested via llama-bench -ngl 99 -fa 1 -ctk q80 -ctv q80.
More from Infra
- MLX over Vulkan hits 124 tok/s on RTX 5060 Ti, beating llama.cpp CUDA by 25% — SumitGup · 2026-11-22
- Intruder swapped Prime Intellect's hosted DeepSeek for an abliterated copy for 53 hours — mawcopolow · 2026-11-22
- Token economics model projects $1.14T global API spend by 2030, but only 1% of users pay beyond coding — AccBalanced · 2026-11-22
- Zero-degradation Strata tuning: fixing layer_split lifts decode ~10-15% and halves CPU expert misses — PathfinderTactician · 2026-11-22
- $1400 decade-old V100 32GB runs Qwen 27B: Q4_K_M +48% gen speed, Q8_0 +16% prefill — PuzzleheadedArm3732 · 2026-11-22
- Inference Stack Launches One-Command Self-Hosted LLM Serving with vLLM, LiteLLM and llm-d — AccBalanced · 2026-11-22