community pool / model answer

Signal-3.8-27B

chat · agentionai/Signal-3.8-27B

50.7 tok/s · model median · 5 runs in the model record.

Answer

69.4tok/s

measuredRadeon AI Pro R9700 32GB ×38-bit · 3 runs · decodebest cell in the pool

  1. Radeon AI Pro R9700 32GB ×369.4 tok/smeasured3 runs · peak VRAM no data yet · 96 GB rig memory
  2. RX 7900 XTX 24GB ×226.3 tok/sreported2 runs · peak VRAM no data yet · 48 GB rig memory
Reference rigQuantizationMedian resultBasisTTFT (ms)Peak VRAMMax contextEngines
Radeon AI Pro R9700 32GB ×38-bit69.4 tok/smeasured397- GB262,144llama.cpp
RX 7900 XTX 24GB ×28-bit26.3 tok/sreported731- GB32,768llama.cpp

Context tested

The table above is the tested context for each exact rig and quantization cell. Empty fields remain empty rather than being filled with an estimate.