DFlash2 speeds Qwen 3.8 27B up to 4 times

llama.cpp pr #27342 adds dflash2, so i rented an rtx 6000 and ran the same four prompts through four decoding setups on qwen3.8 27B median results over the four tasks: baseline 47.4 tok/s mtp 114.7 tok/s dflash 99.3 tok/s dflash2 140.6. tok/s so on average 3x for dflash2 though i have to point out that it's far from a 3x gain some of the time, on one of the test it struggled to achieve a 1.5x gain, it really just depends on the task you give to the model i'm from the atomic.chat team - we publish our own quants on hf and make a desktop and mobile app for running local models. so any feedback welcome - we're building this for you folks about dflash2: inco.ai/blog/dflash2