R9V Update: Created and adopted KVA projections based on Deepseek V4.1 Flash + HySparse2/MiMo-V3 for Qwen3.8 Flash Next. This is a game changer for models that don't natively implement it. 1.45-1.85x speedup in prefill to 3k+ at a small deficit to perplexity. [2x R9700, 128GB DDR5]

R9V has implemented KVA projections, derived from Deepseek V4.1 Flash and HySparse2/MiMo-V3, for Qwen3.8 Flash Next. This innovation significantly boosts prefill speed by 1.45-1.85x up to 3k+ tokens, with a minor impact on perplexity. Performance metrics include a prefill rate of 3,600 t/s at 64k tokens, decode at 74.7 t/s, and a combined concurrent rate of 70.3 t/s across 8 streams. This development is particularly impactful for models lacking native KVA projection support.