Qwen3.8-Flash-NextDGX Spark · NVFP4 · TensorFoldDecode · tok/s74.8Prefill—tok/100W31.2Qwen3.8-Flash-Next (NVFP4, MTP-6). Fixed 2,048-token code task: 74.8 tok/s single-stream decodeRank is the quoted 74.8 single-stream. Context sweep 56–65 skipped as a range. 50.8 at 260k stays in the note.Officialclaimed@redp314TensorFoldWatchCopy linkOpen pair