{"meta":{"count":1,"page":1,"pages":1,"per_page":12,"applied":{"engine":null,"model":null,"quant":null,"storage":null,"hardware":"rtx-pro-6000-blackwell-x3","generation":null,"task":null,"tier":null,"tags":[],"vram":null,"search":null,"sort":"updated","featured":false},"origin":"https://localairecipes.com","openapi":"/openapi.json","updated_at":"2026-10-31T22:48:55+00:00"},"recipes":[{"slug":"deepseek-v4-1-flash-1017-kraken-tp3-3x-rtx-pro-6000-blackwell","title":"DeepSeek-V4.1-Flash-1017 Kraken TP3 3x RTX PRO 6000 Blackwell","summary":"DeepSeek-V4.1-Flash-1017 TP3 on three 96 GB RTX PRO 6000 Blackwell cards, full 1M context, DeepSeek's original weights (FP8 dense + MXFP4 experts, nothing requantized). Expert layers 20-33 park in host RAM against a 1.5 GiB promotion reserve; once healthy their hottest experts promote back to GPU. Promotion order is a blend of real coding and prose traffic (expert-order-blend/, 0.6/0.4), not the image's 17.8K-token sample that left ~15% of parked-layer routes cold: 72 -> 79.5 engine steps/s at 32K C1 (+10%), 214 tok/s, coding 336/374 max, no quality or KV cost. Four KV presets (2x/4x/6x/8x = 2.18M-8.50M tokens) trade promoted experts for cache; 2x default, all pass the eviction stress. Engram layer 1 in host RAM, layer 14 from NVMe. DSpark spec decode, NVFP4 drafter head. 1M window over a 2,185,581-token pool. Image 20261019.3, 2026-10-20 restart: prefill 13,486/12,245 tok/s (32K/128K), 7,193 at 983K; decode 186.7/340.7 c1/c4 at 32K, 179.2/307.1 at 128K, 474 agg, 271.8 coding peak.","tier":"community","beta":false,"generation":"blackwell","decode_tok_s":228.5,"prefill_tok_s":11445.5,"updated_at":"2026-10-31 11:51:47.603558-06","commit_sha":"b55f8454d9aa87af98859d0342dd7c87a98261cf","content_hash":"91affb9e1c071510e9d62dca745c796155565d42","votes":0,"credit":[{"name":"BoxTops4Coochie (recipe, TP3 port, measurements)","url":"https://github.com/BoxTops4Coochie/RTX-PRO-6000-TP3-Inference-Recipes"},{"name":"infernix - ds41f-tp3 expert-parking design on karmic-kraken","url":"https://github.com/infernix"},{"name":"Local Inference Lab - vLLM karmic-kraken fork + FlashInfer b12x kernels","url":"https://github.com/local-inference-lab"},{"name":"DeepSeek - DeepSeek-V4.1-Flash-1017 base model (MIT)","url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017"}],"uploaded_by":"localai-recipes","measured":null,"links":{"self":"/api/v1/recipes/deepseek-v4-1-flash-1017-kraken-tp3-3x-rtx-pro-6000-blackwell.json","alternate":"/r/deepseek-v4-1-flash-1017-kraken-tp3-3x-rtx-pro-6000-blackwell"}}]}