Inovello https://inovello.dev/ Benchmark writeups and failure notes on running large mixture-of-experts models on two RTX 3090s and 192 GB of DDR4. enSat, 05 Sep 2026 00:00:00 +0000 A silent -ot downgrade under mmap, and a 9-line loader fix https://inovello.dev/writeups/pinned-host-experts-under-mmap/ Sat, 05 Sep 2026 00:00:00 +0000 https://inovello.dev/writeups/pinned-host-experts-under-mmap/ Why -ot ...=CUDA_Host did nothing on master, what my PR #28223 changes, and where the 8 minutes of load time went. Qwen3.8-Flash-Next on 2x3090 + DDR4, part 2: 25-29 to 37-41 t/s with UD-Q4_K_XL, the expert cache, and MTP https://inovello.dev/writeups/qwen3-flash-next-2x3090-q4-mtp/ Fri, 04 Sep 2026 00:00:00 +0000 https://inovello.dev/writeups/qwen3-flash-next-2x3090-q4-mtp/ A quant swap, MTP on top of the cache, a 4x faster load, a bug in the cache PR, and the discovery that my RAM had been thermal throttling the whole time. Qwen3.8-Flash-Next on 2x3090 + DDR4: 17 to 25-29 t/s decode with the expert cache PR https://inovello.dev/writeups/qwen3-flash-next-2x3090-expert-cache/ Thu, 03 Sep 2026 00:00:00 +0000 https://inovello.dev/writeups/qwen3-flash-next-2x3090-expert-cache/ All 48 expert layers in host RAM, an LRU cache of hot experts in VRAM, and the VRAM budget that made it pay.