Hardly any benchmarks for Intel B65, so I got 4x and tested them. Long Context Sweeps for Flash-Next and other models
EDIT: Note that I built off the official vllm documentation and I just got done validating the system.
Proof1 community thread
Key points
- That means this is a slow and stable config. Now that i have a baseline control I’m going to begin tweaking settings for performance runs.
- New batch of tests from my 4x B65 rig I built, because I was curious to test how the high memory bandwidth works out in reality.
- So far only doing it with flash-next, but interesting results in the "by model" section of my site for flash-next.
- Qwen3.8 Flash-Next and 27B — Intel Arc Pro B65 context sweeps
Sources (1)
- [1]Hardly any benchmarks for Intel B65, so I got 4x and tested them. Long Context Sweeps for Flash-Next and other modelsr/LocalLLaMA (top, daily) · Oct 20, 03:57 PM
EDIT: Note that I built off the official vllm documentation and I just got done validating the system.
That means this is a slow and stable config. Now that i have a baseline control I’m going to begin tweaking settings for performance runs.
Extractive summary: sentences quoted from the sources.
Before this
- Oct 20, 2026[AINews] DeepSeek DeepInfer: Open Source SOTA Inference running 155T toks/day/cluster making $22m/day at ~80% margin for a 81 day payback
- Oct 19, 2026Qwen3.8-27B and Flash-Next on consumer hardware: engine numbers I could verify, and what they cost with 2026 RAM prices
- Oct 19, 2026Kapitsa: open-source Apple Silicon optimized Metal engine built for Qwen and Nemotron. See benchmarks of Qwen 27B, Nemotron + 8 other models on an M1 Pro (16GB), M4 Pro (48GB) and M5 Max (128GB)
- Oct 19, 2026NVIDIA Open Sources Run:ai Under Apache 2.0 for Kubernetes GPU Orchestration
- Oct 19, 2026sgl-project/sglang v0.5.22
- Oct 19, 2026Qwen 3.8 Flash Next optimized for two CMP 170hx GPUs