Strata qwen 3.8 flash on a 3090 card and I'll add a Pic of sys specs.
Local llm. Code focus. Only like 35gig used included is the OS for the machine, Ubuntu. So this can run while I am running over things for sure, easily. Aug is around 75 to 85tps tokens per sec. Strata is a
Testing Qwen 3.8 Flash IQ3_S GSQ RCO on 2× RTX 5060 Ti 16GB with Strata, running a 256K context window.
I’m testing decode speed, prompt processing, expert-cache behavior, and how well Strata’s multi-GPU layer split scales across 32GB total VRAM. Results coming soon!
Finished perfectly by our custom harness, using Qwen-3.8-27B
It did ask me 2 questions ( Whether I was ok getting an early lunch on Monday as no dinner reservations available). I abandoned as it had $5 no show ( and I dont like them much )
Small clip of nuclear physicist Stanton Friedman talking about UAP landings and creatures, found by Qwen 3.8, from SMU archives.
(I used to teach at SMU, years ago.) WFAA - October 18, 1972 "5:08, 23:51 - (October 18, 1972) Nuclear physicist Stanton T.
You have reached the end of the archive
All of qwen38