Dual RTX 3090 benchmarks: Chain-of-Draft lifts GSM8K from 23% to 77%

Eric SungSep 43 stars

Tests Qwen3.8-27B and Flash-Next on two 3090s and finds a terse prompting style dramatically improves math scores.

Open on GitHub

Preview of Dual RTX 3090 benchmarks: Chain-of-Draft lifts GSM8K from 23% to 77%

Launch: Qwen3.8-Flash-Next
Use case: Benchmarks
Format: Repo
Source: github.com

Eric Sung

Designer / Developer / Filmmaker. Also experimenting with lotsa stuff.

See all

0xSero

@0xSero · Sep 29

Reports prefill and decode speeds for a single 24-32GB GPU plus NVMe, including an Intel Arc B70 result.

Niko1221

@Niko1221 · Sep 24

Open-source inference engine with an OpenAI/Anthropic-style local API that fits the 125B MoE onto a 12GB consumer GPU plus system RAM. The project behind the model's local-running surge.

MiaAI-Lab

@MiaAI-Lab · Sep 4

Step-by-step recipe to squeeze the 125B model onto one DGX Spark instead of two, with config and measured throughput.

peonist-ai

@peonist-ai · Aug 26

Container-based server tuned for the gfx1151 Strix Halo APU, aimed at squeezing maximum tokens per second from the 125B model on a mini-PC.

Open on GitHub

Dual RTX 3090 benchmarks: Chain-of-Draft lifts GSM8K from 23% to 77%

More built with Qwen3.8-Flash-Next