Footer

Your go-to destination for discovering trending open-source projects and uncovering the insights that matter

Legal

  • Terms and conditions
  • Privacy policy

More

  • Pricing
  • Advertise
  • onedump
  • gosnakego
  • Contact us

Social

  • X (Twitter)
  • GitHub
Trendshift owl
What they build speaks louder than words
© 2026 Trendshift, created by Julian Li with ❤️
SignalAdvertise
Trendshift
SignalAdvertise
Sign in
Trendshift owl

Discover trends that matter

DailyWeeklyMonthlyYearly
Live mentionsTopics
RepositoriesDevelopers
InsightsStats

Bifrost
Bifrost
Busbar
Busbar
ClawBench
ClawBench

sybil-solutions/dsv41-flash-offload

DeepSeek-V4.1-Flash EXL3 on one 24 GB RTX 3090 + DDR4 + NVMe: staged-DMA prefill, AVX2 CPU expert tier, elastic VRAM expert cache, Engram on disk, OpenAI API

#LLM inference
Visit GitHub
Python
149
22
2 contributors
MIT License
FeaturedAdvertise to 180K+ monthly visitors
Bifrost

Enterprise AI Gateway Built For Scale: Route, Govern, and Secure All Your AI Traffic

Busbar

Control what AI can do before it acts

ClawBench

a benchmark for personal agents

Momentum

How fast sybil-solutions/dsv41-flash-offload is gaining stars, and where it lands on Trendshift and GitHub Trending

Activity and trending

Stars, forks and merged pull requests each day or month, lined up with the repository's rank on Trendshift Trending

Oct 1 – Oct 22, 2026

Stars +147 in 22 days

0204060
44

Forks +22 in 22 days

0246
6

Merged PRs 1 in 22 days

01
1

Trendshift Trendingwhere the repository ranked each day it made the top 25

Not in the Trendshift top 25 in this period

Oct 1Oct 4Oct 7Oct 10Oct 13Oct 16Oct 19Oct 22
The left and right arrow keys step from day to day; Home and End go to the first and the last

Stars, forks, merged pull requests and Trendshift Trending rank, by day
DayStars gainedForks gainedMerged PRs
Thu, Oct 1, 2026+0+00
Fri, Oct 2, 2026+12+10
Sat, Oct 3, 2026+13+40
Sun, Oct 4, 2026+6+00
Mon, Oct 5, 2026+44+50
Tue, Oct 6, 2026+21+20
Wed, Oct 7, 2026+6+01
Thu, Oct 8, 2026+16+60
Fri, Oct 9, 2026+14+40
Sat, Oct 10, 2026+2+00
Sun, Oct 11, 2026+2+00
Mon, Oct 12, 2026+0+00
Tue, Oct 13, 2026+1+00
Wed, Oct 14, 2026+0+00
Thu, Oct 15, 2026+1+00
Fri, Oct 16, 2026+2+00
Sat, Oct 17, 2026+1+00
Sun, Oct 18, 2026+2+00
Mon, Oct 19, 2026+2+00
Tue, Oct 20, 2026+0+00
Wed, Oct 21, 2026+1+00
Thu, Oct 22, 2026+1+00
Total+147+221
  • 1#1 on the list
  • 7Rank that day
  • Not trending

Social mentions

Recent discussions about this repository across the web
  1. Oct 62 posts
    • 0xSero实测更新:Deepseek-v4.1-Flash 在 CPU + 单卡 3090 环境下运行 - KV 缓存:262k - Prefill 预填充速度:600–800 tok/s - c1 解码速度:26 tok/s - c4 解码速度:41 tok/s 硬件需求: - 1 张 24GB 显存 GPU(3090) - 256GB DDR4/DDR5 内存 - 200GB NVMe 固态

      Read full on X, posted by @MirrortekUK

      @MirrortekUK · x.com

    • 一张3090跑DeepSeek-V4.1-Flash,门槛不在显卡。 0xSero开源dsv41-flash-offload:EXL3 3.0bpw量化,约200GB专家权重放内存,热专家缓存进24GB显存,冷专家交AVX2 CPU算。单流decode 19.3 tok/s。代价:256GB内存、24核以上CPU、约1TB NVMe。 省下的显卡钱,全进了内存条。

      Read full on X, posted by @Alexju4542121

      @Alexju4542121 · x.com

  2. Oct 21 post
    • You thought I was joking 🤓 Here's Deepseek-v4.1-Flash running on CPU + DDR4 with just a single RTX 3090 Wow @MiaAI_lab great quant (:

      Read full on X, posted by @0xSero

      @0xSero · x.com

You might also like

Repositories close to sybil-solutions/dsv41-flash-offload
FeaturedAdvertise to 180K+ monthly visitors
Busbar

Control what AI can do before it acts

Bifrost

Enterprise AI Gateway Built For Scale: Route, Govern, and Secure All Your AI Traffic

ClawBench

a benchmark for personal agents

MiaAI-Lab/DeepSeek-v4.1-Flash-EXL3-2x-DGX-Sparks

Python

DeepSeek v4.1 Flash EXL3 2.9 bpw for 2x DGX Sparks

27944

sybil-solutions/deepseek-v4.1-flash-4x-rtx-pro-6000

Python

DeepSeek V4.1 Flash on 4x RTX PRO 6000 Blackwell: native weights, DSpark, vision, NVMe or locked RAM Engram offload.

7712

MiaAI-Lab/DeepSeek-v4-Flash-One-DGX-Spark

Python

DeepSeek v4 Flash EXL3 on one DGX Spark

34938

sybil-solutions/glm-flash-lite

Python

GLM-5.3-Flash EXL3 on one 24 GB RTX 3090 + DDR4: elastic GPU expert cache, zero-copy experts, AVX2 CPU tier, OpenAI API

32035

yhfgyyf/vllm-deepseek-v4-sm89

Python

Run DeepSeek-V4.1-Flash / DeepSeek-V4-Flash and GLM-5.3-Flash on SM89 (Ada / RTX 4090) and SM120 (RTX PRO 6000) with vLLM

19324

0xSero/deepseek-v4-flash-0731-spark-sparkinfer

Python

Pinned one-DGX-Spark Docker recipe for DeepSeek V4 Flash with EXL3, SparkInfer, and 262K NVFP4 MLA KV cache

16619

0xBakeer/deepseek-v41-flash-spark

Python

DeepSeek-V4.1-Flash on a single DGX Spark (GB10): resident hot experts + NVMe streaming, DSpark, OpenAI API. Work in progress.

12616

ryanzhou/deepseek-v4-flash-mi300x

Python
15718

Entrpi/ds4-on-spark

Shell

Entrpi/ds4, a Blackwell CUDA perf fork of antirez/ds4 on NVIDIA DGX Spark: one-command install, ~3x upstream prefill, ~1.5x decode, DSpark, and full continuous batch support

41032