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

SignalAdvertise
DailyWeeklyMonthlyYearly
Live mentionsTopics
RepositoriesDevelopers
InsightsStats
Busbar
Busbar
Omnigraph
Omnigraph
Websitebench
Websitebench
Bifrost
Bifrost

Picom-code/pico-dgx-ds4

DeepSeek V4.1 Flash on 4x DGX Spark: 438 tok/s at 16 users, ~80/138 single-user, 1M context. Fork of blaesus/dsgx10.

#LLM inference
Visit GitHub
Cuda
105
16
1 contributors
MIT License
AchievementsView all →
FeaturedAdvertise to 180K+ monthly visitors
Busbar

Control what AI can do before it acts

Omnigraph

Versioned graph database on object storage. Git-style workflows and multimodal retrieval

Websitebench

Can AI agents reverse-engineer real websites?

Bifrost

Fastest Enterprise AI Gateway for Routing, Governing, and Securing all your AI Traffic

Momentum

How fast Picom-code/pico-dgx-ds4 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

Nov 4 – Nov 13, 2026

Stars +103 in 10 days

0204060
48

Forks +16 in 10 days

0246
5

Merged PRs 0 in 10 days

01

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

Top 25 in Cuda

14251711
Nov 4Nov 6Nov 8Nov 10Nov 12
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 PRsCuda
Wed, Nov 4, 2026+0+00Not trending
Thu, Nov 5, 2026+2+00#14
Fri, Nov 6, 2026+15+20#2
Sat, Nov 7, 2026+4+30#5
Sun, Nov 8, 2026+0+00Not trending
Mon, Nov 9, 2026+2+10#17
Tue, Nov 10, 2026+4+00#11
Wed, Nov 11, 2026+19+50Not trending
Thu, Nov 12, 2026+48+50Not trending
Fri, Nov 13, 2026+9+00Not trending
Total+103+160
  • 1#1 on the list
  • 7Rank that day
  • Not trending

Mentions and achievements

Where people are talking about Picom-code/pico-dgx-ds4, and every badge it has earned
    1. Nov 122 posts

      • The speed of your recipe is amazing. Blows everything out of the water, both in prefill (yes, I got 10k!) and c1! Where it stalls is when using c3+ agentic work, that mixes in burst of prefill with…

        Read full on X, posted by @pedrib1337

        @pedrib1337 · x.com

      • Well, I will have to spend time trying to figure this out. Chinese characters bleeding out of the outputs. Not everything is just speed. Quality matters too. Even more if we claim to be using the…

        Read full on X, posted by @calvarado2004

        @calvarado2004 · x.com

    2. Nov 61 post

      • pico pico wil take us by storm and then something new will come along

        Read full on X, posted by @ClawOperations

        @ClawOperations · x.com

    3. Nov 51 post

      • 438 tok/s at 16 users DeepSeek V4.1 Flash on 4x DGX Spark, now open source: 80 chat / 138 code single stream 438 tok/s at 16 users 1M context, 10k tok/s prefill image support soon

        Read full on X, posted by @Picomoorhead

        @Picomoorhead · x.com

    You might also like

    Repositories close to Picom-code/pico-dgx-ds4
    FeaturedAdvertise to 180K+ monthly visitors
    Websitebench

    Can AI agents reverse-engineer real websites?

    Busbar

    Control what AI can do before it acts

    Bifrost

    Fastest Enterprise AI Gateway for Routing, Governing, and Securing all your AI Traffic

    Omnigraph

    Versioned graph database on object storage. Git-style workflows and multimodal retrieval

    MiaAI-Lab/DeepSeek-v4.1-Flash-DGX-Sparks

    Python

    DeepSeek-V4.1-Flash on 3-4x NVIDIA DGX Sparks

    29754

    tpurtell/ds41rt

    Rust

    Native DeepSeek V4.1 Flash inference across one RTX PRO 6000 and four DGX Spark expert workers

    425

    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

    41432

    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.

    12817

    elsung/dgx-spark-deepseek-v4-flash

    Python

    2x DGX Spark (GB10) running DeepSeek-V4-Flash: dual-node FP8 TP=2 (~41 tok/s single / ~350 agg), single-node ds4, a 10-model GGUF benchmark suite, and a cross-machine comparison (Mac M2 Ultra, RTX PRO 6000). Settings, recipes, gotchas & benchmarks.

    303

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

    Python

    DeepSeek v4 Flash EXL3 on one DGX Spark

    35039

    MiaAI-Lab/DeepSeek-V4.1-Flash-1017-4x-DGX-Sparks-TensorFold

    Shell

    DeepSeek-V4.1-Flash-1017 on 4 DGX Sparks with TensorFold, switchless support

    7526

    0xSero/DeepSeek-V4.1-Flash-Two-Sparks

    Shell

    DeepSeek-V4.1-Flash on two DGX Sparks: EXL3 routed experts, 262k context, 2M-token KV, vision, tools, MTP, bundled with Pi

    959

    MiaAI-Lab/DeepSeek-v4.1-Flash-1017-2x-DGX-Sparks

    Python

    DeepSeek-V4.1-Flash-1017 EXL3 on 2x DGX Sparks with TensorFold

    47587