Hackernews posts about R1 671B

1. 01
   

   newHow to Run DeepSeek R1 671B Locally on a $2000 EPYC Serverdigitalspaceport.com

   ▲ 463walterbell612d💬 283savesaved
2. 02
   

   newDeepSeek-R1-671B-Q4_K_M with 1 or 2 Arc A770 on Xeongithub.com

   ▲ 297colorant579d💬 112savesaved
3. 03
   

   newDeepSeek R1 671B running on 2 M2 Ultras faster than reading speedtwitter.com

   ▲ 96thyrox615d💬 29savesaved
4. 04
   

   newA step-by-step guide on deploying DeepSeek-R1 671B locallysnowkylin.github.io

   ▲ 37snowkylin613ddiscusssavesaved
5. 05
   

   newDeepSeek R1 671B over 2 tok/s without GPU on local gaming rigold.reddit.com

   ▲ 17tosh613d💬 5savesaved
6. 06
   

   newMac Studio M3 Ultra can run Deepseek R1 671B in memory using <200Wwww.techradar.com

   ▲ 7mizzao562d💬 1savesaved
7. 07
   

   newRunning the Deepseek-R1 671B Model at FP16 Fidelity on AMD EPYC CPUswww.servethehome.com

   ▲ 5tanelpoder550d💬 2savesaved
8. 08
   

   newSambaNova Cloud Launches the Fastest DeepSeek-R1 671B (198 tok/s)sambanova.ai

   ▲ 5mzl599d💬 2savesaved
9. 09
   

   newLambda Chat – Hosted DeepSeek R1 671Blambda.chat

   ▲ 4sabalaba612d💬 1savesaved
10. 10
    

    newDeepSeek R1 671B Running and Testing on a $2000 Local AI Server [video]www.youtube.com

    ▲ 3tosh612ddiscusssavesaved
11. 11
    

    newDeepSeek R1 671B running locally [video]www.youtube.com

    ▲ 3nodesocket616ddiscusssavesaved
12. 12
    

    newR1 671B at Home – EPYC Inference Build Guiderentry.co

    ▲ 3DrPhish618ddiscusssavesaved
13. 13
    

    newFlashMoE: DeepSeek-R1 671B and Qwen3MoE 235B with 1~2 Intel B580 GPU in IPEX-LLMgithub.com

    ▲ 2colorant511d💬 1savesaved
14. 14
    

    newSambaNova Runs DeepSeek-R1 671B on a Single Rack, Slashing AI Inference Costswww.emsi.me

    ▲ 1hochmartinez589ddiscusssavesaved
15. 15
    

    newShow HN: KTransformers:671B DeepSeek-R1 on a Single Machine-286 tokens/s Prefillgithub.com

    ▲ 14sssummer603ddiscusssavesaved
16. 16
    

    new"11 tok/SEC running 671B DeepSeek R1 on two M3 Ultra" – Alex Cheematwitter.com

    ▲ 5ryzvonusef573d💬 1savesaved
17. 17
    

    newQuanxing Technology helps 671B DeepSeek R1 training costs drop by another 95%www.quanxingtech.cn

    ▲ 1almaight603d💬 2savesaved
18. 18
    

    newIs anyone compressing AI models for the 4B people without GPUs or internet?

    ▲ 3yashpxl218d💬 1savesaved
19. 19
    

    newCreating "Hello World" executable bytes for LLMs is so hard?

    ▲ 1artificiology579d💬 3savesaved
20. 20
    

    newRun the Full DeepSeek R1 Locally – With Only 32GB RAMwww.gulla.net

    ▲ 2tomahg605d💬 1savesaved

· edit