Hackernews posts about R1 671B

  1. How to Run DeepSeek R1 671B Locally on a $2000 EPYC Server (digitalspaceport.com)
    463 points by walterbell 5 months ago | 283 comments
  2. DeepSeek-R1-671B-Q4_K_M with 1 or 2 Arc A770 on Xeon (github.com)
    297 points by colorant 3 months ago | 112 comments
  3. DeepSeek R1 671B running on 2 M2 Ultras faster than reading speed (twitter.com)
    96 points by thyrox 5 months ago | 29 comments
  4. A step-by-step guide on deploying DeepSeek-R1 671B locally (snowkylin.github.io)
    37 points by snowkylin 5 months ago | discuss
  5. DeepSeek R1 671B over 2 tok/s without GPU on local gaming rig (old.reddit.com)
    17 points by tosh 5 months ago | 5 comments
  6. Mac Studio M3 Ultra can run Deepseek R1 671B in memory using <200W (www.techradar.com)
    7 points by mizzao 3 months ago | 1 comments
  7. Running the Deepseek-R1 671B Model at FP16 Fidelity on AMD EPYC CPUs (www.servethehome.com)
    5 points by tanelpoder 2 months ago | 2 comments
  8. SambaNova Cloud Launches the Fastest DeepSeek-R1 671B (198 tok/s) (sambanova.ai)
    5 points by mzl 4 months ago | 2 comments
  9. Lambda Chat – Hosted DeepSeek R1 671B (lambda.chat)
    4 points by sabalaba 5 months ago | 1 comments
  10. DeepSeek R1 671B Running and Testing on a $2000 Local AI Server [video] (www.youtube.com)
    3 points by tosh 5 months ago | discuss
  11. DeepSeek R1 671B running locally [video] (www.youtube.com)
    3 points by nodesocket 5 months ago | discuss
  12. R1 671B at Home – EPYC Inference Build Guide (rentry.co)
    3 points by DrPhish 5 months ago | discuss
  13. FlashMoE: DeepSeek-R1 671B and Qwen3MoE 235B with 1~2 Intel B580 GPU in IPEX-LLM (github.com)
    2 points by colorant about 1 month ago | 1 comments
  14. SambaNova Runs DeepSeek-R1 671B on a Single Rack, Slashing AI Inference Costs (www.emsi.me)
    1 points by hochmartinez 4 months ago | discuss
  15. Show HN: KTransformers:671B DeepSeek-R1 on a Single Machine-286 tokens/s Prefill (github.com)
    14 points by sssummer 4 months ago | discuss
  16. "11 tok/SEC running 671B DeepSeek R1 on two M3 Ultra" – Alex Cheema (twitter.com)
    5 points by ryzvonusef 3 months ago | 1 comments
  17. Quanxing Technology helps 671B DeepSeek R1 training costs drop by another 95% (www.quanxingtech.cn)
    1 points by almaight 4 months ago | 2 comments
  18. Creating "Hello World" executable bytes for LLMs is so hard?
    1 points by artificiology 3 months ago | 3 comments
  19. Run the Full DeepSeek R1 Locally – With Only 32GB RAM (www.gulla.net)
    2 points by tomahg 4 months ago | 1 comments