EmilPi

• Which (1 or 2-story) frame to use for 7 GPU rig?

  Posted by EmilPi@reddit | LocalLLaMA | View on Reddit | 5 comments

• Anyone using RTX 8000 (48GB) or MI100 (32GB) cards for LLM inference?

  Posted by EmilPi@reddit | LocalLLaMA | View on Reddit | 44 comments

• How to estimate prompt processing speed for given (multi-)GPU and model?

  Posted by EmilPi@reddit | LocalLLaMA | View on Reddit | 2 comments

• How to run Gemma 3 27B QAT with 128k context window with 3 parallel requests possible on 2x3090

  Posted by EmilPi@reddit | LocalLLaMA | View on Reddit | 6 comments

• 5 commands to run Qwen3-235B-A22B Q3 inference on 4x3090 + 32-core TR + 192GB DDR4 RAM

  Posted by EmilPi@reddit | LocalLLaMA | View on Reddit | 37 comments

• How to use Qwen2.5-Coder-Instruct without frustration in the meantime

  Posted by EmilPi@reddit | LocalLLaMA | View on Reddit | 32 comments

• Is RTX 50xx series intentionally locked for compute / AI ?

  Posted by EmilPi@reddit | LocalLLaMA | View on Reddit | 19 comments

• 2x RTX 3090 + Threadripper 3970X + 256GB RAM LLM inference benchmarks

  Posted by EmilPi@reddit | LocalLLaMA | View on Reddit | 57 comments

• What is best small long-context open-weight model now?

  Posted by EmilPi@reddit | LocalLLaMA | View on Reddit | 15 comments

• PCIe splitter advice

  Posted by EmilPi@reddit | LocalLLaMA | View on Reddit | 6 comments

• A4500 (20 GB VRAM) + RTX 4060 Ti (16 GB) dual GPU setup

  Posted by EmilPi@reddit | buildapc | View on Reddit | 1 comments

• Dual RTX 4070 Ti Super vs RTX 4090

  Posted by EmilPi@reddit | buildapc | View on Reddit | 2 comments

• Run DeepSeek-V3 with 96GB VRAM + 256 GB RAM under Linux

  Posted by EmilPi@reddit | LocalLLaMA | View on Reddit | 27 comments

• Any Pixtral-Large-2411 quants over there?

  Posted by EmilPi@reddit | LocalLLaMA | View on Reddit | 7 comments

• How to simply run your model at startup in Debian/Ubuntu

  Posted by EmilPi@reddit | LocalLLaMA | View on Reddit | 8 comments

• 4x RTX 3090 + Threadripper 3970X + 256 GB RAM LLM inference benchmarks

  Posted by EmilPi@reddit | LocalLLaMA | View on Reddit | 49 comments

• LLM 4 GPUs rig stability problem

  Posted by EmilPi@reddit | LocalLLaMA | View on Reddit | 22 comments

• 4 GPUs stability problem

  Posted by EmilPi@reddit | buildapc | View on Reddit | 0 comments

• Code LLMs with long context window

  Posted by EmilPi@reddit | LocalLLaMA | View on Reddit | 32 comments

• How to know RAM Threaded Read Speed in advance given RAM module, stick count, mobo's RAM channels count and CPU's PCIe lanes count and gen?

  Posted by EmilPi@reddit | buildapc | View on Reddit | 3 comments

• Re: asked LLMs a riddle about Sally and her sisters + plot twist

  Posted by EmilPi@reddit | LocalLLaMA | View on Reddit | 1 comments