HN Explorer

GreenGames

↗ HN profile [ 78.0 ms ] full profile
Karma
0
Created
()
Submissions
0
  1. 1 points 0
  2. 1 points 0
  3. Someone on Reddit is running Qwen 27B at ~20 tok/s on a jailbroken PS5's GPU (reddit.com)
    2 points 0
  4. 1 points 0
  5. $5,999 Ryzen AI MAX+ 395 and Radeon R9700 workstation (videocardz.com)
    2 points 0
  6. 1 points 0
  7. 1 points 0
  8. 1 points 0
  9. 1 points 0
  10. 1 points 0
  11. We got DSV4 Flash 0731 to 32.7 tok/s on AMD Strix Halo (lucebox.com)
    5 points 0
  12. AMD Lucebox Beats Nvidia DGX Spark by 3.63x on DeepSeek V4 Flash (lucebox.com)
    4 points 0
  13. DeepSeek V4 Flash, up to 32 tok/s on AMD Ryzen AI MAX+ 395 (lucebox.com)
    3 points 1
  14. Laguna XS 2.1 33B on a RTX 3090: 296 tok/s peak, 152 tok/s at 256K context (lucebox.com)
    20 points 1
  15. 256K context with 72MiB of KV cache on the GPU (github.com)
    3 points 0
  16. A 35B MoE on a 16 GB GPU, without the offload tax (lucebox.com)
    18 points 0
  17. PFlash: 10x prefill speedup over llama.cpp at 128K on a RTX 3090 (github.com)
    3 points 2
  18. We got 207 tok/s with Qwen3.5-27B on an RTX 3090 (github.com)
    165 points 15
  19. Show HN: OS Megakernel that match M5 Max Tok/w at 2x the Throughput on RTX 3090 (github.com)

    Hey there, we fused all 24 layers of Qwen3.5-0.8B (a hybrid DeltaNet + Attention model) into a single CUDA kernel launch and made it open-source for everyone to try it. On an RTX 3090 power-limited to 220W: - 411 tok/s…

    6 points 1
  20. Cua (YC X25) is hiring an engineer (ycombinator.com)
    1 points 0 job
  21. App-Use, Control Individual Applications with CUA Agents (trycua.com)
    1 points 1
  22. Show HN: Lumier – Run macOS VMs in a Docker (github.com)

    Hey HN, we're excited to share Lumier (https://github.com/trycua/cua/tree/main/libs/lumier), an open-source tool for running macOS and Linux virtual machines in Docker containers on Apple Silicon Macs. When building…

    159 points 52
  23. Microsoft is reportedly about to lay off 3% of its workforce (techcrunch.com)
    13 points 2
  24. Polaris is giving free GPUs/CPUs for everyone (polariscloud.ai)
    3 points 0
  25. Improvements in reasoning AI models may slow down soon, analysis finds (techcrunch.com)
    3 points 0