HN Explorer

buildbot

↗ HN profile [ 5360 ms ] full profile
Karma
0
Created
()
Submissions
0
  1. Co-Designing AI Models Using Speculative Decoding for Faster LLM Inference (developer.nvidia.com)
    1 points 0
  2. Why does it take so long to release black fan versions? (noctua.at)
    778 points 50
  3. Nvidia Groq 3 LPX (nvidia.com)
    1 points 0
  4. Nvidia Nemotron 3 Super Delivers 5x Higher Throughput (blogs.nvidia.com)
    3 points 1
  5. Think Smart About Sparse Compute: LatentMoE for Higher Accuracy per Flop, Param (research.nvidia.com)
    2 points 0
  6. Nvidia Nemotron-3 White Paper [pdf] (research.nvidia.com)
    1 points 0
  7. Webb First to Show 4 Dust Shells 'Spiraling' Apep, Limits Long Orbit (science.nasa.gov)
    2 points 0
  8. Emerging Properties in Unified Multimodal Pretraining (arxiv.org)
    1 points 1
  9. Azure Maia for the era of AI: From silicon to software to systems (azure.microsoft.com)
    2 points 0
  10. Making AI More Accurate: Microscaling on Nvidia Blackwell (youtube.com)
    2 points 0
  11. Blackwell Tensor Cores support Microscaling formats for 4B inference (nvidia.com)
    1 points 0
  12. The Layoff (xeiaso.net)
    1 points 0
  13. Pixel owners report problems after installing January 2024 Google Play update (bleepingcomputer.com)
    163 points 160
  14. Neural Network Quantization and Number Formats from First Principles (semianalysis.com)
    1 points 0
  15. Introduction to MicroScaling (MX) for AI Tensors (fpga.org)
    4 points 0
  16. OpenAI is paying out $51M to a chip startup that Sam Altman invested in (businessinsider.com)
    15 points 9
  17. The first two custom silicon chips designed by Microsoft for its cloud (theverge.com)
    247 points 270
  18. Ask HN: How can x/Twitter use groq for their AI legally?

    Groq is also a company selling AI inference chips - groq.com - isn’t that trademark infringement? They have registered a trademark: https://trademarks.justia.com/880/64/groq-88064920.html It specifically refers to both…

    1 points 0
  19. "Serverless", GPU Accelerated OpenAI API Using Podman, Llama.cpp, and Systemd (blog.maxg.io)
    2 points 0
  20. 1 points 0
  21. Bandwhich: Terminal Bandwidth Utilization Tool (github.com)
    4 points 0
  22. Ask HN: What to do when Google Takeout fails

    I've had a google workspace subscription for years, and stored about 12TB in google drive, mostly backups in a single folder. Recently, google decided to hard lock accounts over the pooled storage limit, unless you pay…

    3 points 5
  23. Ask HN: Gmail Spam Out of Control

    I have an older gmail account that's getting upwards of 80 gross spam email via bnc3.mailjet.com but with spoofed sender/receivers per day - literally multiple per minute sometimes. While they go to spam, I'd like to…

    2 points 1
  24. Microsoft just announced a media event taking place tomorrow with OpenAI (windowscentral.com)
    151 points 135
  25. Show HN: Reverse Engineering an Old Digital Back Raw File Format (blog.maxg.io)

    Perhaps a bit outside of the typical interest of HN readers, but I wanted share how I went about figuring out how to convert the RAW files from a digital camera (well, digital back) that originally came out in 2004 to…

    6 points 0