Hackernews posts about H100s

1. A100s are more cost-efficient than H100s for small training tasks (cloudbreakr.substack.com)
   3 points by jack_nclx 21 days ago | 1 comments
2. The "setup tax" on AWS H100s is killing iterative research
   3 points by miyamotomusashi 1 day ago | discuss
3. Selling H200s to China Is Unwise and Unpopular (thezvi.substack.com)
   6 points by paulpauper 23 days ago | discuss
4. Nvidia can sell H200s to China – if Washington gets a 25 percent cut (www.theregister.com)
   5 points by pseudolus 23 days ago | 2 comments
5. MI300X vs. H100 vs. H200 Benchmark Part 1: Training (newsletter.semianalysis.com)
   5 points by nl 25 days ago | discuss
6. Selling H200s to China Is Unwise and Unpopular (thezvi.wordpress.com)
   2 points by speckx 23 days ago | 1 comments
7. The "4-Hour Barrier": Forensic Audit of Sm_90 Logic Decay on Nvidia H100
   2 points by Stan_Byriukov 3 days ago | discuss
8. Letting Nvidia sell H200s to China is closing the door after horse has bolted (www.theregister.com)
   2 points by pseudolus 22 days ago | discuss
9. Starcloud-1 satellite reaches space, with Nvidia H100 GPU now operating in orbit (www.datacenterdynamics.com)
   1 points by QueensGambit 22 days ago | 1 comments
10. Show HN: Solving the H100 OOM Wall with CTDR – Maxwell Dashboard Included (github.com)
    1 points by Stan_Byriukov 5 days ago | discuss
11. Benchmark: A100 vs. H100 NVMe Random Read throughput during multi-GPU loading
    1 points by pveldandi 22 days ago | discuss
12. Prediction Market Investors – Where Do I Find Them?
    7 points by h100ker 14 days ago | 8 comments
13. Show HN: PolyBets – Prediction Market for Auctions (Bet on Auction Results) (polybets.fun)
    5 points by h100ker 24 days ago | 2 comments
14. $2 H100s: How the GPU Rental Bubble Burst (www.latent.space)
    403 points by swyx about 1 year ago | 279 comments
15. Musk confirms 12K H100s ordered for Tesla were instead prioritized for xAI (www.theregister.com)
    21 points by mfiguiere over 1 year ago | 2 comments
16. Meta buys 600k H100s to train LLaMa3 (twitter.com)
    18 points by yuvalsteuer almost 2 years ago | 15 comments
17. Show HN: TensorDock – GPU Cloud Marketplace, H100s from $2.49/hr (dashboard.tensordock.com)
    10 points by jonathanlei over 1 year ago | 2 comments
18. Cerebras launches processor for AI theoretically equivalent to 62 Nvidia H100s (www.tomshardware.com)
    9 points by reqo almost 2 years ago | 2 comments
19. Show HN: Prime Intellect GPU Cloud. H100s starting at $1.65/HR on demand (app.primeintellect.ai)
    7 points by johannesha over 1 year ago | discuss
20. Show HN: GPU price-per-hour tracker for A100/H100s (computeindex.michaelgiba.com)
    6 points by michaelgiba over 1 year ago | discuss
21. Show HN: A GPU group-buying layer to get enterprise rates (~50% off H100s) (trysaga.ai)
    5 points by DdotR 11 months ago | discuss
22. Ask HN: Are Nvidia H100s that good?
    4 points by jonathanlei over 1 year ago | 5 comments
23. Ask HN: Could advanced M-Line chips ever substitute for Nvidia H100s?
    4 points by Devolver over 1 year ago | 3 comments
24. Ask HN: Features for GPU price-per-hour tracker for A100/H100s
    2 points by michaelgiba over 1 year ago | 3 comments
25. Scaling Coding-Agent RL to 32x H100s. 160% Improvement on Stanford's TBench (github.com)
    2 points by Danau5tin about 2 months ago | 1 comments
26. Free H100s for All .EDU User Emails (app.hyperbolic.xyz)
    2 points by hyperbolic_labs 8 months ago | discuss
27. Who's Hoarding Nvidia H100s? (sherwood.news)
    2 points by jonathanmkeegan over 1 year ago | discuss
28. LLM-Training-Puzzles: What would you do with 1000 H100s? (github.com)
    2 points by sebg over 1 year ago | discuss
29. Meta will have more than 350.000 Nvidia H100s this year (www.instagram.com)
    1 points by jbverschoor almost 2 years ago | 2 comments
30. Flux Fast: Making Flux Go Brrr on H100s (pytorch.org)
    1 points by sayak_paul_hf 6 months ago | discuss
31. Show HN: gpudeploy.com – "Airbnb" for GPUs (www.gpudeploy.com)
    267 points by nicowaltz over 1 year ago | 188 comments
32. Show HN: Terminal-Bench-RL: Training long-horizon terminal agents with RL (github.com)
    125 points by Danau5tin 5 months ago | 12 comments
33. Show HN: H100cloud.com – Dutch Auction Cloud Exchange (h100cloud.com)
    11 points by zandrew over 1 year ago | discuss
34. Show HN: Voltage Park – H100 GPU Orderbook (auction.voltagepark.com)
    10 points by mdu96 over 1 year ago | 3 comments
35. Show HN: I built a website where you can easily fine-tune Llama 3.1 models (www.tunellama.com)
    8 points by hamadafm over 1 year ago | 4 comments
36. Ask HN: Is anyone using AMD GPUs for their AI workloads?
    6 points by technoabsurdist 6 months ago | 2 comments
37. Show HN: Airbnb for GPUs (www.gpudeploy.com)
    2 points by lschneider over 1 year ago | discuss
38. So you want to rent an NVIDIA H100 cluster? 2024 Consumer Guide (www.photoroom.com)
    297 points by ea016 over 1 year ago | 139 comments
39. Deploying DeepSeek on 96 H100 GPUs (lmsys.org)
    285 points by GabrielBianconi 4 months ago | 80 comments
40. AMD's MI300X Outperforms Nvidia's H100 for LLM Inference (www.blog.tensorwave.com)
    280 points by fvv over 1 year ago | 264 comments
41. Processing Strings 109x Faster Than Nvidia on H100 (ashvardanian.com)
    216 points by ashvardanian 3 months ago | 26 comments
42. Google TPU v5p beats Nvidia H100 (www.techradar.com)
    174 points by wslh almost 2 years ago | 133 comments
43. Intel Gaudi2 chips outperform Nvidia H100 on diffusion transformers (stability.ai)
    146 points by memossy almost 2 years ago | 60 comments