ilmoi

14 ms·

1. 31.▲
   by ilmoi 6y ago

   CUDA develpment is totally fine. The service was designed with DL in mind, but that shouldn't stop CUDA dev in any way.
2. 32.▲
   by ilmoi 6y ago

   Wow thanks for posting this! I defo haven't thought about this use case. So interesting. So we actually explicitly prohibit mining in the T&Cs, it's just that in the FAQ I wanted to be a bit more human and dissuade people. Als
3. 33.▲
   by ilmoi 6y ago

   Thanks! I used Tailwind CSS and built the theme from scratch, but I drew inspiration from: - https://nostalgic-css.github.io/NES.css/ - https://jdan.github.io/98.css/
4. 34.▲
   by ilmoi 6y ago

   Yep you're correct that V100s are available as spot instances. If you're willing to write some extra code for saving down weights you could easily go with them. Actually, I'm renting V100s. Got lucky to know the right person
5. 35.▲
   by ilmoi 6y ago

   You are correct that vast.ai is pretty close. The biggest difference is probably security / guaranteed uptime. With vast you're getting what it says on the tin - a machine from a marketplace. Could come from anyone / anywhere
6. 36.▲
   by ilmoi 6y ago

   hah I am European indeed but not Dutch:) But I have spent a year living in the Netherlands.
7. 37.▲
   by ilmoi 6y ago

   Do you think it's the scripts from stripe that are blocking the view? I would guess it's auth0. This was my first front-end project so any feedback 100% welcome.
8. 38.▲
   by ilmoi 6y ago

   Exactly the same as you would be had you been running an EC2 instance at AWS. The machines are hosted, maintained and managed in exactly the same way in a whitelabel DC.
9. 39.▲
   by ilmoi 6y ago

   you'll be making around $6/d, while losing $24h paying for the service.
10. 40.▲
    by ilmoi 6y ago

    Sorry I should have been clearer in my reply. Thanks for clarifying!
11. 41.▲
    by ilmoi 6y ago

    Tesla P100 is the previous generation. P100 > V100 > A100. You can see detailed benchmarking here - http://ai-benchmark.com/ranking_deeplearning.html
12. 42.▲
    by ilmoi 6y ago

    Not really. Firstly it will be way too expensive for you (cloud gamins is like $10/mo, this is $10/10h). Secondly the GPUs installed (Tesla V100s) are not designed for gaming - rather for ML & scientific computing. You should
13. 43.▲
    by ilmoi 6y ago

    thanks for reading the FAQ! Most people don't seem to bother:)
14. 44.▲
    by ilmoi 6y ago

    I’m a self taught ML engineer, and when I was starting on my ML journey I was incredibly frustrated by cloud services like AWS/GCP. a)they were super expensive, b)it took me longer to setup a working GPU instance than to learn to build
15. 45.▲
    Show HN: Cloud GPUs for Deep Learning – At 1/3 the Cost of AWS/GCP(gpu.land)

    168 points by ilmoi 6y ago|93 comments
16. 46.▲
    by ilmoi 6y ago

    In case anyone needs cheap GPU compute, I built a little side project - https://gpu.land/ . You can rent dirt cheap Tesla V100s at $0.99/hr. That's 1/3 of what you'd pay at GCP/AWS/insert big cl
17. 47.▲
    by ilmoi 6y ago

    I always found that annoying. My best best is crypto mining defense. Btw if you're looking for cheap cloud GPUs here's a little side project of mine - https://gpu.land/ . You can rent dirt cheap Tesla V100s at $0.9
18. 48.▲
    Show HN: Crypto Mining Pools Aggregator(github.com)

    3 points by ilmoi 6y ago|1 comments
19. 49.▲
    by ilmoi 6y ago

    Built it out of necessity for one of my projects. With crypto prices rallying there's definitely more mining. Use this to get an up to date list of mining pool IPs / domains for your firewall / hosts file.