Linode GPU Instances • Hacker News Board

Re: Linode GPU Instances

#71

Post by tkulick » Thu, Jun 20, 2019, 12:50 PM UTC

sieabahlpark wrote: post #52

Who_me wrote: post #34

Hey peeps full disclosure I work as one of Linode's RnD engineers. I want to try to get to as many of these as I can. One of the biggest questions is why the Quadro RTX 6000? Few things: 1. Cost it has the same performance as the 8000. The difference is 8 more GB of RAM that comes at a steep premium. Cost is important to us as it allows us to be at a more affordable price point. 2. We have all heard or used the Tesla…

Has linode improved their security intrusion and disclosure policy yet? These are great improvements but are virtually worthless if linode didn't change their behavior.

(Tory from the Linode team here)

We made some improvements to our disclosure / Bug Bounty program last year and launched this on HackerOne. The community and quality of submissions has been great. More information: https://blog.linode.com/2018/05/16/linodes-new-bug-bounty-pr...

We've also been making ongoing improvements to our application security and security infrastructure through the implementation of a DevSecOps culture. This is something we take very seriously.

Re: Linode GPU Instances

#72

Post by Crosseye_Jack » Thu, Jun 20, 2019, 12:54 PM UTC

sannee wrote: post #67

Crosseye_Jack wrote: post #64

Earlier quoted context omitted.

Tech speaking it’s nVidia’s GeForce driver that restricts datacenter usage not the card itself. Not deep dived into it but maybe using nouveau instead of GeForce works around that restriction. You are allowed to use the driver in data centres for cryptocurrency usage. The EULA limited datacenter usage hasn’t really been challenged in court yet. Both sides would have an argument. NVidia are using the Eula to limit an…

You can't use nouveau for CUDA, which sort of negates the whole point of having a GPU in a server in the first place.

I've not deep dived into nouveau for a while so wasn't sure if they added cuda support in the past couple of years since I played with it which is why I only said "maybe".

Re: Linode GPU Instances

#73

Post by Who_me » Thu, Jun 20, 2019, 1:03 PM UTC

trsohmers wrote: post #53

Who_me wrote: post #34

Hey peeps full disclosure I work as one of Linode's RnD engineers. I want to try to get to as many of these as I can. One of the biggest questions is why the Quadro RTX 6000? Few things: 1. Cost it has the same performance as the 8000. The difference is 8 more GB of RAM that comes at a steep premium. Cost is important to us as it allows us to be at a more affordable price point. 2. We have all heard or used the Tesla…

> The difference is 8 more GB of RAM that comes at a steep premium This is incorrect. The RTX 6000 has 24GB of VRAM and is $4000, and the RTX 8000 has 48GB of VRAM (double the amount) and is $5500. Is it worth the price increase? For a lot of people I know it is. Also, the RTX Titan is $2500 and is identical to the RTX 6000 (at the chip level) and also with 24GB of VRAM, with the only difference being software enabli…

I fact checked myself after I read this and you are correct. That's my mistake, Thank you!

Re: Linode GPU Instances

#74

Post by dillonmckay » Thu, Jun 20, 2019, 2:53 PM UTC

bin0 wrote: post #69

dillonmckay wrote: post #60

Earlier quoted context omitted.

There were some compromised accounts via a Coldfusion hack of their admin portal. Not sure if that was isolated. There was something more recent, too. Anyway, happy Linode customer for quite a few years now. My stuff works, no fuss.

Any chance you can provide more information? Linode customer as well; slightly concerned.

Google ‘linode coldfusion’. I think it was over 5 years ago.

Re: Linode GPU Instances

#75

Post by m0zg » Thu, Jun 20, 2019, 6:38 PM UTC

trey-jones wrote: post #70

m0zg wrote: post #27

Still way too much money when a 2x 2080Ti comparably specced machine under my desk costs less than 2.5 months of their billing rate, and 4x 1080Ti servers in my garage cost about 1 month of their 4-GPU machine _and_ have more SSD storage. This pricing is totally insane, especially if not billed per-minute (which in Linode's case it is not) and if there are no cheaper preemptible/spot instances.

What's the cost for power? Serious question, and I'm not suggesting that this cost should account for a large percentage of the price, but genuinely curious. If your GPUs are working every hour of the month for you, how much is it costing you in electricity?

Quad GPU machines draw about 1.3KW each on average when under load. That's about $100/mo where I live, assuming they're 100% loaded 24x7 for the entire month. Realistically it's less. So it's not free, but it's not a crazy amount either.