I want to work on NVIDIA GPU cluster for my thesis. But the cost of clusters are too high for me to bear. Although I can buy a couple of latest GPU cards, but buying cluster is impossible for me at least right now. What is the alternative? Are there any Non profit organization promoting and renting clusters for educational purpose at subsidized pay per use rate?
3
-
$\begingroup$ For general purpose supercomputing, it's usually possible to apply for accounts on various machines and get small allocations to start (say, 10,000 processor-hours). The general idea is to use those hours to get familiar with the machine, generate some results, and then take those results and write a proposal to get a larger allocation there, or on another supercomputer. Presumably, a similar process might exist for GPU clusters. You also might be able to ask around at your university and see if someone in your professional network would be willing to give you an account on their machine. $\endgroup$– Geoff OxberryCommented Sep 17, 2013 at 7:19
-
1$\begingroup$ Where do you work ? Country and institution ? There's no point anyone providing, say, information applicable only in PRC to someone in USA. $\endgroup$– High Performance MarkCommented Sep 17, 2013 at 12:20
-
$\begingroup$ @GeoffOxberry Can you provide links to where I can apply for an account with a small allocation for general purpose supercomputing? Thanks. $\endgroup$– a06eCommented Oct 4, 2014 at 17:37
Add a comment
|
2 Answers
$\begingroup$
$\endgroup$
If you work at an institution in the US or have collaborators at one, you should look into the XSEDE program. This is a federation of the National Science Foundation funded supercomputing centers in the US which provides compute cycles on a variety of large-scale machines to open science research in the US. There are a couple of machines with GPU resources including Keeneland and Stampede.
Startup allocations are readily available to academic researchers and can be applied for on the XSEDE User Portal. As a student researcher, you will need to work with your academic/thesis advisor to get them to apply for time. If your application for an allocation of time is successful, the time itself is free.
answered Sep 17, 2013 at 12:01
$\begingroup$
$\endgroup$
A final alternative is to rent (or apply for time on) a GPU cluster from Amazon EC2. As Bill and others have mentioned, most countries provide some national or regional high performance computing resources for free to their scientists. If you don't have one in your area, it is mostly a matter of finding a collaborator who does.
answered Sep 17, 2013 at 12:31