Hi George,
Some options would be one one of the cloud providers. The obvious is AWS but that can be expensive and their hardware (in the case of GPUs) tends to be out of date and slow. You could try Crunchyard (www.crunchyard.com <
http://www.crunchyard.com/>) or CirraScale (www.cirrascale.com <
http://www.cirrascale.com/>) both of which offer bare metal access on systems with much more modern GPUs. E.g. Titan-XP in the case of Cirrascale which can make them much more cost effective than AWS or GoogleCloud for example. You might even be able to get a free trial from one of them that provides sufficient time for what you need.
Another option would be one of the NSF supercomputing centers. They offer allocations through a quarterly proposal process but one can typically get a few hundred thousand core hours free for development purposed with just an abstract. You need to be affiliated with a US education or research institution though.
That said what you suggest here doesn't sound like it would need a lot of compute and might be easy to do on a small GPU desktop. Can you provide more details of the calculations you want to run and then we can provide some better suggestions. Note for ~$6500 you could get a 4 GPU workstation (see
http://ambermd.org/gpus/recommended_hardware.htm#hardware <
http://ambermd.org/gpus/recommended_hardware.htm#hardware>) that could probably munch through what you need to run in less than a week (assuming what you want to run will work on GPUs).
Hope that helps.
All the best
Ross
> On Jul 20, 2017, at 11:38, George Tzotzos <gtzotzos.me.com> wrote:
>
> Hi everybody,
>
> I’ve submitted a paper to PLOS One. I’ve received mostly positive comments but I have to run 4-5 additional 100ns simulations on a small system (125 residues). I’m afraid my computer power is rather limited, so is the time for submitting the final version of the paper.
>
> Is anyone aware of any possible options for “buying” computer time and use a machine remotely?
>
> Many thanks in advance for any suggestions
>
> George
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Jul 20 2017 - 09:00:06 PDT
