I don't think running hadoop on a GPU cluster is a common use case; the
types of workloads for a hadoop vs. gpu cluster are very different although
a quick google search did turn up some. So this is probably not the best
mailing list for your question.

J
On Thu, Jan 17, 2013 at 5:18 AM, Roberto Nunnari
<[EMAIL PROTECTED]>wrote:

> Roberto Nunnari wrote:
>
>> Hi all.
>>
>> I'm writing to you to ask for advice or a hint to the right direction.
>>
>> In our department, more and more researchers ask us (IT administrators)
>> to assemble (or to buy) GPGPU powered workstations to do parallel computing.
>>
>> As I already manage a small CPU cluster (resources managed using SGE),
>> with my boss we talked about building a new GPU cluster. The problem is
>> that I have no experience at all with GPU clusters.
>>
>> Apart from the already running GPU workstations, we already have some new
>> HW that looks promising to me as a starting point for a GPU cluster.
>>
>> - 1x Dell PowerEdge R720
>> - 1x Dell PowerEdge C410x
>> - 1x NVIDIA M2090 PCIe x16
>> - 1x NVIDIA iPASS Cable Kit
>> (Dell forgot to include the iPASS adapter for the R720!! :-D)
>>
>> I'd be grateful if you could kindly give me some advice and/or hint to
>> the right direction.
>>
>> In particular I'm interested on your opinion on:
>> 1) is the above HW suitable for a small (2 to 4/6 GPUs) GPU cluster?
>> 2) is apache adhoop suitable (or what could we use?) as a queuing and
>> resource management system? We would like the cluster to be usable by many
>> users at once in a way that no user has to worry about resources, just like
>> we do on the CPU cluster with SGE.
>> 3) What distribution of linux would be more appropriate?
>> 4) necessary stack of sw? (cuda, hadoop, other?)
>>
>> Thank you very much for your valuable insight!
>>
>> Best regards.
>> Robi
>>
>
> Anybody on this, please?
> Robi
>
>

Thiago Vieira wrote:
> I've seen some academic researches on this direction, with good results.
> Some computations can be expressed by GPGPU, but it is still a restrict
> number of cases. If is not easy to solve problems using MapReduce, solve
> some problems with SIMD is harder.

Ok.. Thank you all for your time.. I'll keep searching.
Best regards.
Robi
>
> --
> Thiago Vieira
>
>
> On Thu, Jan 17, 2013 at 9:24 PM, Russell Jurney
> <[EMAIL PROTECTED] <mailto:[EMAIL PROTECTED]>> wrote:
>
>     Hadoop streaming can do this, and there's been some discussion in
>     the past, but it's not a core use case. Check the list archives.
>
>     Russell Jurney http://datasyndrome.com
>
>     On Jan 17, 2013, at 9:25 AM, Jeremy Lewi <[EMAIL PROTECTED]
>     <mailto:[EMAIL PROTECTED]>> wrote:
>
>>     I don't think running hadoop on a GPU cluster is a common use
>>     case; the types of workloads for a hadoop vs. gpu cluster are very
>>     different although a quick google search did turn up some. So this
>>     is probably not the best mailing list for your question.
>>
>>     J
>>
>>
>>     On Thu, Jan 17, 2013 at 5:18 AM, Roberto Nunnari
>>     <[EMAIL PROTECTED] <mailto:[EMAIL PROTECTED]>> wrote:
>>
>>         Roberto Nunnari wrote:
>>
>>             Hi all.
>>
>>             I'm writing to you to ask for advice or a hint to the
>>             right direction.
>>
>>             In our department, more and more researchers ask us (IT
>>             administrators) to assemble (or to buy) GPGPU powered
>>             workstations to do parallel computing.
>>
>>             As I already manage a small CPU cluster (resources managed
>>             using SGE), with my boss we talked about building a new
>>             GPU cluster. The problem is that I have no experience at
>>             all with GPU clusters.
>>
>>             Apart from the already running GPU workstations, we
>>             already have some new HW that looks promising to me as a
>>             starting point for a GPU cluster.
>>
>>             - 1x Dell PowerEdge R720
>>             - 1x Dell PowerEdge C410x
>>             - 1x NVIDIA M2090 PCIe x16
>>             - 1x NVIDIA iPASS Cable Kit
>>             (Dell forgot to include the iPASS adapter for the R720!! :-D)
>>
>>             I'd be grateful if you could kindly give me some advice
>>             and/or hint to the right direction.
>>
>>             In particular I'm interested on your opinion on:
>>             1) is the above HW suitable for a small (2 to 4/6 GPUs)
>>             GPU cluster?
>>             2) is apache adhoop suitable (or what could we use?) as a
>>             queuing and resource management system? We would like the
>>             cluster to be usable by many users at once in a way that
>>             no user has to worry about resources, just like we do on
>>             the CPU cluster with SGE.
>>             3) What distribution of linux would be more appropriate?
>>             4) necessary stack of sw? (cuda, hadoop, other?)
>>
>>             Thank you very much for your valuable insight!
>>
>>             Best regards.
>>             Robi
>>
>>
>>         Anybody on this, please?
>>         Robi