You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@mesos.apache.org by Rodrick Brown <ro...@orchard-app.com> on 2016/10/22 13:38:42 UTC

GPU Memory or Fractional offers

Are their any plans to support GPU memory offers? 

Mesos exposes GPUs as a simple scalar resource which doesn't seem that efficient when running smaller jobs on systems with a single GPU even the ability to support fractional values would be more ideal. 

My use case is to be able to have multiple applications share the GPU at the same time. 

-- 
RB 
-- 
*NOTICE TO RECIPIENTS*: This communication is confidential and intended for 
the use of the addressee only. If you are not an intended recipient of this 
communication, please delete it immediately and notify the sender by return 
email. Unauthorized reading, dissemination, distribution or copying of this 
communication is prohibited. This communication does not constitute an 
offer to sell or a solicitation of an indication of interest to purchase 
any loan, security or any other financial product or instrument, nor is it 
an offer to sell or a solicitation of an indication of interest to purchase 
any products or services to any persons who are prohibited from receiving 
such information under applicable law. The contents of this communication 
may not be accurate or complete and are subject to change without notice. 
As such, Orchard App, Inc. (including its subsidiaries and affiliates, 
"Orchard") makes no representation regarding the accuracy or completeness 
of the information contained herein. The intended recipient is advised to 
consult its own professional advisors, including those specializing in 
legal, tax and accounting matters. Orchard does not provide legal, tax or 
accounting advice.

Re: GPU Memory or Fractional offers

Posted by Kevin Klues <kl...@gmail.com>.
As a side note. So long as you don't turn the cgroup devices isolator on,
there is nothing preventing your tasks from just accessing the GPU on the
host without actual allocation of the resource to anyone.

On Saturday, October 22, 2016, Kevin Klues <kl...@gmail.com> wrote:

> Hi rodrick. There are no plans at the moment to support either fractional
> GPU access or GPU memory allocation. The integration with the GPU drivers
> and the Linux isolation methods just isn't there yet.
>
> That said, in the upcoming 1.1 release, we have added support for 'Task
> Groups' (a.k.a. Pods). So you could make all of the Tasks that you want to
> share access to a GPU part of the same task group.
>
> On Saturday, October 22, 2016, Rodrick Brown <rodrick@orchard-app.com
> <javascript:_e(%7B%7D,'cvml','rodrick@orchard-app.com');>> wrote:
>
>> Are their any plans to support GPU memory offers?
>>
>> Mesos exposes GPUs as a simple scalar resource which doesn't seem that
>> efficient when running smaller jobs on systems with a single GPU even the
>> ability to support fractional values would be more ideal.
>>
>> My use case is to be able to have multiple applications share the GPU at
>> the same time.
>> --
>>
>> RB
>>
>> *NOTICE TO RECIPIENTS*: This communication is confidential and intended
>> for the use of the addressee only. If you are not an intended recipient of
>> this communication, please delete it immediately and notify the sender
>> by return email. Unauthorized reading, dissemination, distribution or
>> copying of this communication is prohibited. This communication does not constitute
>> an offer to sell or a solicitation of an indication of interest to purchase
>> any loan, security or any other financial product or instrument, nor is it
>> an offer to sell or a solicitation of an indication of interest to purchase
>> any products or services to any persons who are prohibited from receiving
>> such information under applicable law. The contents of this communication
>> may not be accurate or complete and are subject to change without notice.
>> As such, Orchard App, Inc. (including its subsidiaries and affiliates,
>> "Orchard") makes no representation regarding the accuracy or
>> completeness of the information contained herein. The intended recipient is
>> advised to consult its own professional advisors, including those
>> specializing in legal, tax and accounting matters. Orchard does not
>> provide legal, tax or accounting advice.
>>
>
>
> --
> ~Kevin
>


-- 
~Kevin

Re: GPU Memory or Fractional offers

Posted by Kevin Klues <kl...@gmail.com>.
Hi rodrick. There are no plans at the moment to support either fractional
GPU access or GPU memory allocation. The integration with the GPU drivers
and the Linux isolation methods just isn't there yet.

That said, in the upcoming 1.1 release, we have added support for 'Task
Groups' (a.k.a. Pods). So you could make all of the Tasks that you want to
share access to a GPU part of the same task group.

On Saturday, October 22, 2016, Rodrick Brown <ro...@orchard-app.com>
wrote:

> Are their any plans to support GPU memory offers?
>
> Mesos exposes GPUs as a simple scalar resource which doesn't seem that
> efficient when running smaller jobs on systems with a single GPU even the
> ability to support fractional values would be more ideal.
>
> My use case is to be able to have multiple applications share the GPU at
> the same time.
> --
>
> RB
>
> *NOTICE TO RECIPIENTS*: This communication is confidential and intended
> for the use of the addressee only. If you are not an intended recipient of
> this communication, please delete it immediately and notify the sender by
> return email. Unauthorized reading, dissemination, distribution or copying
> of this communication is prohibited. This communication does not constitute
> an offer to sell or a solicitation of an indication of interest to purchase
> any loan, security or any other financial product or instrument, nor is it
> an offer to sell or a solicitation of an indication of interest to purchase
> any products or services to any persons who are prohibited from receiving
> such information under applicable law. The contents of this communication
> may not be accurate or complete and are subject to change without notice.
> As such, Orchard App, Inc. (including its subsidiaries and affiliates,
> "Orchard") makes no representation regarding the accuracy or completeness
> of the information contained herein. The intended recipient is advised to
> consult its own professional advisors, including those specializing in
> legal, tax and accounting matters. Orchard does not provide legal, tax or
> accounting advice.
>


-- 
~Kevin