[gentoo-soc] GSoC - 2019

[gentoo-soc] GSoC - 2019

[Leonardo Romor]

[-- Attachment #1: Type: text/plain, Size: 581 bytes --]

Hello to everyone,

I would like to propose a project for the Google summer of code 2019.
My idea is to deliver an official packaging for the docker-nvidia runtime.
Would that be a feasible and an interesting project idea?
I have some experience in high performance computing and optimization
techniques, I would also be happy to work on such projects especially if
related
on packaging BLAS implementations or adding a USE flag for the intel
patches for tensorflow.
Which one of the two/three proposed ideas would better fit the GSoC?

Thank you and best regards,

Leonardo Romor

[-- Attachment #2: Type: text/html, Size: 713 bytes --]

Re: [gentoo-soc] GSoC - 2019

[Andrew Savchenko]

[-- Attachment #1: Type: text/plain, Size: 1393 bytes --]

Hi!

On Sat, 6 Apr 2019 17:45:23 +0200 Leonardo Romor wrote:
> Hello to everyone,
> 
> I would like to propose a project for the Google summer of code 2019.
> My idea is to deliver an official packaging for the docker-nvidia runtime.
> Would that be a feasible and an interesting project idea?

It looks interesting. However adding a single package looks too
simple for 3 months worth of work.

> I have some experience in high performance computing and optimization
> techniques, I would also be happy to work on such projects especially if
> related
> on packaging BLAS implementations

We have a related project idea:
https://wiki.gentoo.org/wiki/Google_Summer_of_Code/2019/Ideas#BLAS_and_LAPACK_runtime_switching

> or adding a USE flag for the intel patches for tensorflow.

It is OK to have this. But such task alone is too simple for the
project.

> Which one of the two/three proposed ideas would better fit the GSoC?

Any idea will fit nice as long as:
1) submitted proposal is strong and detailed;
2) you have shown us your skills by fixing some bugs from Gentoo
bugzilla;
3) proposed idea should be complex enough for 3 months worth of
work, though it should be doable at the same time;
4) for your own idea it is advised to look for an interested mentor.

> Thank you and best regards,
> 
> Leonardo Romor


Best regards,
Andrew Savchenko

[-- Attachment #2: Type: application/pgp-signature, Size: 833 bytes --]