Which universities study CUDA?

Question

linux

Hi all! I recently completed my MEPhI undergraduate degree in condensed matter physics. At the moment, there is an opportunity to leave to work not in the specialty in analytics, where they offer starting 60 tyr, or continue their studies in a master's program. I have always been interested in the topic of parallel programming, and I wanted to develop in it. But, as it turned out, the luminaries of Russian science do not need any parallel programming on graphics cores, so I was sent away. Therefore, the question arose whether there are universities and projects in Russia where you can go to study parallel programming on a graph. cores? Or just shut up and work?

Answer 1

But, as it turned out, the luminaries of Russian science do not need any parallel programming on graphics cores.

What for do they need it? They need solutions to problems and problems. So find a problem, the solution of which fits well into the area that interests you - go there.
Do not confuse universities and circles / technical schools / courses

Answer 2

In Russia, training courses on CUDA are given in universities:

1. Moscow Institute of Physics and Technology
   
2. St. Petersburg State Polytechnic University
   
3. High School of Economics
   
4. Lomonosov Moscow State University
   
5. Saint Petersburg State University
   
6. Joint Institute for Nuclear Research
   
7. Moscow Institute of Electronic Technology
   
8. MSTU im. Bauman
   

Answer 3

I have always been interested in the topic of parallel programming.

This doesn't work here. You either know how and have spent years on it, or you don't know how. Wonabies aren't really needed. PM-PU and MIPT have it. Prepare matan for introductory.

Answer 4

When I studied at the SPbSPU at the Department of IMS, one of the research topics that was proposed by one research instructor was the research and development of tools that simplify (or automate) the parallelization of algorithms on CUDA.
I did not dare to do this because it is too difficult

Answer 5

St. Petersburg State University has a resource center with a cluster of tesla, at the physics department for half a year they give introductory information about cuda, employees can drive solutions to specific problems on the cluster. First of all, you need to decide on the tasks that you can and want to solve, and then use such a highly specialized technology.

Answer 6

I know for sure that around 2008, CUDA (and parallel computing in general) were promising areas at Novosibirsk State University (Novosibirsk).
I advise you to find out how things are now there.

Answer 7

Apparently, almost everywhere, and we have particulars: SPbPU, Department of Applied Mathematics.

Answer 8

you can not score but also not dive headlong into something in which you still understand little.
most areas of activity are not as interesting as it seems from the outside. maybe you will quickly dislike this topic after six months of training.
in my opinion, the most adequate thing is:
1) to develop in the direction of what is interesting (comprehensively study the technology itself and all related things).
in the sense of independently studying in your free time or in parallel with work
2) bring work closer to this direction (that is, do not look for only one on which only the technology of interest is used, but something that is close in spirit and principles to what is interesting. Usually there are a lot of analogies in similar directions. And you might be working not with CUDA, but with other parallel computing - there will definitely be common features)
3) with great practical and theoretical experience / knowledge, you will find either an "ideal job" or a scientific association involved in this (most likely foreign, since in the Russian Federation science moves very slowly and cheaply )
and yet to search for people (namely, this should be in priority, and not some universities and institutes), those involved in this should look at top conferences and magazines on topics of interest and read who the author is and where he works. and reach out to specific people (by mail or at the conferences themselves).
almost always, specific topics are moved by specific people, and the place where this happens (university, institute, foreign country, conferences) is secondary.