GPUs and CUDA

GPUs and CUDA

[Rego, P.]

Does Ada have any compiler with support to development under GPUs? CUDA looks to be very promissing and development in Ada on it could be useful.

Re: GPUs and CUDA

[anon]

It seams that Graphics is not a major subsystem for the Ada maintainers.
Maybe that will change for for now the answers are No and No!

You might find some Graphic links in the GNAT Ada for Java. But for 
the most part Ada is Text only, with a few third parties binding 
packages for SDL and openGL graphic engines.

Plus, at this time Ada's RTS is still "Concurrent", so, no to CUDA as well.


In <09ad3bbb-b0a3-438b-9263-b2cb49098e5c@glegroupsg2000goo.googlegroups.com>, "Rego, P." <pvrego@gmail.com> writes:
>Does Ada have any compiler with support to development under GPUs? CUDA looks to be very promissing and development in Ada on it could be useful.

Re: GPUs and CUDA

[Anatoly Chernyshev]

On Jul 9, 3:18 am, a...@att.net wrote:
> It seams that Graphics is not a major subsystem for the Ada maintainers.
> Maybe that will change for for now the answers are No and No!
>
> You might find some Graphic links in the GNAT Ada for Java. But for
> the most part Ada is Text only, with a few third parties binding
> packages for SDL and openGL graphic engines.
>

Well, the most attractive application of CUDA is not the graphics, but
parallel computing using modern graphic chips. The performance gain
for the parallelizable programs, which can use CUDA, is pretty
significant on a PC. I was looking into using it myself, and eagerly
waiting when somebody comes up with the Ada binding.

Re: GPUs and CUDA

[Georg Bauhaus]

On 7/11/11 7:45 PM, Anatoly Chernyshev wrote:

> Well, the most attractive application of CUDA is not the graphics, but
> parallel computing using modern graphic chips. The performance gain
> for the parallelizable programs, which can use CUDA, is pretty
> significant on a PC. I was looking into using it myself, and eagerly
> waiting when somebody comes up with the Ada binding.

Wouldn't a multicore PC with additional Graphics processors be
a candidate for an Ada 2012 Ravenscar multicore runtime?

Re: GPUs and CUDA

[Anatoly Chernyshev]

On Jul 11, 10:38 pm, Georg Bauhaus <rm.dash-bauh...@futureapps.de>
wrote:
> Wouldn't a multicore PC with additional Graphics processors be
> a candidate for an Ada 2012 Ravenscar multicore runtime?

I'm not sure about that, taking into account that Ada's original
purpose is embedded systems, which are often missing GPUs whatsoever.
Yet, I believe, GPU's cores are not quite the same as CPU's, even in
terms of synchronization and communication.

>Err... I think "binding" is perhaps a horrid idea. (Raise your hand if
>you just want a light wrapper around C-headers.)
>The way to go, IMO, is to allow a specialized pragma to indicate to
>the compiler that such and such subprogram is intended for
>distribution to the GPUs for parallel work.

>This would allow something like
> Pragma CUDA( Ralph );

I guess, in this particular case, the binding or port is better. The
suggested pragma makes the language (or at least compiler) hardware-
dependend, which I, personally, would hate to see. Yet, the IT's
hardware sector is much more agile than Ada's standards, so that today
we have CUDA, and tomorrow it will be replaced by who-knows-what...

Re: GPUs and CUDA

[Shark8]

On Jul 12, 12:59 am, Anatoly Chernyshev <achernys...@gmail.com> wrote:
> On Jul 11, 10:38 pm, Georg Bauhaus <rm.dash-bauh...@futureapps.de>
> wrote:
>
> > Wouldn't a multicore PC with additional Graphics processors be
> > a candidate for an Ada 2012 Ravenscar multicore runtime?
>
> I'm not sure about that, taking into account that Ada's original
> purpose is embedded systems, which are often missing GPUs whatsoever.
> Yet, I believe, GPU's cores are not quite the same as CPU's, even in
> terms of synchronization and communication.
>
> >Err... I think "binding" is perhaps a horrid idea. (Raise your hand if
> >you just want a light wrapper around C-headers.)
> >The way to go, IMO, is to allow a specialized pragma to indicate to
> >the compiler that such and such subprogram is intended for
> >distribution to the GPUs for parallel work.
> >This would allow something like
> > Pragma CUDA( Ralph );
>
> I guess, in this particular case, the binding or port is better. The
> suggested pragma makes the language (or at least compiler) hardware-
> dependend, which I, personally, would hate to see.

Yet that is exactly what implementation-defined pragmas are supposed
to allow.
And it needn't force the program to be implementation-dependent;
consider:

Task Type Some_Task [...];

Procedure X is
  Tasks : Array (1..20) of Some_Task;
  Pragma Multiprocessor(Construct => Tasks, Processors => 4, Cores =>
4);
begin
Null; -- We're finished with the procedure when all tasks are
finished.
end X;

A compiler which is written for a single-processor computer is free to
ignore the pragma Multiprocessor...
and yet this will still work. (Though maybe not as-typed, I didn't try
to compiler it, it might need a level of indirection.)

In the particular case of CUDA I've always thought it would have been
better for them to specialize an Ada compiler rather than a C compiler
simply because the TASK construct is very well-suited to such
parallelisms; they could have had some error/complexity-checking which
would cause the compiler to error-out should the TASKs tagged with the
CUDA held things which would be impossible or onerous to implement.

> Yet, the IT's
> hardware sector is much more agile than Ada's standards, so that today
> we have CUDA, and tomorrow it will be replaced by who-knows-what...

True.

Re: GPUs and CUDA

[anon]

Under the CUDA system the parallel code is written in C and can be 
called by C, JAVA, Ruby, etc.  And can include Ada with use of 
"pragma Import" statement.  But why use another hosting language, because 
it is only required to load and start execution of the parallel code, and 
to handle the "Windows" or "X-Windows" environment for the program. Ada 
has third party packages for some windows environment, but most 
programming for Ada do not use them.

For Learning parallel, in my opinion I would use the CUDA C compiler 
to create both parallel code and the hosting code. For the hosting 
code, just use an sample code and alter it to fit you needs.

But the main problem in using CUDA is the memory is to small for 
the parallel design. It mostly setup for graphics.


In <4e1b4333$0$6583$9b4e6d93@newsspool3.arcor-online.net>, Georg Bauhaus <rm.dash-bauhaus@futureapps.de> writes:
>On 7/11/11 7:45 PM, Anatoly Chernyshev wrote:
>
>> Well, the most attractive application of CUDA is not the graphics, but
>> parallel computing using modern graphic chips. The performance gain
>> for the parallelizable programs, which can use CUDA, is pretty
>> significant on a PC. I was looking into using it myself, and eagerly
>> waiting when somebody comes up with the Ada binding.
>
>Wouldn't a multicore PC with additional Graphics processors be
>a candidate for an Ada 2012 Ravenscar multicore runtime?
>

Re: GPUs and CUDA

[Shark8]

On Jul 11, 12:45 pm, Anatoly Chernyshev <achernys...@gmail.com> wrote:
> On Jul 9, 3:18 am, a...@att.net wrote:
>
> > It seams that Graphics is not a major subsystem for the Ada maintainers.
> > Maybe that will change for for now the answers are No and No!
>
> > You might find some Graphic links in the GNAT Ada for Java. But for
> > the most part Ada is Text only, with a few third parties binding
> > packages for SDL and openGL graphic engines.
>
> Well, the most attractive application of CUDA is not the graphics, but
> parallel computing using modern graphic chips. The performance gain
> for the parallelizable programs, which can use CUDA, is pretty
> significant on a PC. I was looking into using it myself, and eagerly
> waiting when somebody comes up with the Ada binding.

Err... I think "binding" is perhaps a horrid idea. (Raise your hand if
you just want a light wrapper around C-headers.)
The way to go, IMO, is to allow a specialized pragma to indicate to
the compiler that such and such subprogram is intended for
distribution to the GPUs for parallel work.

As per http://archive.adaic.com/standards/83lrm/html/lrm-02-08.html#2.8
Named associations are, however, only possible if the argument
identifiers are defined. A name given in an argument must be either a
name visible at the place of the pragma or an identifier specific to
the pragma. [...] A pragma that is not language-defined has no effect
if its identifier is not recognized by the (current) implementation.
Furthermore, a pragma (whether language-defined or implementation-
defined) has no effect if its placement or its arguments do not
correspond to what is allowed for the pragma. The region of text over
which a pragma has an effect depends on the pragma.

This would allow something like
 Pragma CUDA( Ralph );
to indicate that the subprogram "Ralph" is to be run on the GPUs; and
it would also allow non-CUDA enabled compilers to compile the source
regardless of whether or not their compiler actually knows about CUDA.

Re: GPUs and CUDA

[Rego, P.]

>This would allow something like
> Pragma CUDA( Ralph );
>to indicate that the subprogram "Ralph" is to be run on the GPUs; and
>it would also allow non-CUDA enabled compilers to compile the source
>regardless of whether or not their compiler actually knows about CUDA.

That's the idea. It would be very good.

Re: GPUs and CUDA

[ldries46]

|"Rego, P."  schreef in bericht 
news:783efc18-a581-4ae2-bde4-04afa3568ff8@glegroupsg2000goo.googlegroups.com...
|
|>This would allow something like
|> Pragma CUDA( Ralph );
|>to indicate that the subprogram "Ralph" is to be run on the GPUs; and
|>it would also allow non-CUDA enabled compilers to compile the source
|>regardless of whether or not their compiler actually knows about CUDA.
|
|That's the idea. It would be very good.

At this moment I don't know if this is a good idea, because CUDA is a 
language developped only for NVidia GPU's. One of the purposes of Ada I 
think is that a program can always be used on another computer with only the 
a simple recompilation. I would like a Pragma GPU( Ralph ) that can check 
which GPU is used and if that GPU has a CUDA like environment available.

L. Dries 

Re: GPUs and CUDA

[Gene]

On Jul 12, 6:26 am, "ldries46" <bertus.dr...@planet.nl> wrote:
> |"Rego, P."  schreef in berichtnews:783efc18-a581-4ae2-bde4-04afa3568ff8@glegroupsg2000goo.googlegroups.com...
> |
> |>This would allow something like
> |> Pragma CUDA( Ralph );
> |>to indicate that the subprogram "Ralph" is to be run on the GPUs; and
> |>it would also allow non-CUDA enabled compilers to compile the source
> |>regardless of whether or not their compiler actually knows about CUDA.
> |
> |That's the idea. It would be very good.
>
> At this moment I don't know if this is a good idea, because CUDA is a
> language developped only for NVidia GPU's. One of the purposes of Ada I
> think is that a program can always be used on another computer with only the
> a simple recompilation. I would like a Pragma GPU( Ralph ) that can check
> which GPU is used and if that GPU has a CUDA like environment available.
>
> L. Dries

Indeed, GPUs are highly specialized processors that excel mostly at
certain kinds of floating point computations with limited branching
behaviors.  Compiler tricks to find good mappings for general code
onto GPUs are very much a current topic of research.  Nvidia's CUDA
implementation relies on language extensions to C and Fortran buit
into GCC.  I don't think it would be useful or even possible to map
GPU-style threads into Ada tasks.  So getting access to CUDA
primitives from Ada looks like a tough problem, certainly more than
adding a Pragma and tweaking the GNAT code generator.

On the other hand, there are many numerical libraries already
rewritten and tuned to exploit CUDA.  Ada bindings for these would be
very useful and straingthforwrd to build.

Re: GPUs and CUDA

[Shark8]

On Jul 12, 5:26 am, "ldries46" <bertus.dr...@planet.nl> wrote:
> |"Rego, P."  schreef in berichtnews:783efc18-a581-4ae2-bde4-04afa3568ff8@glegroupsg2000goo.googlegroups.com...
> |
> |>This would allow something like
> |> Pragma CUDA( Ralph );
> |>to indicate that the subprogram "Ralph" is to be run on the GPUs; and
> |>it would also allow non-CUDA enabled compilers to compile the source
> |>regardless of whether or not their compiler actually knows about CUDA.
> |
> |That's the idea. It would be very good.
>
> At this moment I don't know if this is a good idea, because CUDA is a
> language developped only for NVidia GPU's. One of the purposes of Ada I
> think is that a program can always be used on another computer with only the
> a simple recompilation. I would like a Pragma GPU( Ralph ) that can check
> which GPU is used and if that GPU has a CUDA like environment available.
>
> L. Dries

True; I was thinking more on the side of "what NVidia should have
done" in making their compiler.
The CUDA compiler is, after all, a a C-compiler with a bunch of
[nasty] add-ons.

Re: GPUs and CUDA

[ldries46]

"Shark8"  schreef in bericht 
news:98b32073-e1ef-4e97-80f8-ce068fba2738@h12g2000vbx.googlegroups.com...

On Jul 12, 5:26 am, "ldries46" <bertus.dr...@planet.nl> wrote:
> |"Rego, P."  schreef in 
> berichtnews:783efc18-a581-4ae2-bde4-04afa3568ff8@glegroupsg2000goo.googlegroups.com...
> |
> |>This would allow something like
> |> Pragma CUDA( Ralph );
> |>to indicate that the subprogram "Ralph" is to be run on the GPUs; and
> |>it would also allow non-CUDA enabled compilers to compile the source
> |>regardless of whether or not their compiler actually knows about CUDA.
> |
> |That's the idea. It would be very good.
>
> At this moment I don't know if this is a good idea, because CUDA is a
> language developped only for NVidia GPU's. One of the purposes of Ada I
> think is that a program can always be used on another computer with only 
> the
> a simple recompilation. I would like a Pragma GPU( Ralph ) that can check
> which GPU is used and if that GPU has a CUDA like environment available.
>
> L. Dries

|done" in making their compiler.
|The CUDA compiler is, after all, a a C-compiler with a bunch of
|[nasty] add-ons.

Perhaps this discussions must be continued on another forum. The best way to 
import GPU calculations in a regular program would be lijk graphics with an 
OpenGPU library to which you can link from any language.

:. Dries 

Re: GPUs and CUDA

[Rego, P.]

Yes, I was looking to packages to implement paralell computation for finite elements calculations for very big sparse matrices (something like 300x300x300). C has several implementations for CUDA, but surely it lacks real-time Ada facilities, mainly in access and control of multi-process GPU constraints. Guess that it could be very useful to have implemented in Ada (and even better native implementation).

Re: GPUs and CUDA

[Nicholas Collin Paul de Glouceſter]

On July 8th, 2011, anon@ATT.net sent:
|---------------------------------------------------------------------------|
|"It seams that Graphics is not a major subsystem for the Ada maintainers.  |
|Maybe that will change for for now the answers are No and No!              |
|                                                                           |
|You might find some Graphic links in the GNAT Ada for Java. But for        |
|the most part Ada is Text only, with a few third parties binding           |
|packages for SDL and openGL graphic engines.                               |
|                                                                           |
|Plus, at this time Ada's RTS is still "Concurrent", so, no to CUDA as well.|
|                                                                           |
|                                                                           |
|In                                                                         |
|<09ad3bbb-b0a3-438b-9263-b2cb49098e5c@glegroupsg2000goo.googlegroups.com>, |
|"Rego, P." <pvrego@gmail.com> writes:                                      |
|>Does Ada have any compiler with support to development under GPUs?        |
|CUDA looks to be very promissing and development in Ada on it could be     |
|useful."                                                                   |
|---------------------------------------------------------------------------|



Concurrent Computer Corporation did not have a binding for its Ada95
compiler to its CUDA hardware when I asked in July 2010 (almost a year
ago).