Re: [Gomp-discuss] OpenMP: Examples 1

[Lars Segerlund]

 Scott, you are my hero :-) ...  don't give up !

 Excellent, stuff like this is what we need.

 I can do some translating of the code, while keeping a keen eye on the intel 
compiler ( it cheats :-) ). 

 I think youre on to the main thing here, about breaking things down into parts 
which we have to get done.
 It is always best to first go for the 'low hanging fruit'.

 A question though, does it sound reasonable that we should go for support for 
both threads & MPI ? This would mainly concearn the lib(s), and perhaps we have 
to stay out of trouble during the implementation.

 / Lars Segerlund.


On Thu, 26 Aug 2004 13:28:41 -0400
Scott Robert Ladd <address@hidden> wrote:

> Okay,
> 
> Attached hereto are the most basic examples: A "Hello, world" program in 
> OpenMP, and as implement with PThreads and Unix-style processes. I also 
> have an MPI and Windows Processes version, but those are on a machine 
> that is busy at the moment.
> 
> Yes, this is incredible simple -- which is my point. We need to figure 
> out what the baseline is for doing even the simplest program. What is 
> common between different parallelization models?
> 
> Later tonight, I'll pull a more advanced example (a basic loop) from CVS.
> 
> My goal was pretty simple: To see precisely what sort of code we would 
> expect under different conditions, so we could identify the common 
> features for constructing a universal model.
> 
> I largely gave up on GOMP earlier this year, due to a lack of interest 
> by others.
> 
> ..Scott
> 
> -- 
> Scott Robert Ladd
> Coyote Gulch Productions (http://www.coyotegulch.com)
> Software Invention for High-Performance Computing
>