Re: [Starpu-devel] Combined workers and OpenMP code

[Keisuke Fukuda <fukuda@matsulab.is.titech.ac.jp>]

Thank you for your reply, Samuel.

> I notice that your program uses a regression-based performance model,
> but it only submits tasks of the same size.  The scheduler does not
> realize that it's still a task of the same size, which will not help the
> performance model, and thus continues to try with 1, in the hope the
> performance model will eventually provide estimations.

I changed my code to use various array size and it started to use
multiple cores. Great.
https://gist.github.com/2488259

But mine and vector_scal still use only upto 4 cores. I investigated a
little and found that
StarPU only recognized 4 cores.

Here's a result from starpu_machine_display :
$ starpu_machine_display
StarPU has found :
        4 CPU cores
                CPU 0
                CPU 1
                CPU 2
                CPU 3
        2 CUDA devices
                CUDA 0 (Tesla C2050 2.6 GiB 01:00.0)
                CUDA 1 (Tesla C2050 2.6 GiB 02:00.0)
        0 OpenCL devices
(snip)

I checked "lstopo" command and made sure that hwloc recognizes 6 cores.
Why StarPU doesn't recognize the other 2 cores?
Is it a compile-time prblem or environmental configuration problem?


> You can use the execute_on_a_specific_worker field to force using a
> combined worker.  Nicolas, is there any way to print the combined
> workers tree?  Also, it seems that STARPU_MIN/MAX_WORKERSIZE are not
> documented at all, could you handle that?

I've not yet tried these but they must be useful !


Regards.
Keisuke

-- 
------------------------------------------
FUKUDA, Keisuke<fukuda@matsulab.is.titech.ac.jp>
Dept. of Math. & Comp. Sciences
Satoshi Matsuoka Lab.,
Tokyo Institute of Technology