[Omp] why cannot use openmp

[Shah, Sanjiv]

Hello,

An old friend, John Larson, reminds me that Double Precision on Cray
used to be 128 bit floating point.  I am not sure what it is today.  The
rest of the world means 64 bits when using the term double precision.

Could you please clarify whether you are talking about 128 bits or 64
bits?

If it is 128 bits, the term used by the rest of the world is Quad
precision.  This is usually done via software emulation.  Quad precision
in multi-processor mode could fail if the software emulation layer is
not thread safe.

Thanks,
Sanjiv

--
Sanjiv, 217-403-4244

-----Original Message-----
From: Omp-bounces at openmp.org [mailto:Omp-bounces at openmp.org] On Behalf
Of Shah, Sanjiv
Sent: Tuesday, March 07, 2006 10:33 PM
To: lizhou; Omp at openmp.org
Subject: Re: [Omp] why cannot use openmp

There's no inherent reason why double precision on Linux with multiple
CPU's shouldn't work.  Most likely it is related to either your
application, your compiler or a combination of both.

Have you tried to create a small test case to identify what fails?

Sanjiv

--
Sanjiv, 217-403-4244

-----Original Message-----
From: Omp-bounces at openmp.org [mailto:Omp-bounces at openmp.org] On Behalf
Of lizhou
Sent: Tuesday, March 07, 2006 8:43 PM
To: Omp at openmp.org
Subject: [Omp] why cannot use openmp

Hi,

The model is double precision at Unix operation system in Cray. Now it
is
transferred to Linux operation system. Does anyone know why the model
cannot
use double precision in Linux with multiple CPU? but it is ok just using
single CPU at Linux or just use single precision with multiple cup. Any
suggestion are very appreciated.

Best wishes,
ZHOU, Li


_______________________________________________
Omp mailing list
Omp at openmp.org
http://openmp.org/mailman/listinfo/omp_openmp.org

_______________________________________________
Omp mailing list
Omp at openmp.org
http://openmp.org/mailman/listinfo/omp_openmp.org