PGI User Forum
 SearchSearch   MemberlistMemberlist     RegisterRegister   ProfileProfile    Log inLog in 

Free OpenACC Webinar

Meaning of target compiler option cuda2.3/ cuda3.0?

 
Post new topic   Reply to topic    PGI User Forum Forum Index -> Accelerator Programming
View previous topic :: View next topic  
Author Message
xray



Joined: 21 Jan 2010
Posts: 85

PostPosted: Fri Apr 09, 2010 5:29 am    Post subject: Meaning of target compiler option cuda2.3/ cuda3.0? Reply with quote

Hello,
what do the target compiler options cuda2.3 and cuda3.0 do?
I tried to find it out but couldn't find anything except: "Use CUDA 2.3/3.0 Toolkit compatibility"...

I changed my Fortran PGI Accelerator compilation from
pgf90 -g -ta=nvidia,cc11 -Minfo -fastsse -c ./jacobi.F90 -o jacobi.o
to
pgf90 -g -ta=nvidia,cc11,3.0 -Minfo -fastsse -c ./jacobi.F90 -o jacobi.o
and suddenly my performance increased from 3300 MFlops to 4600 MFlops!

So, is cuda2.3 the default option? And what happens internally when switching to 3.0?

Bye, Sandra
Back to top
View user's profile
mkcolg



Joined: 30 Jun 2004
Posts: 6208
Location: The Portland Group Inc.

PostPosted: Fri Apr 09, 2010 9:23 am    Post subject: Reply with quote

Hi Sandra,

Quote:
So, is cuda2.3 the default option?

Yes. The default is to use the CUDA 2.3 toolkit.
Quote:

And what happens internally when switching to 3.0?
The PGI compiler will use the NVIDIA CUDA 3.0 toolkit. CUDA 3.0 was just released and is new in PGI 10.4. For Fermi, you'll need to use CUDA 3.0.

Note that you will need to have a NVIDIA driver that supports CUDA 3.0. If you are unsure which version you have, the PGI utility "pgaccelinfo" will show your version number.

Hope this helps,
Mat

Code:
% pgaccelinfo
CUDA Driver Version            3000   

Device Number:                 0
Device Name:                   Tesla T10 Processor
Device Revision Number:        1.3
Global Memory Size:            4294770688
Number of Multiprocessors:     30
Number of Cores:               240
Concurrent Copy and Execution: Yes
Total Constant Memory:         65536
Total Shared Memory per Block: 16384
Registers per Block:           16384
Warp Size:                     32
Maximum Threads per Block:     512
Maximum Block Dimensions:      512, 512, 64
Maximum Grid Dimensions:       65535 x 65535 x 1
Maximum Memory Pitch:          2147483647B
Texture Alignment              256B
Clock Rate:                    1296 MHz
Initialization time:           4424750 microseconds
Current free memory            4256304896
Upload time (4MB)               998 microseconds ( 796 ms pinned)
Download time                  1194 microseconds ( 769 ms pinned)
Upload bandwidth               4202 MB/sec (5269 MB/sec pinned)
Download bandwidth             3512 MB/sec (5454 MB/sec pinned)

Back to top
View user's profile
xray



Joined: 21 Jan 2010
Posts: 85

PostPosted: Mon Apr 12, 2010 12:34 am    Post subject: Reply with quote

Thanks for your reply!
Back to top
View user's profile
Display posts from previous:   
Post new topic   Reply to topic    PGI User Forum Forum Index -> Accelerator Programming All times are GMT - 7 Hours
Page 1 of 1

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum


Powered by phpBB © phpBB Group