+- +-
Say hello if visiting :) by Gecko
11 Jan 2023, 07:43:05 pm

Seti is down again by Mike
09 Aug 2017, 10:02:44 am

Some considerations regarding OpenCL MultiBeam app tuning from algorithm view by Raistmer
11 Dec 2016, 06:30:56 am

Loading APU to the limit: performance considerations by Mike
05 Nov 2016, 06:49:26 am

Better sleep on Windows - new round by Raistmer
26 Aug 2016, 02:02:31 pm

Author Topic: Fermi compatibility for linux cuda client  (Read 8151 times)

Offline sunu

  • Alpha Tester
  • Knight who says 'Ni!'
  • ***
  • Posts: 771
Fermi compatibility for linux cuda client
« on: 21 Mar 2010, 05:11:20 pm »
As the fermi release nears, just a heads up for those that will get the new cards.

Quoting nvidia's "Fermi Compatibility Guide" available from http://developer.nvidia.com/object/gpucomputing.html :

Quote
My application uses the CUDA Runtime API with CUDA Toolkit 2.1, 2.2, or 2.3.
How can I confirm that my application is ready to run on Fermi?
Answer: CUDA applications built using the CUDA Toolkit versions 2.1 through
2.3 are compatible with Fermi as long as they are built to include PTX versions of
their kernels. NVIDIA Driver versions 195.xx or newer allow the application to use
the PTX JIT code path. To test that PTX JIT is working for your application, you
can do the following:
- Go to the NVIDIA website, and install the latest R195 driver.
- Set the system environment flag CUDA_FORCE_PTX_JIT=1
- Launch your application.
When starting a CUDA application for the first time with the above environment
flag, the CUDA driver will JIT compile the PTX for each CUDA kernel that is used
into native CUBIN code. The generated CUBIN for the target GPU architecture is
cached by the CUDA driver. This cache persists across system shutdown/restart
events.
If this test passes, then your application is ready for Fermi.

Well, I got the new cuda 3.0 toolkit and did the following runs with my 9600GT:
No CUDA_FORCE_PTX_JIT flag:
cuda 2 OK
cuda 3 OK

CUDA_FORCE_PTX_JIT flag enabled:
cuda 2 OK
cuda 3 FAIL

What does this tell us? My estimate is that fermi will run the linux client at least with the 2.x libraries. Cuda 3 + ENV flag might be reserved only for fermi hardware, probably that's why it failed on my 9600GT.

Offline Raistmer

  • Working Code Wizard
  • Volunteer Developer
  • Knight who says 'Ni!'
  • *****
  • Posts: 14349
Re: Fermi compatibility for linux cuda client
« Reply #1 on: 21 Mar 2010, 05:21:13 pm »
Did you do some benchmarks already?
If driver can cache native binary for GPU, it could save some time on startup...

Offline sunu

  • Alpha Tester
  • Knight who says 'Ni!'
  • ***
  • Posts: 771
Re: Fermi compatibility for linux cuda client
« Reply #2 on: 21 Mar 2010, 05:39:39 pm »
Unfortunately I didn't think of that beforehand and didn't keep time. Maybe one of these days I'll do a rerun.

Offline Devaster

  • Volunteer Developer
  • Knight who says 'Ni!'
  • *****
  • Posts: 653
  • I like Duke !!!
Re: Fermi compatibility for linux cuda client
« Reply #3 on: 21 Mar 2010, 08:02:09 pm »
so  cufft can be connected to the specified stream ....

 

Welcome, Guest.
Please login or register.
 
 
 
Forgot your password?
Members
Total Members: 97
Latest: ToeBee
New This Month: 0
New This Week: 0
New Today: 0
Stats
Total Posts: 59559
Total Topics: 1672
Most Online Today: 24
Most Online Ever: 983
(20 Jan 2020, 03:17:55 pm)
Users Online
Members: 0
Guests: 36
Total: 36
Powered by EzPortal