Post your BOINC Startup 'CUDA' Info

Message boards : Number crunching : Post your BOINC Startup 'CUDA' Info
Message board moderation

To post messages, you must log in.

Previous · 1 . . . 7 · 8 · 9 · 10 · 11 · 12 · Next

AuthorMessage
Joe10169

Send message
Joined: 2 Dec 05
Posts: 1
Credit: 6,510
RAC: 0
United States
Message 950834 - Posted: 29 Nov 2009, 18:46:28 UTC

11/29/2009 10:48:12 AM Processor: 4 AuthenticAMD AMD Phenom(tm) 9550 Quad-Core Processor [x86 Family 16 Model 2 Stepping 3]

11/29/2009 10:48:12 AM NVIDIA GPU 0: GeForce 9800 GT (driver version 19562, CUDA version 3000, compute capability 1.1, 512MB, 336 GFLOPS peak)

11/29/2009 10:48:12 AM NVIDIA GPU 1: GeForce 9800 GT (driver version 19562, CUDA version 3000, compute capability 1.1, 512MB, 336 GFLOPS peak)

ID: 950834 · Report as offensive
jrmy

Send message
Joined: 14 Jun 04
Posts: 10
Credit: 21,475
RAC: 0
United States
Message 950872 - Posted: 29 Nov 2009, 22:05:42 UTC

11/29/2009 1:03:23 PM NVIDIA GPU 0: GeForce GTX 260 (driver version 19038, CUDA version 2030, compute capability 1.3, 896MB, 583 GFLOPS peak)
ID: 950872 · Report as offensive
jrmy

Send message
Joined: 14 Jun 04
Posts: 10
Credit: 21,475
RAC: 0
United States
Message 950873 - Posted: 29 Nov 2009, 22:09:02 UTC - in response to Message 950431.  
Last modified: 29 Nov 2009, 22:09:28 UTC

Not that I use my Graphics card for crunching, but I'll post here anyway.
28/11/2009 3:17:45 p.m.		NVIDIA GPU 0: GeForce 8600M GT (driver version 18681, CUDA version 2020, compute capability 1.1, 256MB, 61 GFLOPS peak)

Hopefully the GTX260 C216 Superclock in my gaming build will do a bit better!
i have a core 216 and, as posted above, it's specs are: 11/29/2009 1:03:23 PM NVIDIA GPU 0: GeForce GTX 260 (driver version 19038, CUDA version 2030, compute capability 1.3, 896MB, 583 GFLOPS peak)
ID: 950873 · Report as offensive
Luke
Volunteer developer
Avatar

Send message
Joined: 31 Dec 06
Posts: 2546
Credit: 817,560
RAC: 0
New Zealand
Message 950879 - Posted: 29 Nov 2009, 22:23:55 UTC - in response to Message 950873.  
Last modified: 29 Nov 2009, 22:24:05 UTC

Not that I use my Graphics card for crunching, but I'll post here anyway.
28/11/2009 3:17:45 p.m.		NVIDIA GPU 0: GeForce 8600M GT (driver version 18681, CUDA version 2020, compute capability 1.1, 256MB, 61 GFLOPS peak)

Hopefully the GTX260 C216 Superclock in my gaming build will do a bit better!
i have a core 216 and, as posted above, it's specs are: 11/29/2009 1:03:23 PM NVIDIA GPU 0: GeForce GTX 260 (driver version 19038, CUDA version 2030, compute capability 1.3, 896MB, 583 GFLOPS peak)

Quite a substantial increase in GFLOPS then. Perhaps with a bit of OCing you could reach 10x that of my 8600M GT? It is only 27 GFLOPS away. GTX 260 rocks the GPU world, it is the all round gaming card.
- Luke.
ID: 950879 · Report as offensive
Profile Sutaru Tsureku
Volunteer tester

Send message
Joined: 6 Apr 07
Posts: 7105
Credit: 147,663,825
RAC: 5
Germany
Message 950883 - Posted: 29 Nov 2009, 22:35:58 UTC
Last modified: 29 Nov 2009, 22:37:35 UTC


'Message 950396'

GIGABYTE GTX260(-216) SOC (manufacturer OCed)
680/1500/1250 [GPU/shader/RAM]
BOINC V6.6.x -> 117 GFLOPS
BOINC V6.10.x -> 653 GFLOPS

AFAIK, this is the highest manufacturer OCed GTX260-216.

Who (which manufacturer) can beat GIGABYTE?

ID: 950883 · Report as offensive
Luke
Volunteer developer
Avatar

Send message
Joined: 31 Dec 06
Posts: 2546
Credit: 817,560
RAC: 0
New Zealand
Message 950885 - Posted: 29 Nov 2009, 22:49:55 UTC - in response to Message 950883.  


'Message 950396'

GIGABYTE GTX260(-216) SOC (manufacturer OCed)
680/1500/1250 [GPU/shader/RAM]
BOINC V6.6.x -> 117 GFLOPS
BOINC V6.10.x -> 653 GFLOPS

AFAIK, this is the highest manufacturer OCed GTX260-216.

Who (which manufacturer) can beat GIGABYTE?


I'm not sure how many GFLOPS it pushes, and it isn't a 260-C216, but this is the self-proclaimed "fastest graphics card in the world"
Asus Mars GTX 295 X2 @ Computex
I can find them available at all. And if there is only 1000 of them made, they would be very rare.

It packs 4GB of memory under the chassis, and I can make out that is has an Engine clock of 648Mhz, a Shader clock of 1476(?)Mhz, and a Memory clock of 2.4Ghz!
- Luke.
ID: 950885 · Report as offensive
Profile Sutaru Tsureku
Volunteer tester

Send message
Joined: 6 Apr 07
Posts: 7105
Credit: 147,663,825
RAC: 5
Germany
Message 950889 - Posted: 29 Nov 2009, 22:57:29 UTC


I thought the highest manufacturer OCed GTX260-216 !

BTW. The GIGABYTE have higher settings (all three).

ID: 950889 · Report as offensive
jrmy

Send message
Joined: 14 Jun 04
Posts: 10
Credit: 21,475
RAC: 0
United States
Message 950970 - Posted: 30 Nov 2009, 2:58:15 UTC - in response to Message 950873.  

Not that I use my Graphics card for crunching, but I'll post here anyway.
28/11/2009 3:17:45 p.m.		NVIDIA GPU 0: GeForce 8600M GT (driver version 18681, CUDA version 2020, compute capability 1.1, 256MB, 61 GFLOPS peak)

Hopefully the GTX260 C216 Superclock in my gaming build will do a bit better!
i have a core 216 and, as posted above, it's specs are: 11/29/2009 1:03:23 PM NVIDIA GPU 0: GeForce GTX 260 (driver version 19038, CUDA version 2030, compute capability 1.3, 896MB, 583 GFLOPS peak)

mine is the evga version that comes "superclocked". i'm not sure how much farther i'd want to push it
ID: 950970 · Report as offensive
Luke
Volunteer developer
Avatar

Send message
Joined: 31 Dec 06
Posts: 2546
Credit: 817,560
RAC: 0
New Zealand
Message 951031 - Posted: 30 Nov 2009, 8:04:36 UTC - in response to Message 950970.  

Not that I use my Graphics card for crunching, but I'll post here anyway.
28/11/2009 3:17:45 p.m.		NVIDIA GPU 0: GeForce 8600M GT (driver version 18681, CUDA version 2020, compute capability 1.1, 256MB, 61 GFLOPS peak)

Hopefully the GTX260 C216 Superclock in my gaming build will do a bit better!
i have a core 216 and, as posted above, it's specs are: 11/29/2009 1:03:23 PM NVIDIA GPU 0: GeForce GTX 260 (driver version 19038, CUDA version 2030, compute capability 1.3, 896MB, 583 GFLOPS peak)

mine is the evga version that comes "superclocked". I'm not sure how much farther I'd want to push it


That's the card I am purchasing. You might be able to squeeze a bit more juice out of it, because it is still at EVGA factory defaults, and no matter how much "superclocked" or "overclocked" they advertise it as, you can always push it an inch further. Also, if you have a GPU cooler that can help with the overclocking process as well. Just don't fry it if you try!
- Luke.
ID: 951031 · Report as offensive
angler

Send message
Joined: 19 Oct 00
Posts: 33
Credit: 880,214
RAC: 0
United States
Message 951098 - Posted: 30 Nov 2009, 15:09:47 UTC - in response to Message 950795.  

Results from the same platform

29-Oct-2009 13:23:37 [---] NVIDIA GPU 0: GeForce 9600 GT (driver version 19107, compute capability 1.1, 512MB, est. 37GFLOPS)
11/26/2009 3:20:20 PM NVIDIA GPU 0: GeForce 9600 GT (driver version 19107, CUDA version 2030, compute capability 1.1, 512MB, 208 GFLOPS peak)
11/30/2009 10:00:04 AM NVIDIA GPU 0: GeForce 9600 GT (driver version 19562, CUDA version 3000, compute capability 1.1, 512MB, 208 GFLOPS peak)

ID: 951098 · Report as offensive
Profile screaam

Send message
Joined: 1 Dec 09
Posts: 1
Credit: 2,377,730
RAC: 0
Latvia
Message 951644 - Posted: 2 Dec 2009, 12:33:24 UTC

Hi. I've joined this project yesterday, it's inetresting. I would like to share my result about 9600GT perfomance, but I'm not sure about the result because I have only peak perfomance. How can I get the numbers for the usual perfomance?
02.12.2009 13:58:56 NVIDIA GPU 0: GeForce 9600 GT (driver version 19562, CUDA version 3000, compute capability 1.1, 512MB, 235 GFLOPS peak)

The card is overclocked.
ID: 951644 · Report as offensive
Profile Mamluk
Avatar

Send message
Joined: 10 Sep 09
Posts: 80
Credit: 2,448,048
RAC: 0
South Africa
Message 951647 - Posted: 2 Dec 2009, 13:10:16 UTC

Don't seem to have this one:

NVIDIA GPU 0: GeForce 8800 GS (driver version 19107, CUDA version 2030, compute capability 1.1, 384MB, 326 GFLOPS peak)
ID: 951647 · Report as offensive
Profile hiamps
Volunteer tester
Avatar

Send message
Joined: 23 May 99
Posts: 4292
Credit: 72,971,319
RAC: 0
United States
Message 951654 - Posted: 2 Dec 2009, 13:23:46 UTC - in response to Message 946403.  

11/7/2009 4:56:05 PM BOINC 6.10.17
NVIDIA GPU 0: GeForce 9600 GSO (driver version 19107, CUDA version 2030, compute capability 1.1, 768MB, 276 GFLOPS peak)


I have one of these and have OC'd it to B&*&$*#, and coded all kind of experimental things to see what It can do.... 276GFlops my ...... Clearly these are 'marketing flops'


12/1/2009 6:13:39 PM NVIDIA GPU 0: GeForce GTX 260 (driver version 19562, CUDA version 3000, compute capability 1.3, 896MB, 537 GFLOPS peak)
12/1/2009 6:13:39 PM NVIDIA GPU 1: GeForce 9600 GSO (driver version 19562, CUDA version 3000, compute capability 1.1, 768MB, 276 GFLOPS peak)
Have them in the same machine now and it seems to be about 1/2 as fast or a bit more as the GTX 260.
Official Abuser of Boinc Buttons...
And no good credit hound!
ID: 951654 · Report as offensive
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14655
Credit: 200,643,578
RAC: 874
United Kingdom
Message 951662 - Posted: 2 Dec 2009, 13:52:14 UTC - in response to Message 951654.  

12/1/2009 6:13:39 PM NVIDIA GPU 0: GeForce GTX 260 (driver version 19562, CUDA version 3000, compute capability 1.3, 896MB, 537 GFLOPS peak)
12/1/2009 6:13:39 PM NVIDIA GPU 1: GeForce 9600 GSO (driver version 19562, CUDA version 3000, compute capability 1.1, 768MB, 276 GFLOPS peak)
Have them in the same machine now and it seems to be about 1/2 as fast or a bit more as the GTX 260.

So you have two different CUDA cards in the same machine, and BOINC can correctly enumerate the different compute capabilities, memory capacities, and speed ratings?

Rather puts the kybosh on the B^S posted in your other thread "GPU" in stats page misleading, doesn't it?

I'm sure that the web page issue is purely to do with storing multiple copies of the equivalent data in the back-end database, and displaying it efficiently/compactly on the website. And knowing that they can't store/display it properly, it may not all be passed back in the comms channel either - but it's available at the head-end, so of course it could be passed back, stored and used if the will was there.
ID: 951662 · Report as offensive
Profile hiamps
Volunteer tester
Avatar

Send message
Joined: 23 May 99
Posts: 4292
Credit: 72,971,319
RAC: 0
United States
Message 951700 - Posted: 2 Dec 2009, 17:26:41 UTC - in response to Message 951662.  

12/1/2009 6:13:39 PM NVIDIA GPU 0: GeForce GTX 260 (driver version 19562, CUDA version 3000, compute capability 1.3, 896MB, 537 GFLOPS peak)
12/1/2009 6:13:39 PM NVIDIA GPU 1: GeForce 9600 GSO (driver version 19562, CUDA version 3000, compute capability 1.1, 768MB, 276 GFLOPS peak)
Have them in the same machine now and it seems to be about 1/2 as fast or a bit more as the GTX 260.

So you have two different CUDA cards in the same machine, and BOINC can correctly enumerate the different compute capabilities, memory capacities, and speed ratings?

Rather puts the kybosh on the B^S posted in your other thread "GPU" in stats page misleading, doesn't it?

I'm sure that the web page issue is purely to do with storing multiple copies of the equivalent data in the back-end database, and displaying it efficiently/compactly on the website. And knowing that they can't store/display it properly, it may not all be passed back in the comms channel either - but it's available at the head-end, so of course it could be passed back, stored and used if the will was there.



No it doesn't. I was talking about the the STATISTICS page like up there on your top right link on this page. Check out the top computers, pretty soon mine will say 2 GeForce 9600GSO...On the PUBLIC stat pages it does not distinguish what kind of cards you have it just tells how many and assumes they are the same as what you put in PCIE slot 1. Not a bug just not as useful as it could be because you can't compare yourself with other with multple cards as you don't know what the second and third card may be.
Official Abuser of Boinc Buttons...
And no good credit hound!
ID: 951700 · Report as offensive
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14655
Credit: 200,643,578
RAC: 874
United Kingdom
Message 951703 - Posted: 2 Dec 2009, 17:42:16 UTC - in response to Message 951700.  

12/1/2009 6:13:39 PM NVIDIA GPU 0: GeForce GTX 260 (driver version 19562, CUDA version 3000, compute capability 1.3, 896MB, 537 GFLOPS peak)
12/1/2009 6:13:39 PM NVIDIA GPU 1: GeForce 9600 GSO (driver version 19562, CUDA version 3000, compute capability 1.1, 768MB, 276 GFLOPS peak)
Have them in the same machine now and it seems to be about 1/2 as fast or a bit more as the GTX 260.

So you have two different CUDA cards in the same machine, and BOINC can correctly enumerate the different compute capabilities, memory capacities, and speed ratings?

Rather puts the kybosh on the B^S posted in your other thread "GPU" in stats page misleading, doesn't it?

I'm sure that the web page issue is purely to do with storing multiple copies of the equivalent data in the back-end database, and displaying it efficiently/compactly on the website. And knowing that they can't store/display it properly, it may not all be passed back in the comms channel either - but it's available at the head-end, so of course it could be passed back, stored and used if the will was there.

No it doesn't. I was talking about the the STATISTICS page like up there on your top right link on this page. Check out the top computers, pretty soon mine will say 2 GeForce 9600GSO...On the PUBLIC stat pages it does not distinguish what kind of cards you have it just tells how many and assumes they are the same as what you put in PCIE slot 1. Not a bug just not as useful as it could be because you can't compare yourself with other with multple cards as you don't know what the second and third card may be.

Sorry, speed-reading and posting replies in a hurry before going out for an appointment, again.

I know you were talking about the statistics page: I think you'll also find the same on the 'host details' page. Looking at your host 5185924, I see:

Coprocessors [2] NVIDIA GeForce GTX 260 (896MB) driver: 19562

which is rather different from the message log snip at the head of this post.

The B^S I was referring to - sorry Gundolf - was the first reply you got: "IIRC, it's a matter of the CUDA drivers". Oh no it isn't: the 'unless' clause which follows (but which I missed the first time round) certainly applies: "there's a difference between what BOINC manager says and what the 'Computers belonging to ...' pages tell".

So, as I said, the problem is simply that the BOINC database and website haven't been designed to store/display all the available information.
ID: 951703 · Report as offensive
angler

Send message
Joined: 19 Oct 00
Posts: 33
Credit: 880,214
RAC: 0
United States
Message 951956 - Posted: 3 Dec 2009, 14:45:38 UTC

results from progression on single machine.

12/2/2009 2:05:51 PM NVIDIA GPU 0: GeForce GTS 250 (driver version 19107, CUDA version 2030, compute capability 1.1, 512MB, 470 GFLOPS peak)

06-Oct-2009 10:21:36 [---] NVIDIA GPU 0: GeForce GTS 250 (driver version 19062, compute capability 1.1, 512MB, est. 84GFLOPS)

27-Jun-2009 10:32:13 [---] CUDA device: GeForce 9600 GSO (driver version 18585, compute capability 1.1, 768MB, est. 49GFLOPS)


ID: 951956 · Report as offensive
Profile Mamluk
Avatar

Send message
Joined: 10 Sep 09
Posts: 80
Credit: 2,448,048
RAC: 0
South Africa
Message 952952 - Posted: 7 Dec 2009, 10:21:29 UTC

NVIDIA GPU 0: GeForce GT 240 (driver version 19562, CUDA version 3000, compute capability 1.2, 512MB, 279 GFLOPS peak)
ID: 952952 · Report as offensive
Crun-chi
Volunteer tester
Avatar

Send message
Joined: 3 Apr 99
Posts: 174
Credit: 3,037,232
RAC: 0
Croatia
Message 952965 - Posted: 7 Dec 2009, 12:27:43 UTC - in response to Message 952952.  

NVIDIA GPU 0: GeForce GT 240 (driver version 19562, CUDA version 3000, compute capability 1.2, 512MB, 257 GFLOPS peak)

I am cruncher :)
I LOVE SETI BOINC :)
ID: 952965 · Report as offensive
Profile hiamps
Volunteer tester
Avatar

Send message
Joined: 23 May 99
Posts: 4292
Credit: 72,971,319
RAC: 0
United States
Message 952974 - Posted: 7 Dec 2009, 13:39:11 UTC - in response to Message 952965.  

NVIDIA GPU 0: GeForce GT 240 (driver version 19562, CUDA version 3000, compute capability 1.2, 512MB, 257 GFLOPS peak)

Seem about the same as my 9600GSO hopefully the GT 240 is faster. Can't tell as you both have hidden computers...
Official Abuser of Boinc Buttons...
And no good credit hound!
ID: 952974 · Report as offensive
Previous · 1 . . . 7 · 8 · 9 · 10 · 11 · 12 · Next

Message boards : Number crunching : Post your BOINC Startup 'CUDA' Info


 
©2024 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.