V10/11 of modified SETI MB CUDA + opt AP package for full multi-GPU+CPU use

[k6xt]

My own Q9450 did ~4400 with MB tasks before CUDA start. Now it ~5200 with 9600GSO addition. But it still not stabilized (I expect it to be higher).

Am I correct to thinik the important GPU spec is core clock?

You hinted in earlier post that stream processors are not involved in GPU crunching. How important is memory clock?

Thanks Art

[Raistmer]

My own Q9450 did ~4400 with MB tasks before CUDA start. Now it ~5200 with 9600GSO addition. But it still not stabilized (I expect it to be higher).

Am I correct to thinik the important GPU spec is core clock?

You hinted in earlier post that stream processors are not involved in GPU crunching. How important is memory clock?

Thanks Art

Not correct to think stream processors not included. I said "texture-related hardware", not stream processors. There is dedicated hardware that will do linear approximation between 2 dots in texture for example. This particular part can be used inside CUDA BTW, but I don't think this app use it. But there are another parts of GPU that can't be accessed via CUDA at all.
Will try to find relevant info in CUDA FAQ or forums (I seen that there).

[Raistmer]

Ok, here:
"
What texture features does CUDA support?

CUDA supports 1D and 2D textures, which can be accessed with normalized (0..1) or integer coordinates. Textures can also be bound to linear memory and accessed with the "tex1Dfetch" function.

The hardware only supports 1, 2 and 4-component textures, not 3-component textures.

3D textures, cube maps, texture arrays, compressed textures and mip-maps are not currently supported.
"

and

"
Are graphics operations such as z-buffering and alpha blending supported in CUDA?

No. Access to video memory in CUDA is done via the load/store mechanism, and doesn't go through the normal graphics raster operations like blending. We don't have any plans to expose blending or any other raster ops in CUDA.
"

AND AFAIK any game will do blending a lot

[Raistmer]

Am I correct to thinik the important GPU spec is core clock?

Think no.
look beta forum for GPU OCing questions. It was showed that shader clock most important and engine clock least important.
My own GPU can support high memory freqs (for now it works on 2x1140 and 2x1120 known to be stable already). With shader freqs it still awaiting evaluation...

[k6xt]

Ok, here:
"
What texture features does CUDA support?

CUDA supports 1D and 2D textures, which can be accessed with normalized (0..1) or integer coordinates. Textures can also be bound to linear memory and accessed with the "tex1Dfetch" function.

The hardware only supports 1, 2 and 4-component textures, not 3-component textures.

3D textures, cube maps, texture arrays, compressed textures and mip-maps are not currently supported.
"

and

"
Are graphics operations such as z-buffering and alpha blending supported in CUDA?

No. Access to video memory in CUDA is done via the load/store mechanism, and doesn't go through the normal graphics raster operations like blending. We don't have any plans to expose blending or any other raster ops in CUDA.
"

AND AFAIK any game will do blending a lot

From this maybe I can infer CUDA uses some but not all capabilities. Just considering SETI or Einstein processing, then, there must be a point beyond which it is wasted money to buy more performance. Would it be true that increased core clock is always valuable, memory less so? And, is it possible to decide how many stream processors is enough, or is the case "more is better"?

I am trying to decide how best to implement the new PC with regard to SETI processing. Given I don't understand much about video processing, guessing which specs will yield the most result per dollar is a task. The spec sheets that I've read for example don't talk about texture so it is hard to relate to specs.

Regards Art

[Raistmer]

Look Richard's thread on SETI main about CUDA completion times for this.
You can see there if high end card give adequate performance boost or not comparing to their prices....
Maybe 2 midrange cards could work better than one more costly high-end card... Performance-related info still not so big to give easy answers on such questions...

BTW, if GPU supportd computation capability 1.3 - the SETI CUDA MB app can make use of it.
In my understandig it's a rough analogue to CPU SSE level (that is, new instructions added). And CUDA MB can use them.
In other aspects - the more processors GPU have and the higher its freqs - the more faster it will be (also memory bus bandwith should matter).

[toomnymods]

Hello I'm wondering if my computer is really crunching as fast as it should

http://setiathome.berkeley.edu/show_host_detail.php?hostid=4686284&ipprivate=1 Link to my desktop /w 2 EVGA GTX260 gpu's and a quadcore Q6600 intel running on Windows XP sp3
I have downloaded your new V10a x86 sse3x from your first post (the top one i downloaded is this correct for 32 bit windows?)
I am running the 6.4.7 version of boinc
I have the cc_config.xml set to 6 processes

just seems awfully slow to be running 6 processes at once and only have a 8K RAC which is excactly the same as when i had 1 gpu.. i have gotten absolutely no gain from running 2 gpu's thus far..
thanks

[MAOJC]

Did you ever build on for older SSE2 AMD CPUs + a GPU?

What config do you have? For what CPU you need SSE2 app ?

Opteron 170/270/280 AMD

Its motherboard has PCI-E slots?

Yes running a 9600XT on it now and standard seticuda under XP.

Ok, CPU-Z output (in html form or as text file) would be nice for this host.

Sorry it does SSE3 My Duh!

[k6xt]

Hello I'm wondering if my computer is really crunching as fast as it should

http://setiathome.berkeley.edu/show_host_detail.php?hostid=4686284&ipprivate=1 Link to my desktop /w 2 EVGA GTX260 gpu's and a quadcore Q6600 intel running on Windows XP sp3
I have downloaded your new V10a x86 sse3x from your first post (the top one i downloaded is this correct for 32 bit windows?)
I am running the 6.4.7 version of boinc
I have the cc_config.xml set to 6 processes

just seems awfully slow to be running 6 processes at once and only have a 8K RAC which is excactly the same as when i had 1 gpu.. i have gotten absolutely no gain from running 2 gpu's thus far..
thanks

Same here. So far I've not seen any action from the GPU.

Have you had any trouble with the EVGA card? I saw a post suggesting EVGA was having some issues.

Thanks Art

[Geek@Play]

Off subject a bit..............

I remember there existed a NON TEAM build somewhere but can't find it.  Can anyone help me out?

Looking for NON TEAM cuda app

[toomnymods]

Hello I'm wondering if my computer is really crunching as fast as it should

http://setiathome.berkeley.edu/show_host_detail.php?hostid=4686284&ipprivate=1 Link to my desktop /w 2 EVGA GTX260 gpu's and a quadcore Q6600 intel running on Windows XP sp3
I have downloaded your new V10a x86 sse3x from your first post (the top one i downloaded is this correct for 32 bit windows?)
I am running the 6.4.7 version of boinc
I have the cc_config.xml set to 6 processes

just seems awfully slow to be running 6 processes at once and only have a 8K RAC which is excactly the same as when i had 1 gpu.. i have gotten absolutely no gain from running 2 gpu's thus far..
thanks

Same here. So far I've not seen any action from the GPU.

Have you had any trouble with the EVGA card? I saw a post suggesting EVGA was having some issues.

Thanks Art

so far no issues but my gpu's arent crunching wu's in 2-4 mins like they where when i just had a single gpu.. after adding second it is taking like 20-30 mins to crunch them now.. and 1hr 30 mins for cpu to complete a wu, where before it would do it in 45mins to 58 mins tops..

i dont see the purpose of running 2 gpu processes if it slows everything else down it isnt worth the expense of extra card and extra electricity to run the damn thing

[Raistmer]

Hello I'm wondering if my computer is really crunching as fast as it should

http://setiathome.berkeley.edu/show_host_detail.php?hostid=4686284&ipprivate=1 Link to my desktop /w 2 EVGA GTX260 gpu's and a quadcore Q6600 intel running on Windows XP sp3
I have downloaded your new V10a x86 sse3x from your first post (the top one i downloaded is this correct for 32 bit windows?)
I am running the 6.4.7 version of boinc
I have the cc_config.xml set to 6 processes

just seems awfully slow to be running 6 processes at once and only have a 8K RAC which is excactly the same as when i had 1 gpu.. i have gotten absolutely no gain from running 2 gpu's thus far..
thanks

"
setiathome_CUDA: No device specified, determined to use CUDA device 1: GeForce GTX 260
SETI@home using CUDA accelerated device GeForce GTX 260
CPU-GPU team mod by Raistmer
"
It's not V10 output. Update to V10b version.

ADDON: just for example, true V10 output:
"
setiathome_CUDA: CUDA Device 1 specified, checking...
   Device 1: GeForce 9600 GSO is okay
SETI@home using CUDA accelerated device GeForce 9600 GSO
V10 modification by Raistmer
"

[Raistmer]

Off subject a bit..............

I remember there existed a NON TEAM build somewhere but can't find it.  Can anyone help me out?

Looking for NON TEAM cuda app

The same CUDA app can be used w/o its team CPU counterpart. But CPU teamed app can not be used separately.

[Raistmer]

Hello I'm wondering if my computer is really crunching as fast as it should

http://setiathome.berkeley.edu/show_host_detail.php?hostid=4686284&ipprivate=1 Link to my desktop /w 2 EVGA GTX260 gpu's and a quadcore Q6600 intel running on Windows XP sp3
I have downloaded your new V10a x86 sse3x from your first post (the top one i downloaded is this correct for 32 bit windows?)
I am running the 6.4.7 version of boinc
I have the cc_config.xml set to 6 processes

just seems awfully slow to be running 6 processes at once and only have a 8K RAC which is excactly the same as when i had 1 gpu.. i have gotten absolutely no gain from running 2 gpu's thus far..
thanks

Same here. So far I've not seen any action from the GPU.

Have you had any trouble with the EVGA card? I saw a post suggesting EVGA was having some issues.

Thanks Art

so far no issues but my gpu's arent crunching wu's in 2-4 mins like they where when i just had a single gpu.. after adding second it is taking like 20-30 mins to crunch them now.. and 1hr 30 mins for cpu to complete a wu, where before it would do it in 45mins to 58 mins tops..

i dont see the purpose of running 2 gpu processes if it slows everything else down it isnt worth the expense of extra card and extra electricity to run the damn thing

Is it your result?
http://setiathome.berkeley.edu/result.php?resultid=1180810147
it takes ~500seconds to complete, 500 realtime seconds, and 120.2188 CPU seconds.
What you talk about?

[toomnymods]

"
setiathome_CUDA: No device specified, determined to use CUDA device 1: GeForce GTX 260
SETI@home using CUDA accelerated device GeForce GTX 260
CPU-GPU team mod by Raistmer
"
It's not V10 output. Update to V10b version.

ADDON: just for example, true V10 output:
"
setiathome_CUDA: CUDA Device 1 specified, checking...
   Device 1: GeForce 9600 GSO is okay
SETI@home using CUDA accelerated device GeForce 9600 GSO
V10 modification by Raistmer
"

i dont understand what you are asking me to do?? where is version 10b?