Things aren't looking very good right now in the SETI world. I just about have everything together for my new Red rig (Q9550, 3x GTX 295) but now I'm having 2nd thoughts since I can't even keep my current two rigs crunching 24/7.

So Red might become a Bitcoin miner instead. Anyone tried this yet? It seems to have a lot in common with S@H as it uses GPU HP. I've even seen mining rigs using PCIe extenders so they can run 8 cards off of one MB.

TA-I've been able to get WU's without a problem....I don't see an issue with running Seti along with a separate project to fill in the gaps.

to the tune of "Here We Go Again"

<cmdline>-period_iterations_num 20 -instances_per_device 1</cmdline>
<coproc>
<type>ATI</type>
<count>1</count>

missed the last one

What should the count be (0.5 like the others?), I just went back to look at prior posts and someones removed a bunch of posts.

I don't remember that being one of the one's to change, but of course I could have missed it.

Dan/Philly

I'd say, leave that set of instructions as is. For now

<cmdline>-period_iterations_num 20 -instances_per_device 1</cmdline>
<coproc>
<type>ATI</type>
<count>1</count>

missed the last one

there are 2 sets of this.

I'm sorry i only told you to edit the <cmdline> in 2 places

Please add <flops> and the <count> to the second set as well

Sorry I forgot...

you have "instances per device" set to 2 but you have your count set to one so you are asking boinc to run 2 but each Wu uses the whole GPU which means you are basically having 2 WU's active but only processing 1 at a time.

use Notepad to edit your app_info, as i said earlier.

Claggy

Here is my current app file, should I just copy the other app file into it. I did take notice that suggested file has SSE2 in it VS SSSE3.


<cmdline>-instances_per_device 1 -unroll 6 -ffa_block 4096 -ffa_block_fetch 2048</cmdline>
- <coproc>
<type>ATI</type>
<count>1</count>
</coproc>

I think you could keep your file and just edit a couple of lines.

<cmdline>-instances_per_device 2 -unroll 8 -ffa_block 8192 -ffa_block_fetch 4096</cmdline>
<flops>209876543210</flops>
<coproc>
<type>ATI</type>
<count>0.5</count>
</coproc>


Edit your file where you find the <cmdline> to my suggested line. There are 2 such lines.

Add the line <flops>209876543210</flops> accordingly where it is in my suggestion.

And make sure you edit the <count>0.5</count> accordingly

I recomend 11.9 that I am pleased with.
But you may choose 11.2-11.4 as they also work(maybe a little bit faster).

11.9 uses APP SDK2.5 for OpenCl

Before then, please make sure OpenCL is available on your GPU, Cat 10.10 drivers were released with and without OpenCL support, if you haven't got OpenCL support then the ATI GPU tasks will error out.

Claggy

SSE2x

Claggy, thanks for highlighting the SSE2x, I'm sure I'd have missed it. At your convenience, let me know if the changes I made were correct, obviously I think they were :)

Yep, the app has changed to the SSSE3x app, and the times have come down a bit too:

<core_client_version>6.10.60</core_client_version>
<![CDATA[
<stderr_txt>
Windows optimized S@H Enhanced application by Alex Kan
Version info: v8b2-SSSE3x (Intel, Core 2-optimized v8-nographics) V5.13 by Alex Kan
SSSE3x Win32 Build 386 , Ported by : Jason G, Raistmer, JDWhale

CPUID: Intel(R) Core(TM)2 CPU 6600 @ 2.40GHz
Speed: 2 x 2400 MHz
Cache: L1=64K L2=4096K
Features: MMX SSE SSE2 SSE3 SSSE3

Work Unit Info:
...............
Credit multiplier is : 2.85
WU true angle range is : 2.591619

Flopcounter: 10473321425573.945000

Spike count: 6
Pulse count: 0
Triplet count: 0
Gaussian count: 0
called boinc_finish
boinc_exit(): requesting safe worker shutdown ->
boinc_exit(): received safe worker shutdown acknowledge ->

</stderr_txt>
]]>

Did you install the ATI GPU apps?, as you haven't had any Wu's yet, i assume your GPU is busy on other projects and your cache is full for it,

Claggy

Philadelphia:

I suggest you run the astropulse r.521 here since they give the best credits as of now.
Check seti server status if there is any available task.
Set your work cache to as many days of work you want to have in your cache.

Some says ATI driver 11.2-11,4 is the best I use ATI driver 11.9. Do not use 11.10-11.11 with OpenCl


Create or edit your app_info.xml so it runs 2 tasks at the same time on the GPU.

Here is an example of the app_info.xml that I use.

_________________________
<app_info>
<app>
<name>astropulse_v505</name>
</app>
<file_info>
<name>ap_5.06_win_x86_SSE2_OpenCL_ATI_r521.exe</name>
<executable/>
</file_info>
<file_info>
<name>AstroPulse_Kernels_r521.cl</name>
<executable/>
</file_info>
<app_version>
<app_name>astropulse_v505</app_name>
<version_num>506</version_num>
<avg_ncpus>0.04</avg_ncpus>
<max_ncpus>0.2</max_ncpus>
<plan_class>ati13ati</plan_class>
<cmdline>-instances_per_device 2 -unroll 8 -ffa_block 8192 -ffa_block_fetch 4096 -hp</cmdline>
<flops>209876543210</flops>
<coproc>
<type>ATI</type>
<count>0.5</count>
</coproc>
<file_ref>
<file_name>ap_5.06_win_x86_SSE2_OpenCL_ATI_r521.exe</file_name>
<main_program/>
</file_ref>
<file_ref>
<file_name>AstroPulse_Kernels_r521.cl</file_name>
<copy_file/>
</file_ref>
</app_version>
<app_version>
<app_name>astropulse_v505</app_name>
<version_num>505</version_num>
<avg_ncpus>0.04</avg_ncpus>
<max_ncpus>0.2</max_ncpus>
<plan_class>ati13ati</plan_class>
<cmdline>-instances_per_device 2 -unroll 8 -ffa_block 8192 -ffa_block_fetch 4096 -hp</cmdline>
<flops>209876543210</flops>
<coproc>
<type>ATI</type>
<count>0.5</count>
</coproc>
<file_ref>
<file_name>ap_5.06_win_x86_SSE2_OpenCL_ATI_r521.exe</file_name>
<main_program/>
</file_ref>
<file_ref>
<file_name>AstroPulse_Kernels_r521.cl</file_name>
<copy_file/>
</file_ref>
</app_version>
</app_info>
_______________________________

Your app_info.xml might look a bit different. My app only runs on the GPU so it misses the Lines for CPU.

//TRuEQ