AP V7

Message boards : Number crunching : AP V7
Message board moderation

To post messages, you must log in.

Previous · 1 · 2 · 3 · 4 · 5 · 6 . . . 20 · Next

AuthorMessage
Chris Adamek
Volunteer tester

Send message
Joined: 15 May 99
Posts: 251
Credit: 434,772,072
RAC: 236
United States
Message 1584132 - Posted: 9 Oct 2014, 19:53:39 UTC - in response to Message 1583534.  

Could you post the Mac versions to your site as well?

Thanks,

Chris

If there are no optimized applications available at the moment, then give at least a link to download the standard distribution AP7 application(s) for the anonymous platform users to be prepared. ;)

If Mike and Arkayn could mirror the final release packages for each individual app on their sites, please, it would save me breaking away from trying to knock the whole installer into shape.


I have them up on my site now as well.
http://www.arkayn.us/forum/index.php?action=tpmod;dl=cat105

ID: 1584132 · Report as offensive
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14649
Credit: 200,643,578
RAC: 874
United Kingdom
Message 1584224 - Posted: 9 Oct 2014, 21:42:21 UTC

The Lunatics installer has gone into closed Beta testing, just as the first tape starts to split....

Testing mostly takes place in European time zones, so we're off to bed now (via the pub). Please leave us some AP v7 tasks for full testing in the morning.
ID: 1584224 · Report as offensive
Claggy
Volunteer tester

Send message
Joined: 5 Jul 99
Posts: 4654
Credit: 47,537,079
RAC: 4
United Kingdom
Message 1584226 - Posted: 9 Oct 2014, 21:46:09 UTC
Last modified: 9 Oct 2014, 22:06:14 UTC

I've received a couple of AstroPulse v7 v7.03 (sse), and AstroPulse v7 v7.05 (opencl_nvidia_100) tasks, But no AstroPulse v7 7.04 (opencl_ati_100) tasks (yet) which is strange. Edit: a little later one is sent.

All AstroPulse v7 tasks for computer 5427475

Claggy
ID: 1584226 · Report as offensive
woohoo
Volunteer tester

Send message
Joined: 30 Oct 13
Posts: 972
Credit: 165,671,404
RAC: 5
United States
Message 1584238 - Posted: 9 Oct 2014, 22:05:06 UTC

I did receive one AstroPulse v7 7.04 ati_opencl_100 but I was going to let it sit there until my computer got through the other MBs. The time estimate says 11:37:30 which is rather long but nothing to worry about now.
ID: 1584238 · Report as offensive
Claggy
Volunteer tester

Send message
Joined: 5 Jul 99
Posts: 4654
Credit: 47,537,079
RAC: 4
United Kingdom
Message 1584239 - Posted: 9 Oct 2014, 22:07:13 UTC - in response to Message 1584238.  

First one is complete, just awaiting my wingman.

Claggy
ID: 1584239 · Report as offensive
Profile arkayn
Volunteer tester
Avatar

Send message
Joined: 14 May 99
Posts: 4438
Credit: 55,006,323
RAC: 0
United States
Message 1584243 - Posted: 9 Oct 2014, 22:18:17 UTC - in response to Message 1584132.  

Could you post the Mac versions to your site as well?

Thanks,

Chris

If there are no optimized applications available at the moment, then give at least a link to download the standard distribution AP7 application(s) for the anonymous platform users to be prepared. ;)

If Mike and Arkayn could mirror the final release packages for each individual app on their sites, please, it would save me breaking away from trying to knock the whole installer into shape.


I have them up on my site now as well.
http://www.arkayn.us/forum/index.php?action=tpmod;dl=cat105


I have a couple of them up, but you will have to make your own app_info for now.

ID: 1584243 · Report as offensive
Profile HAL9000
Volunteer tester
Avatar

Send message
Joined: 11 Sep 99
Posts: 6534
Credit: 196,805,888
RAC: 57
United States
Message 1584251 - Posted: 9 Oct 2014, 22:47:24 UTC - in response to Message 1584238.  

I did receive one AstroPulse v7 7.04 ati_opencl_100 but I was going to let it sit there until my computer got through the other MBs. The time estimate says 11:37:30 which is rather long but nothing to worry about now.

It went a bit the other direction with my iGPU. Giving me an estimated 15 hours rather than the 21.5 hours it normally takes with AP v6. The CPU is clocked in rather high, as expected, at 158 hours. So that should get whittled down to 23 hours over time.
SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[
ID: 1584251 · Report as offensive
Profile BilBg
Volunteer tester
Avatar

Send message
Joined: 27 May 07
Posts: 3720
Credit: 9,385,827
RAC: 0
Bulgaria
Message 1584253 - Posted: 9 Oct 2014, 22:55:13 UTC - in response to Message 1584238.  

The time estimate says 11:37:30 which is rather long ...

Mine is 12 days :)
But I know it will finish in ~2 h if run alone or ~3:30 if run along AstroPulse v6 (APv6 0.48 + APv7 0.51)

From BoincTasks:
7.04 astropulse_v7 (opencl_ati_100) ap_01au10ad_B1_P0_00088_20141009_18779.wu_2 - (-) 0.00 0.000    12d,03:42:56    03-11-2014 23:48 0.04C + 0.51ATI   Ready to start
 


- ALF - "Find out what you don't do well ..... then don't do it!" :)
 
ID: 1584253 · Report as offensive
Profile HAL9000
Volunteer tester
Avatar

Send message
Joined: 11 Sep 99
Posts: 6534
Credit: 196,805,888
RAC: 57
United States
Message 1584265 - Posted: 9 Oct 2014, 23:22:27 UTC

ID: 1584265 · Report as offensive
Wedge009
Volunteer tester
Avatar

Send message
Joined: 3 Apr 99
Posts: 451
Credit: 431,396,357
RAC: 553
Australia
Message 1584274 - Posted: 9 Oct 2014, 23:48:54 UTC

Interesting observation.

I like how your host finished the task in half an hour despite it being 87% blanked. Looking forward to no more blanking processing on the CPU.
Soli Deo Gloria
ID: 1584274 · Report as offensive
Urs Echternacht
Volunteer tester
Avatar

Send message
Joined: 15 May 99
Posts: 692
Credit: 135,197,781
RAC: 211
Germany
Message 1584280 - Posted: 10 Oct 2014, 0:02:08 UTC - in response to Message 1584265.  

Recocognized. Sadly the error happens outside of the APv7 app in /usr/lib/system/libsystem_kernel.dylib
_\|/_
U r s
ID: 1584280 · Report as offensive
Profile ivan
Volunteer tester
Avatar

Send message
Joined: 5 Mar 01
Posts: 783
Credit: 348,560,338
RAC: 223
United Kingdom
Message 1584287 - Posted: 10 Oct 2014, 0:13:59 UTC - in response to Message 1584239.  

First one is complete, just awaiting my wingman.

I've had two complete and pending (3,900 secs each/400 secs CPU) and seven in progress on my "stock" Linux box.
ID: 1584287 · Report as offensive
Lionel

Send message
Joined: 25 Mar 00
Posts: 680
Credit: 563,640,304
RAC: 597
Australia
Message 1584309 - Posted: 10 Oct 2014, 0:59:18 UTC - in response to Message 1584287.  

I don't know if this is an issue and has been picked up or if it is a null event but there is an inconsistency in the readme file for r2721 for NV.

Under the "NV Titan ..." title it has unroll 16 ...16384 ...

Under the "For NV x80/x70" title it has ... unroll 18 ... 16384

=====================


NV Titan and 780 TI

-unroll 16 -ffa_block 16384 -ffa_block_fetch 8192

* Bigger unroll values < 20 doesn`t necessarily result in better run times.
Further testing required.

NV x70/x80

-unroll 12 -ffa_block 12288 -ffa_block_fetch 6144

NV x50/x60 and x60 TI

-unroll 10 -ffa_block 6144 -ffa_block_fetch 1536

NV x30/x40

-unroll 4 -ffa_block 2048 -ffa_block_fetch 1024



-tune switch

Tune values must be equal or less than max work group size.
Most modern Nvidia cards have work group size of 1024.

possible values:

-tune 1 256 4 1
-tune 1 128 8 1
-tune 1 64 16 1
-tune 1 32 32 1
-tune 1 16 64 1


Intensive testing highlighted -tune 1 64 8 1 -tune 2 64 8 1 to be fastest on mid range and high end GPU`s.
On entry level cards -tune 1 128 8 1 -tune 2 128 8 1 should be fastest.
Further testing required for other GPU`s.

-oclFFT_plan switch

Use at your own risk !
------------------------


FFT kernels are processed in 8 point fft kernels by default.
Using different fft kernel planning can speed up processing significantly.
In most cases 16 point fft kernels are fastest for AstroPulse V7.

-oclFFT_plan 256 16 256

On entry level cards like the NV 610 using full work group size is faster.

-oclFFT_plan 256 16 1024

For those suffering from high CPU usage please use -use_sleep switch.

Note:
Using -use_sleep switch requires bigger unroll values to get better GPU utilization.

Example:

For NV x80/x70
-use_sleep -unroll 18 -oclFFT_plan 256 16 256 -ffa_block 16384 -ffa_block_fetch 8192 -tune 1 64 8 1 -tune 2 64 8 1
ID: 1584309 · Report as offensive
W-K 666 Project Donor
Volunteer tester

Send message
Joined: 18 May 99
Posts: 19012
Credit: 40,757,560
RAC: 67
United Kingdom
Message 1584316 - Posted: 10 Oct 2014, 1:17:10 UTC
Last modified: 10 Oct 2014, 1:22:03 UTC

Well that went smooothly, can report a "percent blanked: 74.54" done in under 28 mins on a 670.
and a "percent blanked: 70.98" in ~32 min

edit} how was the high blanked outliers detected in version 6, because I have a Ver7 validated but not showing up in Application Details, "Number of tasks completed"
ID: 1584316 · Report as offensive
Profile Zeus Fab3r
Avatar

Send message
Joined: 17 Jan 01
Posts: 649
Credit: 275,335,635
RAC: 597
Serbia
Message 1584320 - Posted: 10 Oct 2014, 1:48:20 UTC

Is there any minimum driver requirements for running V7 NV app? I've got 4 tasks which failed out immediately with ominous "output file absent" error message. The only thing I can think of is that I'm using an rather old driver.

Who the hell is General Failure and why is he reading my harddisk?¿
ID: 1584320 · Report as offensive
Profile Zalster Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 27 May 99
Posts: 5517
Credit: 528,817,460
RAC: 242
United States
Message 1584323 - Posted: 10 Oct 2014, 1:56:20 UTC - in response to Message 1584320.  
Last modified: 10 Oct 2014, 2:08:34 UTC

Sorry, are you running stock?
ID: 1584323 · Report as offensive
Profile HAL9000
Volunteer tester
Avatar

Send message
Joined: 11 Sep 99
Posts: 6534
Credit: 196,805,888
RAC: 57
United States
Message 1584324 - Posted: 10 Oct 2014, 1:58:46 UTC - in response to Message 1584316.  

Well that went smooothly, can report a "percent blanked: 74.54" done in under 28 mins on a 670.
and a "percent blanked: 70.98" in ~32 min

edit} how was the high blanked outliers detected in version 6, because I have a Ver7 validated but not showing up in Application Details, "Number of tasks completed"

I know not all valid tasks get counted as "completed", but I had done nearly 50 valid non 30/30 early exit tasks with AP v6 before I had 11 "completed" on one system. Since it was just one system I had just ignored it, but perhaps there is something else going on.
SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[
ID: 1584324 · Report as offensive
Profile Zeus Fab3r
Avatar

Send message
Joined: 17 Jan 01
Posts: 649
Credit: 275,335,635
RAC: 597
Serbia
Message 1584328 - Posted: 10 Oct 2014, 2:07:28 UTC - in response to Message 1584323.  

Sorry, are you running stock?


No, I just switched to ap7_win_x86_sse2_opencl_nv_r2721
Here is portion of app_info:
    <app>
    <name>astropulse_v7</name>
    </app>
    <file_info>
    <name>AP7_win_x86_SSE2_OpenCL_NV_r2721.exe</name>
    <executable/>
    </file_info>
    <file_info>
    <name>libfftw3f-3-3-4_x86.dll</name>
    <executable/>
    </file_info>
    <file_info>
    <name>ap_cmdline_win_x86_SSE2_OpenCL_NV.txt</name>
    </file_info>
    <file_info>
    <name>AstroPulse_Kernels_r2721.cl</name>
    </file_info>
    <app_version>
    <app_name>astropulse_v7</app_name>
    <version_num>705</version_num>
    <flops>855000000000</flops>
    <platform>windows_intelx86</platform>
    <avg_ncpus>0.04</avg_ncpus>
    <max_ncpus>0.2</max_ncpus>
    <plan_class>cuda_opencl_100</plan_class>
    <cmdline></cmdline>
    <coproc>
    <type>CUDA</type>
    <count>1</count>
    </coproc>
    <file_ref>
    <file_name>AP7_win_x86_SSE2_OpenCL_NV_r2721.exe</file_name>
    <main_program/>
    </file_ref>
    <file_ref>
    <file_name>libfftw3f-3-3-4_x86.dll</file_name>
    </file_ref>
    <file_ref>
    <file_name>ap_cmdline_win_x86_SSE2_OpenCL_NV.txt</file_name>
    <open_name>ap_cmdline.txt</open_name>
    </file_ref>
    <file_ref>
    <file_name>AstroPulse_Kernels_r2721.cl</file_name>
    </file_ref>
    </app_version>
    <app_version>
    <app_name>astropulse_v7</app_name>
    <version_num>705</version_num>
    <flops>855000000000</flops>
    <platform>windows_x86_64</platform>
    <avg_ncpus>0.04</avg_ncpus>
    <max_ncpus>0.2</max_ncpus>
    <plan_class>cuda_opencl_100</plan_class>
    <cmdline></cmdline>
    <coproc>
    <type>CUDA</type>
    <count>1</count>
    </coproc>
    <file_ref>
    <file_name>AP7_win_x86_SSE2_OpenCL_NV_r2721.exe</file_name>
    <main_program/>
    </file_ref>
    <file_ref>
    <file_name>libfftw3f-3-3-4_x86.dll</file_name>
    </file_ref>
    <file_ref>
    <file_name>ap_cmdline_win_x86_SSE2_OpenCL_NV.txt</file_name>
    <open_name>ap_cmdline.txt</open_name>
    </file_ref>
    <file_ref>
    <file_name>AstroPulse_Kernels_r2721.cl</file_name>
    </file_ref>
    </app_version>



Who the hell is General Failure and why is he reading my harddisk?¿
ID: 1584328 · Report as offensive
Profile Zalster Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 27 May 99
Posts: 5517
Credit: 528,817,460
RAC: 242
United States
Message 1584331 - Posted: 10 Oct 2014, 2:09:32 UTC - in response to Message 1584328.  
Last modified: 10 Oct 2014, 2:15:22 UTC

Next question.. why did you get v7 instead of v7.05? v7 is for CPU..That's unusual


Edit...
Could it be the naming of the AP in your app_info? Or did you try to transfer a CPU work unit to the GPU? Just throwing some ideas out there...
ID: 1584331 · Report as offensive
Profile Zeus Fab3r
Avatar

Send message
Joined: 17 Jan 01
Posts: 649
Credit: 275,335,635
RAC: 597
Serbia
Message 1584339 - Posted: 10 Oct 2014, 2:20:37 UTC - in response to Message 1584331.  
Last modified: 10 Oct 2014, 2:28:49 UTC

Next question.. why did you get v7 instead of v7.05? v7 is for CPU..That's unusual


Edit...
Could it be the naming of the AP in your app_info? Or did you try to transfer a CPU work unit to the GPU? Just throwing some ideas out there...


I didn't mess with anything like that, knowing how sensitive new apps can be. I just downloaded nv app from Mike's link put files in place and modified app_info.

Edit: Strange how simple solutions can make wonders. Simple restart did the thing and now everything is fine. thx anyway!

Who the hell is General Failure and why is he reading my harddisk?¿
ID: 1584339 · Report as offensive
Previous · 1 · 2 · 3 · 4 · 5 · 6 . . . 20 · Next

Message boards : Number crunching : AP V7


 
©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.