What do you think about moving the 4670 to the lower slot and installing the Stock 7.0.62? It actually works better when the 4670 is in the lower slot due to it being further from the heat of the processor and the airflow from the front case fan is directed onto the 4670 heatsink better.

This doesn't seem to work on my XP Host either. I tried;
<device_nums>0</device_nums> & <device_nums>0 0</device_nums>

<coproc>
<type>ati_opencl_100</type>
<count>1</count>
<device_nums>0</device_nums>
</coproc>

Both times I received;

4/17/2013 3:06:02 AM |  | Starting BOINC client version 7.0.62 for windows_intelx86
4/17/2013 3:06:02 AM |  | OS: Microsoft Windows XP: Professional x86 Edition, Service Pack 3, (05.01.2600.00)
4/17/2013 3:06:02 AM |  | CUDA: NVIDIA GPU 0: GeForce 8800 GT (driver version 266.58, CUDA version 3.20, compute capability 1.1, 512MB, 468MB available, 504 GFLOPS peak)
4/17/2013 3:06:02 AM |  | CAL: ATI GPU 0: AMD Radeon HD 6790/6850/6870 series (Barts) (CAL version 1.4.1646, 1024MB, 989MB available, 2976 GFLOPS peak)
4/17/2013 3:06:02 AM |  | OpenCL: NVIDIA GPU 0: GeForce 8800 GT (driver version 266.58, device version OpenCL 1.0 CUDA, 512MB, 468MB available, 504 GFLOPS peak)
4/17/2013 3:06:02 AM |  | OpenCL: AMD/ATI GPU 0: AMD Radeon HD 6790/6850/6870 series (Barts) (driver version CAL 1.4.1646, device version OpenCL 1.1 AMD-APP (831.4), 1024MB, 989MB available, 2976 GFLOPS peak)
4/17/2013 3:06:02 AM | SETI@home | Found app_info.xml; using anonymous platform
4/17/2013 3:06:02 AM |  | app version refers to missing GPU type ati_opencl_100
4/17/2013 3:06:02 AM | SETI@home | Missing coprocessor for task ap_24ja13ad_B5_P0_00015_20130414_19002.wu_1
4/17/2013 3:06:02 AM | SETI@home | Missing coprocessor for task ap_25fe13aa_B1_P1_00312_20130414_20647.wu_0
.
.

?

Have you changed the according app_info.xml entrys as mentioned?

---

- Performance is not a simple linear function of the number of CPUs you throw at the problem. -

First of all, the GPU type is ATI. I think it should be

<type>ATI</type>

Second, do the app_info.xml files for the Brook app and the OpenCL app both contain this?

<name>astropulse_v6</name>

If you want to distinguish between them, they must have different names. I think this is where the 'ati_opencl_100' and 'ati_no_opencl' names come into play, though I would suggest instead the names

<name>astropulse_v6_opencl</name>

and

<name>astropulse_v6_brook</name>

However, I don't see any place to put the name of the application you wish to run on that GPU in the description of the cc_config.xml <coproc> option.

I have written to Dr. Anderson asking for clarification. I'll post when I hear back.

---

Charlie Fenton
BOINC / SETI@home Macintosh & Windows Programmer

First of all, the GPU type is ATI. I think it should be

<type>ATI</type>

Second, do the app_info.xml files for the Brook app and the OpenCL app both contain this?

<name>astropulse_v6</name>

If you want to distinguish between them, they must have different names. I think this is where the 'ati_opencl_100' and 'ati_no_opencl' names come into play, though I would suggest instead the names

<name>astropulse_v6_opencl</name>

and

<name>astropulse_v6_brook</name>

However, I don't see any place to put the name of the application you wish to run on that GPU in the description of the cc_config.xml <coproc> option.

I have written to Dr. Anderson asking for clarification. I'll post when I hear back.

Even if he get's it working client side, ie gets AP Wu's, some being for each app_version, there is only app_version server side, ie 'AstroPulse v6 (anonymous platform, ATI GPU)' and only one APR,
the OpenCL AP app is almost 100% a GPU app, while the Hybrid AP app is almost 100% a CPU app with a small GPU acceleration, i expect each Hybrid AP Wu will end up maximum time exeeded.

Claggy

I think we have to be very careful with our terminology here.

My understanding is that the "application" entity defined by the <name> tags is the broad data processing class - thus, looking at the SETI Applications page (and converting from 'friendly' to 'short' format), there are currently just four applications defined here:

setiathome_enhanced (current - also called 'MultiBeam', in common parlance)
astropulse_v505 (deprecated - no new work being generated)
setiathome_v7 (work in progress - not yet deployed)
astropulse_v6 (current)

The <app><name> tag in app_info.xml has to be one of those two/four, otherwise scheduler requests can't specify what type of task the client wishes to be allocated (and has the tools to process).

Then, the subsequent <app_name> tag in <app_version> has to match the <app><name>. I don't see any way of shoe-horning 'astropulse_v6_opencl' or 'astropulse_v6_brook' in at this level.

Unfortunately, we also have the common plain-english usage of the word 'application' to describe a binary executable file (or collection of files) - I think this leads to some wishful thinking in discussions like this.

The same applies to the usage of appname in an <exclude_gpu> block in cc_config.xml, and <app><name> in app_config.xml: I don't think either construct allows sufficient granularity to achieve what you're trying to do here.

The only places I can see where an actual program filename is supported are <exclusive_app> and <exclusive_gpu_app>.

First of all, the GPU type is ATI. I think it should be

<type>ATI</type>

Second, do the app_info.xml files for the Brook app and the OpenCL app both contain this?

<name>astropulse_v6</name>

If you want to distinguish between them, they must have different names. I think this is where the 'ati_opencl_100' and 'ati_no_opencl' names come into play, though I would suggest instead the names

<name>astropulse_v6_opencl</name>

and

<name>astropulse_v6_brook</name>

However, I don't see any place to put the name of the application you wish to run on that GPU in the description of the cc_config.xml <coproc> option.

I have written to Dr. Anderson asking for clarification. I'll post when I hear back.

Even if he get's it working client side, ie gets AP Wu's, some being for each app_version, there is only app_version server side, ie 'AstroPulse v6 (anonymous platform, ATI GPU)' and only one APR,
the OpenCL AP app is almost 100% a GPU app, while the Hybrid AP app is almost 100% a CPU app with a small GPU acceleration, i expect each Hybrid AP Wu will end up maximum time exeeded.

Claggy

One of the first requirements I came across was that the App name <name>astropulse_v6</name> HAS to be exact, down to the case setting. You can't change that entry. I don't have any CAL tasks, I think that's still in beta. I would think to distinguish between OpenCL & CAL astropulse_v6's you'd rely on the <plan_class></plan_class> just as you do with ati_opencl_100 and opencl_nvidia_100. I have seen a class cal_ati in beta, since the 4670 uses class ati_opencl_100 and the 3650 would use cal_ati, I don't see why there would be a problem with maximum time exeeded. On my Host, the GPU & CPU times are much closer together than that other example. You can change the plan_class to anything you want, I have downloaded astropulse_v6 plan_class 'code_404' for my nVivda card from the server before, it was a test.

I'm still pondering the exact meaning of 'specify a GPU other than those detected and used by BOINC.' Has anyone been able to get this;
<coproc>
<type>some_name</type>
<count>1</count>
<device_nums>0 2</device_nums>
</coproc>
to work with a GPU that IS detected and used by BOINC?

I spoke with Dr. Anderson today, and we have determined that the use of the <coproc> option in cc_config.xml won't work with the past or current code for the boinc_get_opencl_ids() API, so cannot be used with the current astropulse_v6 anonymous platform apps. We will work on fixing this, but it might take a while to get to it.

---

Charlie Fenton
BOINC / SETI@home Macintosh & Windows Programmer

I finally obtained access to the Brook+ App with the New Lunatics installer. I'm unable to get it to work in Host 6796475. Every attempt results in BOINC trying to start an OpenCL Astropulse on the currently installed 2600XT while ignoring the Brook+ App. It appears the only way it might work is to not have mixed OpenCL/CAL Apps, as in, removing the 4670...

I finally obtained access to the Brook+ App with the New Lunatics installer. I'm unable to get it to work in Host 6796475. Every attempt results in BOINC trying to start an OpenCL Astropulse on the currently installed 2600XT while ignoring the Brook+ App. It appears the only way it might work is to not have mixed OpenCL/CAL Apps, as in, removing the 4670...

Tricky situation if you want to run OpenCL AP on one and Brook AP on the other.
Only way around that I know of is to use two running boinc clients and disable one card in each.
With newer boinc you can use -allow_multiple_clients - that way you only need two data directories, not two program folders as well.

---

A person who won't read has no advantage over one who can't read. (Mark Twain)

I finally obtained access to the Brook+ App with the New Lunatics installer. I'm unable to get it to work in Host 6796475. Every attempt results in BOINC trying to start an OpenCL Astropulse on the currently installed 2600XT while ignoring the Brook+ App. It appears the only way it might work is to not have mixed OpenCL/CAL Apps, as in, removing the 4670...

Tricky situation if you want to run OpenCL AP on one and Brook AP on the other.
Only way around that I know of is to use two running boinc clients and disable one card in each.
With newer boinc you can use -allow_multiple_clients - that way you only need two data directories, not two program folders as well.

I found the <allow_multiple_clients>0|1</allow_multiple_clients> part, where are the instructions for setting up a second data folder?
I was looking here, GPU-load went to 100%... and noticed the 'Memory Usage 311 MB' with 2 instances. My 2600 only has 256 MB, so, I guess I can't run 2 on the 2600? The 3650 has 1024 MB, but, I was thinking about using the 3650 on another Host. Most of the Brook+ Run-times I've seen are much longer than my current CPU AP Times, I'm assuming my Brook+ times will be better than my current CPU AP Times? If not, I might as well stop now...
Thanks.

Since the CPU times are close to the 4670 times, would I be better off to just replace the 4670 with the 3650 and run the 2600 & 3560 together? How would that work out? Then I could place the 4670 in the other host...

Oh joy, more fun.

6/6/2013 2:06:53 AM |  | Starting BOINC client version 7.0.64 for windows_intelx86
6/6/2013 2:06:53 AM |  | Data directory: C:\ProgramData\BOINC
6/6/2013 2:06:53 AM |  | Processor: 2 GenuineIntel Pentium(R) Dual-Core  CPU      E5200  @ 2.50GHz [Family 6 Model 23 Stepping 6]
6/6/2013 2:06:53 AM |  | OS: Microsoft Windows 8: Professional with Media Center x86 Edition, (06.02.9200.00)
6/6/2013 2:06:53 AM |  | CAL: ATI GPU 0: ATI Radeon HD 2600/3650 (RV630/RV635) (CAL version 1.4.1734, 1024MB, 992MB available, 348 GFLOPS peak)
6/6/2013 2:06:53 AM |  | CAL: ATI GPU 1: ATI Radeon HD 2600/3650 (RV630/RV635) (CAL version 1.4.1734, 256MB, 224MB available, 336 GFLOPS peak)
6/6/2013 2:06:53 AM | SETI@home | Found app_info.xml; using anonymous platform
6/6/2013 2:06:53 AM |  | App version needs OpenCL but GPU doesn't support it
6/6/2013 2:06:53 AM |  | App version needs OpenCL but GPU doesn't support it
6/6/2013 2:06:53 AM | SETI@home | Missing coprocessor for task ap_12mr10ab_B5_P1_00073_20130531_01409.wu_1
6/6/2013 2:06:53 AM | SETI@home | Missing coprocessor for task ap_12mr10ab_B6_P0_00149_20130531_03801.wu_0
6/6/2013 2:06:53 AM | SETI@home | Missing coprocessor for task ap_12mr10ab_B6_P1_00149_20130531_03849.wu_0
6/6/2013 2:06:53 AM | SETI@home | Missing coprocessor for task ap_18mr10ad_B0_P0_00165_20130531_14358.wu_1
6/6/2013 2:06:53 AM | SETI@home | Missing coprocessor for task ap_12mr10ab_B5_P1_00149_20130531_01409.wu_0
6/6/2013 2:06:53 AM | SETI@home | Missing coprocessor for task ap_17mr10aa_B1_P0_00392_20130531_20221.wu_0
6/6/2013 2:06:53 AM | SETI@home | Missing coprocessor for task ap_12mr10ab_B6_P0_00185_20130531_03801.wu_0
6/6/2013 2:06:53 AM | SETI@home | Missing coprocessor for task ap_18mr10ad_B0_P0_00187_20130531_14358.wu_0
6/6/2013 2:06:53 AM | SETI@home | Missing coprocessor for task ap_12mr10ab_B6_P1_00222_20130531_03849.wu_1
6/6/2013 2:06:53 AM | SETI@home | Missing coprocessor for task ap_18mr10aa_B0_P1_00161_20130531_19221.wu_1
6/6/2013 2:06:53 AM | SETI@home | Missing coprocessor for task ap_17mr10aa_B2_P0_00055_20130531_23620.wu_0
6/6/2013 2:06:53 AM | SETI@home | Missing coprocessor for task ap_18mr10ad_B0_P0_00385_20130531_14358.wu_1
6/6/2013 2:06:53 AM | SETI@home | Missing coprocessor for task ap_17mr10aa_B2_P0_00016_20130531_23620.wu_2
6/6/2013 2:06:53 AM | SETI@home | Missing coprocessor for task ap_17mr10af_B1_P0_00278_20130531_11060.wu_1
6/6/2013 2:06:53 AM | SETI@home | Missing coprocessor for task ap_17mr10aa_B1_P1_00197_20130531_14949.wu_0
6/6/2013 2:06:53 AM | SETI@home | Missing coprocessor for task ap_17mr10aa_B2_P0_00137_20130531_23620.wu_2
6/6/2013 2:06:53 AM | SETI@home | Missing coprocessor for task ap_17mr10aa_B2_P0_00213_20130531_23620.wu_0
6/6/2013 2:06:53 AM | SETI@home | Missing coprocessor for task ap_17mr10aa_B1_P1_00350_20130531_14949.wu_0
6/6/2013 2:06:53 AM |  | Config: use all coprocessors
6/6/2013 2:06:53 AM | SETI@home | Restarting task ap_19mr10ab_B1_P1_00210_20130531_03972.wu_3 using astropulse_v6 version 601 in slot 1
6/6/2013 2:06:53 AM | SETI@home | Sending scheduler request: To fetch work.
6/6/2013 2:06:53 AM | SETI@home | Requesting new tasks for ATI
6/6/2013 2:06:56 AM | SETI@home | Scheduler request completed: got 6 new tasks
6/6/2013 2:06:56 AM | SETI@home | [error] Missing coprocessor for task ap_13my10ab_B6_P0_00075_20130605_23717.wu_0; aborting
6/6/2013 2:06:56 AM | SETI@home | [error] Missing coprocessor for task ap_13my10ab_B6_P0_00076_20130605_23717.wu_0; aborting
6/6/2013 2:06:56 AM | SETI@home | [error] Missing coprocessor for task ap_14my10aa_B3_P1_00395_20130605_27581.wu_0; aborting
6/6/2013 2:06:56 AM | SETI@home | [error] Missing coprocessor for task ap_17my10ae_B0_P1_00273_20130605_00769.wu_0; aborting
6/6/2013 2:06:56 AM | SETI@home | [error] Missing coprocessor for task ap_24ap10ac_B4_P0_00007_20130603_01335.wu_4; aborting
6/6/2013 2:06:56 AM | SETI@home | [error] Missing coprocessor for task ap_14my10aa_B3_P1_00394_20130605_27581.wu_1; aborting
6/6/2013 2:06:58 AM | SETI@home | Started download of ap_13my10ab_B6_P0_00075_20130605_23717.wu
6/6/2013 2:06:58 AM | SETI@home | Started download of ap_13my10ab_B6_P0_00076_20130605_23717.wu
6/6/2013 2:06:58 AM | SETI@home | Started download of ap_14my10aa_B3_P1_00395_20130605_27581.wu
6/6/2013 2:06:58 AM | SETI@home | Started download of ap_17my10ae_B0_P1_00273_20130605_00769.wu
6/6/2013 2:06:58 AM | SETI@home | Started download of ap_24ap10ac_B4_P0_00007_20130603_01335.wu
6/6/2013 2:06:58 AM | SETI@home | Started download of ap_14my10aa_B3_P1_00394_20130605_27581.wu
6/6/2013 2:07:10 AM | SETI@home | Finished download of ap_14my10aa_B3_P1_00395_20130605_27581.wu
6/6/2013 2:12:00 AM | SETI@home | Sending scheduler request: To fetch work.
6/6/2013 2:12:00 AM | SETI@home | Reporting 6 completed tasks
6/6/2013 2:12:00 AM | SETI@home | Requesting new tasks for ATI
6/6/2013 2:12:02 AM | SETI@home | Scheduler request completed: got 4 new tasks
6/6/2013 2:12:02 AM | SETI@home | [error] Missing coprocessor for task ap_12my10aa_B4_P0_00216_20130605_32761.wu_2; aborting
6/6/2013 2:12:02 AM | SETI@home | [error] Missing coprocessor for task ap_20my10aa_B2_P0_00266_20130604_28855.wu_4; aborting
6/6/2013 2:12:02 AM | SETI@home | [error] Missing coprocessor for task ap_20my10aa_B5_P0_00216_20130605_10477.wu_2; aborting
6/6/2013 2:12:02 AM | SETI@home | [error] Missing coprocessor for task ap_08my10ad_B6_P1_00238_20130605_32515.wu_2; aborting
6/6/2013 2:12:04 AM | SETI@home | Started download of ap_12my10aa_B4_P0_00216_20130605_32761.wu....

So, why does Brook+ insist it requires OpenCL?

I used <version_num>604</version_num> <plan_class>ati_opencl_100</plan_class> because that's what the old files were classed.

<app_info> 
    <app>
      <name>astropulse_v6</name>
    </app>
    <file_info>
        <name>AP6_win_x86_SSE_Brook_r1854.exe</name>
        <executable/>
    </file_info>
    <file_info>
        <name>brook.dll</name>
        <executable/>
    </file_info>
    <file_info>
        <name>brook_cal.dll</name>
        <executable/>
    </file_info>
    <app_version>
        <app_name>astropulse_v6</app_name>
        <version_num>604</version_num>
        <plan_class>ati_opencl_100</plan_class>
        <platform>windows_intelx86</platform>
        <avg_ncpus>0.5</avg_ncpus>
        <max_ncpus>0.5</max_ncpus>
        <flops>36000000000</flops>
        <coproc>
            <type>ATI</type>
            <count>1</count>
        </coproc>
        <file_ref>
            <file_name>AP6_win_x86_SSE_Brook_r1854.exe</file_name>
            <main_program/>
        </file_ref>
        <file_ref>
            <file_name>brook.dll</file_name>
        </file_ref>
        <file_ref>
            <file_name>brook_cal.dll</file_name>
        </file_ref>
    </app_version>
  <app>
    <name>astropulse_v6</name>
 </app>
    <file_info>
       <name>ap_6.01r557_SSE2_331_AVX.exe</name>
        <executable/>
    </file_info>
    <app_version>
        <app_name>astropulse_v6</app_name>
        <version_num>601</version_num>
        <platform>windows_intelx86</platform>
        <flops>38000000000</flops>
        <file_ref>
            <file_name>ap_6.01r557_SSE2_331_AVX.exe</file_name>
	  <main_program/>
	</file_ref>
    </app_version>
</app_info>

Why is it Downloading new Files just so it can delete them?

Great. I decided to remove the app_info file and reset the project. Now I have;

6/6/2013 2:59:51 AM |  | Starting BOINC client version 7.0.64 for windows_intelx86
6/6/2013 2:59:51 AM |  | OS: Microsoft Windows 8: Professional with Media Center x86 Edition, (06.02.9200.00)
6/6/2013 2:59:51 AM |  | CAL: ATI GPU 0: ATI Radeon HD 2600/3650 (RV630/RV635) (CAL version 1.4.1734, 1024MB, 992MB available, 348 GFLOPS peak)
6/6/2013 2:59:51 AM |  | CAL: ATI GPU 1: ATI Radeon HD 2600/3650 (RV630/RV635) (CAL version 1.4.1734, 256MB, 224MB available, 336 GFLOPS peak)
6/6/2013 2:59:51 AM |  | Config: use all coprocessors
6/6/2013 2:59:51 AM | SETI@home | URL http://setiathome.berkeley.edu/; Computer ID 6796475; resource share 100
6/6/2013 2:59:51 AM | SETI@home | General prefs: from SETI@home (last modified 04-May-2013 22:44:16)
6/6/2013 2:59:51 AM | SETI@home | Computer location: school
6/6/2013 2:59:51 AM |  | General prefs: using separate prefs for school
6/6/2013 2:59:51 AM |  | Reading preferences override file
6/6/2013 3:01:13 AM |  | Suspending network activity - user request
6/6/2013 3:01:20 AM |  | Resuming network activity
6/6/2013 3:01:38 AM | SETI@home | update requested by user
6/6/2013 3:01:40 AM | SETI@home | Sending scheduler request: Requested by user.
6/6/2013 3:01:40 AM | SETI@home | Not requesting tasks: some download is stalled
6/6/2013 3:01:43 AM | SETI@home | Scheduler request completed
6/6/2013 3:01:57 AM |  | [error] Can't create HTTP response output file projects/setiathome.berkeley.edu/astropulse_6.01_windows_intelx86.exe
6/6/2013 3:01:57 AM | SETI@home | Backing off 15 min 35 sec on download of astropulse_6.01_windows_intelx86.exe
6/6/2013 3:02:03 AM |  | [error] Can't create HTTP response output file projects/setiathome.berkeley.edu/ap_graphics_6.01_windows_intelx86.exe
6/6/2013 3:02:03 AM | SETI@home | Backing off 8 min 19 sec on download of ap_graphics_6.01_windows_intelx86.exe
6/6/2013 3:02:09 AM |  | [error] Can't create HTTP response output file projects/setiathome.berkeley.edu/libfftw3f-3-1-1a_upx.dll
6/6/2013 3:02:09 AM | SETI@home | Backing off 14 min 12 sec on download of libfftw3f-3-1-1a_upx.dll
6/6/2013 3:02:41 AM |  | [error] Can't create HTTP response output file projects/setiathome.berkeley.edu/astropulse_6.01_windows_intelx86.exe
6/6/2013 3:02:41 AM | SETI@home | Backing off 20 min 31 sec on download of astropulse_6.01_windows_intelx86.exe
6/6/2013 3:02:41 AM |  | [error] Can't create HTTP response output file projects/setiathome.berkeley.edu/ap_graphics_6.01_windows_intelx86.exe
6/6/2013 3:02:41 AM | SETI@home | Backing off 30 min 3 sec on download of ap_graphics_6.01_windows_intelx86.exe
6/6/2013 3:02:41 AM |  | [error] Can't create HTTP response output file projects/setiathome.berkeley.edu/libfftw3f-3-1-1a_upx.dll
6/6/2013 3:02:41 AM | SETI@home | Backing off 19 min 3 sec on download of libfftw3f-3-1-1a_upx.dll...

This has been a perfectly working system since Oct, now it can't even download a SETI file.

I thought that was fixed?


Oh, I see, the Project folder is Missing. How can it download a file to a Missing Folder? Why is the Folder Missing...

Detached and got the Folder, DLed the stock AP CPU App, quit and ran the New Lunatics installer for AP CPU & Brook.

6/6/2013 3:32:17 AM |  | Starting BOINC client version 7.0.64 for windows_intelx86
6/6/2013 3:32:17 AM |  | log flags: file_xfer, sched_ops, task
6/6/2013 3:32:17 AM |  | Libraries: libcurl/7.25.0 OpenSSL/1.0.1 zlib/1.2.6
6/6/2013 3:32:17 AM |  | Data directory: C:\ProgramData\BOINC
6/6/2013 3:32:17 AM |  | Processor: 2 GenuineIntel Pentium(R) Dual-Core  CPU      E5200  @ 2.50GHz [Family 6 Model 23 Stepping 6]
6/6/2013 3:32:17 AM |  | Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss htt tm pni ssse3 cx16 nx lm tm2 pbe
6/6/2013 3:32:17 AM |  | OS: Microsoft Windows 8: Professional with Media Center x86 Edition, (06.02.9200.00)
6/6/2013 3:32:17 AM |  | CAL: ATI GPU 0: ATI Radeon HD 2600/3650 (RV630/RV635) (CAL version 1.4.1734, 1024MB, 992MB available, 348 GFLOPS peak)
6/6/2013 3:32:17 AM |  | CAL: ATI GPU 1: ATI Radeon HD 2600/3650 (RV630/RV635) (CAL version 1.4.1734, 256MB, 224MB available, 336 GFLOPS peak)
6/6/2013 3:32:17 AM | SETI@home | Found app_info.xml; using anonymous platform
6/6/2013 3:32:17 AM |  | App version needs OpenCL but GPU doesn't support it
6/6/2013 3:32:17 AM |  | Config: use all coprocessors
6/6/2013 3:32:17 AM | SETI@home | URL http://setiathome.berkeley.edu/; Computer ID 6796475; resource share 100
6/6/2013 3:32:17 AM | SETI@home | General prefs: from SETI@home (last modified 04-May-2013 22:44:16)
6/6/2013 3:32:17 AM | SETI@home | Computer location: school
6/6/2013 3:32:17 AM |  | General prefs: using separate prefs for school
6/6/2013 3:32:17 AM |  | Reading preferences override file
6/6/2013 3:32:17 AM |  | File projects/setiathome.berkeley.edu/astropulse_6.01_windows_intelx86.exe not found
6/6/2013 3:32:17 AM |  | File projects/setiathome.berkeley.edu/ap_graphics_6.01_windows_intelx86.exe not found
6/6/2013 3:32:17 AM |  | File projects/setiathome.berkeley.edu/libfftw3f-3-1-1a_upx.dll not found
6/6/2013 3:32:17 AM |  | Not using a proxy
6/6/2013 3:32:17 AM | SETI@home | Restarting task ap_23ap11ab_B1_P1_00169_20130511_17897.wu_3 using astropulse_v6 version 601 in slot 0
6/6/2013 3:32:17 AM | SETI@home | Sending scheduler request: To fetch work.
6/6/2013 3:32:17 AM | SETI@home | Requesting new tasks for ATI
6/6/2013 3:32:18 AM | SETI@home | Scheduler request completed: got 3 new tasks
6/6/2013 3:32:18 AM | SETI@home | [error] Missing coprocessor for task ap_14my10aa_B5_P0_00034_20130605_29681.wu_0; aborting
6/6/2013 3:32:18 AM | SETI@home | [error] Missing coprocessor for task ap_14my10aa_B5_P0_00032_20130605_29681.wu_0; aborting
6/6/2013 3:32:18 AM | SETI@home | [error] Missing coprocessor for task ap_14my10aa_B4_P1_00033_20130605_29680.wu_1; aborting
6/6/2013 3:32:20 AM | SETI@home | Started download of ap_14my10aa_B5_P0_00034_20130605_29681.wu
6/6/2013 3:32:20 AM | SETI@home | Started download of ap_14my10aa_B5_P0_00032_20130605_29681.wu
6/6/2013 3:32:26 AM | SETI@home | Finished download of ap_14my10aa_B5_P0_00032_20130605_29681.wu
6/6/2013 3:32:26 AM | SETI@home | Started download of ap_14my10aa_B4_P1_00033_20130605_29680.wu
6/6/2013 3:32:29 AM | SETI@home | Finished download of ap_14my10aa_B5_P0_00034_20130605_29681.wu
6/6/2013 3:32:30 AM | SETI@home | Finished download of ap_14my10aa_B4_P1_00033_20130605_29680.wu
6/6/2013 3:32:57 AM |  | Suspending GPU computation - user request
6/6/2013 3:33:01 AM |  | Suspending network activity - user request...

Same, Same.

Just a note. Remember how GPUz was reporting the 3650 & 2600 as having OpenCL back here;



Since removing the 4670, Both cards are now reported by GPUz as NOT having OpenCL. If you are testing this App with an OpenCL GPU installed alongside the CAL GPU, you might not be receiving Real World Experiences...

Um. I think you need Raistmer...

As to getting a second client with second data dir running:

From a command prompt:

cd to the boinc PROGRAM folder.

boinc --allow_multiple_clients --gui_rpc_port <port> --dir <new data folder> --detach_console

boinc will create and populate a new data folder in the specified dir
In that dir you will find a file named gui_rpc_auth.cfg either edit the contents to a pwd you can remember or just c/p.
Now you need BoincTasks - BoincManager can't attach to a second localhost at a different port ::) In BoincTasks open another host and give it the portnumber you gave further up and the password you took/copied.

I suppose you can copy the project folder over from the main data dir and also config etc. edit app_info.xml for that other app.
Don't forget to attach to seti ;)

A bit bare, I'm mainly working from memory.


We may have botched the brook aistub - we couldn't actually test that bit.
Putting to Raistmer for review.

---

A person who won't read has no advantage over one who can't read. (Mark Twain)

ok please try this updated aistub:

    <app>
        <name>astropulse_v6</name>
    </app>
    <file_info>
        <name>AP6_win_x86_SSE_Brook_r1854.exe</name>
        <executable/>
    </file_info>
    <file_info>
        <name>brook.dll</name>
        <executable/>
    </file_info>
    <file_info>
        <name>brook_cal.dll</name>
        <executable/>
    </file_info>

    <app_version>
        <app_name>astropulse_v6</app_name>
        <version_num>606</version_num>
        <avg_ncpus>0.8</avg_ncpus>
        <max_ncpus>0.8</max_ncpus>
        <plan_class>ati13ati</plan_class>
        <coproc>
            <type>ATI</type>
            <count>1</count>
        </coproc>
        <file_ref>
            <file_name>AP6_win_x86_SSE_Brook_r1854.exe</file_name>
            <main_program/>
        </file_ref>
        <file_ref>
            <file_name>brook.dll</file_name>
        </file_ref>
        <file_ref>
            <file_name>brook_cal.dll</file_name>
        </file_ref>
    </app_version>

If you have any work in progress, check the version numbers match.

edit: additionally Richard found a dependency on libfftw3f-3.dll which we failed to reference and which the installer wouldn't put in on a single install. Make sure you've got that dll in the project folder as well.

---

A person who won't read has no advantage over one who can't read. (Mark Twain)

That seems to work, to a point. Now I'm getting the old "Task ap_14my10aa_B4_P1_00033_20130605_29680.wu_1 exited with zero status but no 'finished' file If this happens repeatedly you may need to reset the project." Memory usage is less than 100 MB on both cards, so, I don't think it's memory. The CPU App must have called for the libfftw3f-3.dll file, because it's in the folder.

6/6/2013 7:38:41 AM |  | CAL: ATI GPU 0: ATI Radeon HD 2600/3650 (RV630/RV635) (CAL version 1.4.1734, 1024MB, 992MB available, 348 GFLOPS peak)
6/6/2013 7:38:41 AM |  | CAL: ATI GPU 1: ATI Radeon HD 2600/3650 (RV630/RV635) (CAL version 1.4.1734, 256MB, 224MB available, 336 GFLOPS peak)
6/6/2013 7:38:41 AM | SETI@home | Found app_info.xml; using anonymous platform
6/6/2013 7:38:41 AM |  | App version needs OpenCL but GPU doesn't support it
6/6/2013 7:38:41 AM | SETI@home | [error] No application found for task: windows_intelx86 604 ati_opencl_100; discarding
6/6/2013 7:38:41 AM | SETI@home | [error] No application found for task: windows_intelx86 604 ati_opencl_100; discarding
6/6/2013 7:38:41 AM | SETI@home | [error] No application found for task: windows_intelx86 604 ati_opencl_100; discarding
6/6/2013 7:38:41 AM |  | Config: use all coprocessors
6/6/2013 7:38:41 AM | SETI@home | URL http://setiathome.berkeley.edu/; Computer ID 6796475; resource share 100
6/6/2013 7:38:41 AM | SETI@home | General prefs: from SETI@home (last modified 04-May-2013 22:44:16)
6/6/2013 7:38:41 AM | SETI@home | Computer location: school
6/6/2013 7:38:41 AM |  | General prefs: using separate prefs for school
6/6/2013 7:38:41 AM |  | Reading preferences override file
6/6/2013 7:38:41 AM |  | Preferences:
6/6/2013 7:38:41 AM |  | max memory usage when active: 1023.58MB
6/6/2013 7:38:41 AM |  | max memory usage when idle: 1842.44MB
6/6/2013 7:38:41 AM |  | max disk usage: 20.00GB
6/6/2013 7:38:41 AM |  | suspend work if non-BOINC CPU load exceeds 50 %
6/6/2013 7:38:41 AM |  | (to change preferences, visit a project web site or select Preferences in the Manager)
6/6/2013 7:38:41 AM |  | Not using a proxy
6/6/2013 7:38:42 AM |  | Suspending GPU computation - user request
6/6/2013 7:38:42 AM |  | Suspending network activity - user request
6/6/2013 7:38:42 AM | SETI@home | Restarting task ap_23ap11ab_B1_P1_00169_20130511_17897.wu_3 using astropulse_v6 version 601 in slot 0
6/6/2013 7:38:42 AM | SETI@home | Restarting task ap_20my10aa_B4_P1_00283_20130605_03161.wu_3 using astropulse_v6 version 601 in slot 1
6/6/2013 7:39:17 AM | SETI@home | General prefs: from SETI@home (last modified 04-May-2013 22:44:16)
6/6/2013 7:39:17 AM | SETI@home | Computer location: school
6/6/2013 7:39:17 AM |  | General prefs: using separate prefs for school
6/6/2013 7:39:17 AM |  | Reading preferences override file
6/6/2013 7:39:17 AM |  | Preferences:
6/6/2013 7:39:17 AM |  | max memory usage when active: 1023.58MB
6/6/2013 7:39:17 AM |  | max memory usage when idle: 1842.44MB
6/6/2013 7:39:17 AM |  | max disk usage: 20.00GB
6/6/2013 7:39:17 AM |  | suspend work if non-BOINC CPU load exceeds 50 %
6/6/2013 7:39:17 AM |  | (to change preferences, visit a project web site or select Preferences in the Manager)
6/6/2013 7:39:30 AM |  | Resuming GPU computation
6/6/2013 7:39:35 AM |  | Resuming network activity
6/6/2013 7:39:35 AM | SETI@home | Sending scheduler request: To fetch work.
6/6/2013 7:39:35 AM | SETI@home | Requesting new tasks for ATI
6/6/2013 7:39:37 AM | SETI@home | Scheduler request completed: got 3 new tasks
6/6/2013 7:39:37 AM | SETI@home | Resent lost task ap_14my10aa_B5_P0_00032_20130605_29681.wu_0
6/6/2013 7:39:37 AM | SETI@home | Resent lost task ap_14my10aa_B4_P1_00033_20130605_29680.wu_1
6/6/2013 7:39:37 AM | SETI@home | Resent lost task ap_14my10aa_B5_P0_00034_20130605_29681.wu_0
6/6/2013 7:39:39 AM | SETI@home | Started download of ap_14my10aa_B5_P0_00032_20130605_29681.wu
6/6/2013 7:39:39 AM | SETI@home | Started download of ap_14my10aa_B4_P1_00033_20130605_29680.wu
6/6/2013 7:39:46 AM | SETI@home | Finished download of ap_14my10aa_B4_P1_00033_20130605_29680.wu
6/6/2013 7:39:46 AM | SETI@home | Started download of ap_14my10aa_B5_P0_00034_20130605_29681.wu
6/6/2013 7:39:46 AM | SETI@home | Starting task ap_14my10aa_B4_P1_00033_20130605_29680.wu_1 using astropulse_v6 version 606 (ati13ati) in slot 2
6/6/2013 7:39:50 AM | SETI@home | Finished download of ap_14my10aa_B5_P0_00032_20130605_29681.wu
6/6/2013 7:39:50 AM | SETI@home | Starting task ap_14my10aa_B5_P0_00032_20130605_29681.wu_0 using astropulse_v6 version 606 (ati13ati) in slot 3
6/6/2013 7:39:51 AM | SETI@home | Finished download of ap_14my10aa_B5_P0_00034_20130605_29681.wu
6/6/2013 7:41:43 AM | SETI@home | Task ap_14my10aa_B4_P1_00033_20130605_29680.wu_1 exited with zero status but no 'finished' file
6/6/2013 7:41:43 AM | SETI@home | If this happens repeatedly you may need to reset the project.
6/6/2013 7:41:43 AM | SETI@home | Restarting task ap_14my10aa_B4_P1_00033_20130605_29680.wu_1 using astropulse_v6 version 606 (ati13ati) in slot 2
6/6/2013 7:41:44 AM | SETI@home | Task ap_14my10aa_B5_P0_00032_20130605_29681.wu_0 exited with zero status but no 'finished' file
6/6/2013 7:41:44 AM | SETI@home | If this happens repeatedly you may need to reset the project.
6/6/2013 7:41:44 AM | SETI@home | Restarting task ap_14my10aa_B5_P0_00032_20130605_29681.wu_0 using astropulse_v6 version 606 (ati13ati) in slot 3
6/6/2013 7:42:16 AM | SETI@home | Task ap_14my10aa_B5_P0_00032_20130605_29681.wu_0 exited with zero status but no 'finished' file
6/6/2013 7:42:16 AM | SETI@home | If this happens repeatedly you may need to reset the project.
6/6/2013 7:42:16 AM | SETI@home | Restarting task ap_14my10aa_B5_P0_00032_20130605_29681.wu_0 using astropulse_v6 version 606 (ati13ati) in slot 3
6/6/2013 7:42:17 AM | SETI@home | Task ap_14my10aa_B4_P1_00033_20130605_29680.wu_1 exited with zero status but no 'finished' file
6/6/2013 7:42:17 AM | SETI@home | If this happens repeatedly you may need to reset the project.
6/6/2013 7:42:17 AM | SETI@home | Restarting task ap_14my10aa_B4_P1_00033_20130605_29680.wu_1 using astropulse_v6 version 606 (ati13ati) in slot 2
6/6/2013 7:42:49 AM | SETI@home | Task ap_14my10aa_B5_P0_00032_20130605_29681.wu_0 exited with zero status but no 'finished' file
6/6/2013 7:42:49 AM | SETI@home | If this happens repeatedly you may need to reset the project.
6/6/2013 7:42:49 AM | SETI@home | Restarting task ap_14my10aa_B5_P0_00032_20130605_29681.wu_0 using astropulse_v6 version 606 (ati13ati) in slot 3
6/6/2013 7:42:51 AM | SETI@home | Task ap_14my10aa_B4_P1_00033_20130605_29680.wu_1 exited with zero status but no 'finished' file
6/6/2013 7:42:51 AM | SETI@home | If this happens repeatedly you may need to reset the project.
6/6/2013 7:42:51 AM | SETI@home | Restarting task ap_14my10aa_B4_P1_00033_20130605_29680.wu_1 using astropulse_v6 version 606 (ati13ati) in slot 2
6/6/2013 7:43:21 AM | SETI@home | Task ap_14my10aa_B5_P0_00032_20130605_29681.wu_0 exited with zero status but no 'finished' file
6/6/2013 7:43:21 AM | SETI@home | If this happens repeatedly you may need to reset the project.
6/6/2013 7:43:21 AM | SETI@home | Restarting task ap_14my10aa_B5_P0_00032_20130605_29681.wu_0 using astropulse_v6 version 606 (ati13ati) in slot 3
6/6/2013 7:43:27 AM | SETI@home | Task ap_14my10aa_B4_P1_00033_20130605_29680.wu_1 exited with zero status but no 'finished' file
6/6/2013 7:43:27 AM | SETI@home | If this happens repeatedly you may need to reset the project.
6/6/2013 7:43:27 AM | SETI@home | Restarting task ap_14my10aa_B4_P1_00033_20130605_29680.wu_1 using astropulse_v6 version 606 (ati13ati) in slot 2
6/6/2013 7:43:37 AM |  | Suspending GPU computation - user request...

Then you have complete extended unresponsiveness with one CPU task running. You have to disable all CPU tasks to get it down to mildly unresponsive. I'll see how it goes for a little while.

----------------------------

Still getting the restarts with just 1 and 1 running. At least the screen 'lag' appears to be gone;

6/6/2013 8:18:46 AM |  | Starting BOINC client version 7.0.64 for windows_intelx86
6/6/2013 8:18:46 AM |  | Processor: 2 GenuineIntel Pentium(R) Dual-Core  CPU      E5200  @ 2.50GHz [Family 6 Model 23 Stepping 6]
6/6/2013 8:18:46 AM |  | OS: Microsoft Windows 8: Professional with Media Center x86 Edition
6/6/2013 8:18:46 AM |  | CAL: ATI GPU 0: ATI Radeon HD 2600/3650 (RV630/RV635) (CAL version 1.4.1734, 1024MB, 992MB available, 348 GFLOPS peak)
6/6/2013 8:18:46 AM |  | CAL: ATI GPU 1: (not used) ATI Radeon HD 2600/3650 (RV630/RV635) (CAL version 1.4.1734, 256MB, 224MB available, 336 GFLOPS peak)
6/6/2013 8:18:46 AM | SETI@home | Found app_info.xml; using anonymous platform
6/6/2013 8:18:46 AM |  | max CPUs used: 1
6/6/2013 8:18:46 AM | SETI@home | Restarting task ap_23ap11ab_B1_P1_00169_20130511_17897.wu_3 using astropulse_v6 version 601 in slot 0
6/6/2013 8:18:46 AM | SETI@home | Starting task ap_14my10aa_B4_P1_00033_20130605_29680.wu_1 using astropulse_v6 version 606 (ati13ati) in slot 2
6/6/2013 8:19:27 AM | SETI@home | Task ap_14my10aa_B4_P1_00033_20130605_29680.wu_1 exited with zero status but no 'finished' file
6/6/2013 8:19:27 AM | SETI@home | If this happens repeatedly you may need to reset the project.
6/6/2013 8:19:27 AM | SETI@home | Restarting task ap_14my10aa_B4_P1_00033_20130605_29680.wu_1 using astropulse_v6 version 606 (ati13ati) in slot 2
6/6/2013 8:20:08 AM | SETI@home | Task ap_14my10aa_B4_P1_00033_20130605_29680.wu_1 exited with zero status but no 'finished' file
6/6/2013 8:20:08 AM | SETI@home | If this happens repeatedly you may need to reset the project.
6/6/2013 8:20:08 AM | SETI@home | Restarting task ap_14my10aa_B4_P1_00033_20130605_29680.wu_1 using astropulse_v6 version 606 (ati13ati) in slot 2
6/6/2013 8:20:47 AM | SETI@home | Task ap_14my10aa_B4_P1_00033_20130605_29680.wu_1 exited with zero status but no 'finished' file
6/6/2013 8:20:47 AM | SETI@home | If this happens repeatedly you may need to reset the project.
6/6/2013 8:20:47 AM | SETI@home | Restarting task ap_14my10aa_B4_P1_00033_20130605_29680.wu_1 using astropulse_v6 version 606 (ati13ati) in slot 2
6/6/2013 8:20:54 AM |  | Suspending GPU computation - user request

Doesn't look good...

Ok, so you're still getting that 'needs OpenCL' msg - I'm putting that down to some check David is doing then - we know that we don't need OpenCL for brook.

David must be doing some checks on the plan_class we know nothing about (at least not without having a good scream at the code) if a simple plan class change gets it to work. (for a given value of 'work').

regarding unresponsiveness - worse than with the previous version? I've no idea (again) what Raistmer did to the code, but I suppose if the old one was better you could drop down to it.

---

A person who won't read has no advantage over one who can't read. (Mark Twain)

Ok, so you're still getting that 'needs OpenCL' msg - I'm putting that down to some check David is doing then - we know that we don't need OpenCL for brook.

David must be doing some checks on the plan_class we know nothing about (at least not without having a good scream at the code) if a simple plan class change gets it to work. (for a given value of 'work').

regarding unresponsiveness - worse than with the previous version? I've no idea (again) what Raistmer did to the code, but I suppose if the old one was better you could drop down to it.

The 'Requires OpenCL' went away when I changed the Plan Class. It had downloaded 3 of the 604 class files just as I quit the program, so, they were there when I restarted the program. It deleted those 3 604s, then downloaded and used the ati13 class. The program would be usable when running 1 CPU task and 1 Brook task, if it weren't for the restarts. Any idea why it's restarting the Brook task?