Posts by BilBg


log in
21) Questions and Answers : GPU applications : Running 2 WorkUnits on a GPU is using 2 cores to feed GPU. Is there a quick way to figure out if that's an improvement? (Message 1795666)
Posted 18 days ago by Profile BilBg
<cpu_usage>0.25</cpu_usage>
.....
It works!!! ...but it is using 1 core to feed each of the the GPU WUs.

It's not clear what you mean by "using" and what you mean by "it".
The <cpu_usage>0.25 tells BOINC to "free one core" (run one less CPU task) per 4 GPU tasks running.


BOINC sees 8 cores and I was letting it use 6 or 7

So you tell BOINC to "free cores" by 2 ways (which add-up)
- by Setting: e.g. "Use at most 99% of the CPUs" ("hard" Setting - will free 1 core always)
- by app_config.xml ("soft" Setting - (if <cpu_usage>0.25) will free 1 core only if 4 GPU tasks are running)

<cpu_usage> in app_config.xml don't tell anything to the app, because the app don't know what is app_config.xml (which is read only by BOINC)
22) Message boards : Number crunching : Anything relating to AstroPulse tasks (Message 1794595)
Posted 22 days ago by Profile BilBg
and kept hammering the retry button

If you use BoincTasks this hammering can be automated, I have this set at ~15 minutes (refresh_asap) in config.xml (in BoincTasks install directory, next to BoincTasks.exe)
<config> <refresh> <downloads> 1709 </downloads> <uploads> 1307 </uploads> <refresh> 3571 </refresh> <refresh_asap> 907 </refresh_asap> </refresh> </config>


Note:
For me <refresh> don't work (is ignored) when <refresh_asap> is present (on BoincTasks 1.61)
23) Questions and Answers : Windows : LOSS OF PERFORMANCE (Message 1794539)
Posted 22 days ago by Profile BilBg
You may just read "Average Credit Decreasing?":
http://setiathome.berkeley.edu/forum_thread.php?id=79418
24) Message boards : Number crunching : OpenCL MB v8.12 issues thread attempt 2 (Message 1794472)
Posted 23 days ago by Profile BilBg
Max compute units: 24

24 is shown for the TITAN X
For GeForce GTX 1080 is shown 20
Max compute units: 20 Max work group size: 1024 Max clock frequency: 1733Mhz Max memory allocation: 2147483648 Cache type: Read/Write Cache line size: 128 Cache size: 327680 Global memory size: 8589934592 Constant buffer size: 65536 Max number of constant args: 9 Local memory type: Scratchpad Local memory size: 49152 Queue properties: Out-of-Order: Yes Name: GeForce GTX 1080
25) Message boards : Number crunching : Optimizing SBS and Period Iterations for the Fury X (Message 1794418)
Posted 23 days ago by Profile BilBg
Any better recommendation for future posts?

I just click on your link:
https://goo.gl/photos/ToZxyJgxrxDFNGSa7

... then on Thumbnail, then Right-Click on image -> Open image in new tab
26) Message boards : Number crunching : Optimizing SBS and Period Iterations for the Fury X (Message 1794401)
Posted 23 days ago by Profile BilBg
You used some very long link for your image which don't work, here is your image:

27) Questions and Answers : GPU applications : Wrong information about GPU on Computer Info page (Message 1794386)
Posted 23 days ago by Profile BilBg
I think this is only cosmetic - to simplify the web pages only the "best" GPU is shown.
In your case the "best" GPU is determined because of "OpenCL 2.0"

BOINC reports all the details to the server, I recently made 2 posts about that:
http://setiathome.berkeley.edu/forum_thread.php?id=79591&postid=1789475#1789475
http://setiathome.berkeley.edu/forum_thread.php?id=79738&postid=1793876#1793876

Applications know where they run and detect both GPUs OK:

Running on device number: 0
http://setiathome.berkeley.edu/result.php?resultid=4970229273
Low-performance GPU detected, default period_iterations_num set to 500 Used GPU device parameters are: Number of compute units: 2 Single buffer allocation size: 128MB Total device global memory: 1648MB max WG size: 256 local mem type: Real LotOfMem path: yes LowPerformanceGPU path: yes period_iterations_num=500


Running on device number: 1
http://setiathome.berkeley.edu/result.php?resultid=4970875081
Used GPU device parameters are: Number of compute units: 5 Single buffer allocation size: 128MB Total device global memory: 2048MB max WG size: 256 local mem type: Real LotOfMem path: no LowPerformanceGPU path: no period_iterations_num=50
28) Message boards : Number crunching : No more guppi's=vlars on the gpu please (Message 1794147)
Posted 24 days ago by Profile BilBg
. . BTW, can you tell me how to insert a graphic/image into a message?

Example:

1) Go to:
http://postimage.org/

Upload your file (for programs' windows the best format is .png)

2) Copy/Paste the "Direct Link" here, in my case it was:
http://s33.postimg.org/xv9h331wv/ATI_Memory_Viewer_07_06_2016.png

3) Mark/select the whole line ([Home], Shift+[End] on the keyboard)

4) Click [Img] button here, you will get:
[img]http://s33.postimg.org/xv9h331wv/ATI_Memory_Viewer_07_06_2016.png[/img]


5) Use the [Preview] button to see if all is correct.

And you will get:




P.S.
- To see the 'code' of my post - use the [Quote] button under it and read the raw text.

- The original filename was:
ATI MemoryViewer - 07.06.2016.png
It was changed automatically by postimage.org to ATI_Memory_Viewer_07_06_2016.png

- On the first usage of postimage.org you have to select "FAMILY safe" before [Upload It!]
(it is remembered for the next visits)
 
29) Message boards : Number crunching : Some considerations regarding OpenCL MultiBeam app tuning from algorithm view (Message 1794138)
Posted 24 days ago by Profile BilBg
-period_iterations_num N
"splits single call to N subsequent calls each of those process only 1/Nth of all available periods"

What is the max and min # of "all available periods"?
(1/50 of 1000 can be easy on GPU and 1/50 of 1000000 can be hard (lags))

1/Nth is relative to the # of all available periods

Do we have a way to set the (absolute) max # of periods in single kernel call?
e.g. something like:
-max_num_of_periods_per_call 1000
-num_of_periods_per_call_max 1000

(i.e. if "1/Nth of all available periods" gives more than 1000 the N should be changed automatically (for this sequence of calls) so less than 1000 periods be computed in single kernel call.
(I don't know if 1000 is anywhere near the reality.)
)
30) Message boards : Number crunching : CPU time difference (Message 1793876)
Posted 25 days ago by Profile BilBg
*edit* Display it on the server system summary page that is, it displays it locally just fine, every time for me.

BOINC not only "displays it locally just fine", it also reports all this info to the server.
Check your sched_request_setiathome.berkeley.edu.xml
Find in it strings <host_info> and <coproc_

(I have only one <coproc_ati> but I think you'll find in your file several GPUs listed)
31) Message boards : Number crunching : LotzaCores and a GTX 1080 FTW (Message 1793783)
Posted 25 days ago by Profile BilBg
for whatever reason, BOINC can't seem to identify more than one video card correctly.

Not really true, especially for GPUs from one vendor.
(i.e. if all are NVIDIA they use the same NVIDIA driver)

Maybe you mean not "can't seem to identify" but "can't seem to use"?
Because if the GPUs are all NVIDIA but different - for BOINC to use all of them you will need:
<use_all_gpus>1</use_all_gpus> in cc_config.xml

If in doubt - post the start of Event Log (Ctrl+Shift+E)


More info on "use_all_gpus" (by Ageless):
https://setiathome.berkeley.edu/forum_thread.php?id=78800&postid=1756194#1756194

... and later on the same thread about "Dummy Plug" (by me):
https://setiathome.berkeley.edu/forum_thread.php?id=78800&postid=1774759#1774759
32) Message boards : Number crunching : What drivers are people using with the 1080 (Message 1793293)
Posted 27 days ago by Profile BilBg
Plus, they have released a new version of Precision X 16 for the new hardware too. The last version was released in 2013, so it's about time for a new version. Not sure if it is a 64 bit version or not, but I hope so, and if not, I hope they are working on one and will release it soon.


That new Precision X requires your name, email and SN of the card to install, lol

I get the file from (no need for any name, email, etc.):
http://www.majorgeeks.com/files/details/evga_precision.html

On the page you can see this "funny" comment:
"Editor's Note:
Wow, what a nightmare to download.
Register with them asking for your email, address and phone number, confirm your email, answer a survey.
That sucks but we did it all for you so you don't have to.
Enjoy!
"

The file is:
EVGA_PrecisionX_16_Setup_v5.3.11.zip
https://www.virustotal.com/en/file/8409bf60e09e6935fefb7a68737f1f246c3039f0ea573582ab6f85b086197eef/analysis/

... and inside of the .zip is:
EVGA_PrecisionX_16_Setup.exe (28 682 104 bytes) 12.04.2016
https://www.virustotal.com/en/file/0cbbcd8703681829ed2ebee2a405ed32777c2c342892ffd0b743ddee69fd1c7e/analysis/

I tested installing it on Windows XP - again no need for any name, email, etc.


(The installed PrecisionX.exe can't start on Windows XP but that is another/different issue:
D:\Program Files\EVGA\PrecisionX 16\PrecisionX.exe is not a valid Win32 application.
)
33) Questions and Answers : Web site : Why is re-logging in necessary? (Message 1793254)
Posted 27 days ago by Profile BilBg
It's simple: as per Message 1787070
- something on your system deletes cookies (Do you use some "cleaner" program? Do you use the same Antivirus with same Settings on all computers?)
- you have your browser set to discard or block cookies (Do you use some extension/add-on in the browser that may do this?)

The password (in fact 'account key') is remembered in a browser cookie:
Name: auth
Content: {your 'account key'}   http://setiathome.berkeley.edu/weak_auth.php
Domain: setiathome.berkeley.edu

... and this cookie have (normally) life of one year:
Created: March 3, 2016 at 7:20:10 AM
Expires: March 3, 2017 at 7:20:10 AM


P.S.
You didn't tell exactly which browsers do you use on your computers (you mentioned "Internet Explorer" but is this the browser you use now?)
If "Internet Explorer" have something like "incognito" mode/window - in that mode no cookies or history is preserved.
(Any Chromium clone (I use SRWare Iron/IronPortable and Slimjet Browser) have "incognito" mode invoked by Ctrl+Shift+N)

I don't use IE (and "my IE" is the old Windows XP version) so don't know if something like "incognito" mode exists in it.
 
34) Questions and Answers : Web site : possible bug for webpages listing the 10,000 Top participants & Top Computers (Message 1791913)
Posted 30 May 2016 by Profile BilBg
Another similar thread ("Top Hosts"):
http://setiathome.berkeley.edu/forum_thread.php?id=79692
35) Questions and Answers : Web site : possible bug for webpages listing the 10,000 Top participants & Top Computers (Message 1791808)
Posted 29 May 2016 by Profile BilBg
OK, I see you now at 2880 and 2888 with different RAC and Credit (no matter if I use Ctrl+F5)
http://setiathome.berkeley.edu/top_users.php?sort_by=expavg_credit&offset=2860
http://setiathome.berkeley.edu/top_users.php?sort_by=expavg_credit&offset=2880

2880 Stubbles69 9,267 278,507 Canada 29 Nov 1999, 23:28:13 UTC ..... 2888 Stubbles69 9,245 277,942 Canada 29 Nov 1999, 23:28:13 UTC


At the left (here) I see yet another "version" of the numbers:
Credit: 282,078
RAC: 9,269

So there seem to really exist bug on "Top participants" pages (in top_users.php or in database)

P.S.
I just sent Message to David Anderson
36) Questions and Answers : GPU applications : kann man die gpu Auslastung begrenzen? (Message 1791038)
Posted 27 May 2016 by Profile BilBg
kann man die gpu Auslastung begrenzen? wenn ja wie?

you can limit the gpu utilization? if yes how?

Not really
But you may try TThrottle and set lower temperature for the GPU
(this will lower the GPU utilization if the temperature reaches the limit set by you)
37) Questions and Answers : Web site : possible bug for webpages listing the 10,000 Top participants & Top Computers (Message 1791033)
Posted 27 May 2016 by Profile BilBg
Also, sometimes I am not listed at all since my RAC falls in between pages (slightly too high on a certain page but slightly too low on the "previous" page).

Did you try F5 or Ctrl+F5 (before and after "Previous 20 · Next 20")?

At the moment of post you are at 3045
http://setiathome.berkeley.edu/top_users.php?sort_by=expavg_credit&offset=3040
38) Questions and Answers : Web site : Is there are way or a place to get the bulk data from: Statistics | Top Computers/Hosts? (Message 1791028)
Posted 27 May 2016 by Profile BilBg
Not "high level of accuracy" but you may spot the most active areas on this animation:
http://setiathome.berkeley.edu/kiosk/

If you look at the page code Ctrl+U
view-source:http://setiathome.berkeley.edu/kiosk/

... at the bottom are these links:
http://setiathome.berkeley.edu/kiosk/js/in-locs.js
http://setiathome.berkeley.edu/kiosk/js/out-locs.js

They are tables/lists of
... {"la":"46.81228","lo":"-71.21454","cc":"CA"}, {"la":"51.45625","lo":"-0.97113","cc":"GB"}, ...

I think they change every day, you may save them and find a way to analyse.
39) Questions and Answers : Web site : Is there are way or a place to get the bulk data from: Statistics | Top Computers/Hosts? (Message 1790997)
Posted 27 May 2016 by Profile BilBg
Not "high level of accuracy" but you may spot the most active areas on this animation:
http://setiathome.berkeley.edu/kiosk/

"By country" map:
(I doubt "they" (stats sites) can supply better accuracy as IP is private and projects keep it secret)
http://boincstats.com/pl/stats/0/project/detail/country

("By time zones" may apply only to "wide" (W-E) countries (e.g. Russia) and not to small or "tall" (N-S) countries (e.g. Chile))

P.S.
The page at stats.free-dc do not work for me ("Loading Stats Data..." never finishes):
http://stats.free-dc.org/stats.php?page=countries&proj=sah
40) Questions and Answers : Windows : Task Stalled at 99.730% Complete (Message 1790160)
Posted 25 May 2016 by Profile BilBg
Failed to load 'user_logo'.
Failed to load 'background'.

Nothing to worry about these, they are normal Messages (they are "by default") because user_logo and background are custom files which normally don't exist.
Only if the user wants he may supply in SETI@home preferences "URL of background image" and "URL of logo image"


I tried suspend/resume about 50 times with no change.

How long do you wait (e.g. 2-3 minutes should be enough) after resume?

I'm almost sure the app is hanging in "Optimal function choices" (which don't depend on task), this is long (years) existing bug in stock CPU apps.
If the app passes the "Optimal function choices" (startup phase) it will continue with the task and not hang during its processing.

You can see the app is doing real progress by:
- in Windows Task Manager, on Performance tab - from View menu check "Show Kernel Times" - then "too much" Red will mean the app is hang in "Optimal function choices"
- date/time of state.sah will change every minute (that is the checkpoint file)
- the <prog>0.99646484</prog> inside state.sah will change (increase)


Did you try to reboot?

P.S.
I know how to avoid this hang (and compute much faster) but you will lose the SETI@home screensaver

To try this "fix"/improvement - use the Lunatics installer (to install optimized applications):
http://setiathome.berkeley.edu/forum_thread.php?id=71867&postid=1758149#1758149

Again: optimized applications don't have screensaver


Previous 20 · Next 20

Copyright © 2016 University of California