Posts by Sean

1) Questions and Answers : GPU applications : Gpu overheat on AMD Sapphire Tri-X R9 290 (Message 1903760)
Posted 30 Nov 2017 by Profile Sean
Post:
In your BOINC data directory there may (or may not) be a file called cc_config.xml

https://boinc.berkeley.edu/wiki/Client_configuration has in depth details on creating and editing this file, but adding:

    <ignore_ati_dev>0</ignore_ati_dev>


should stop BOINC from using your Radeon.

2) Questions and Answers : GPU applications : AMD Rx Vega 56 issues (Message 1893506)
Posted 6 Oct 2017 by Profile Sean
Post:
Scanning thru the first page of your results under the inconclusive column, the RX280 is throwing more than the Vega 56.

I hate to admit it, but that's because I've had the Vega solely on mining duty for the past couple weeks.
The 280x and CPU I've kept dedicated to SETI tasks. The 280x seems particularly well suited for SETI - I can't remember the last time it gave an invalid result and I'm quite happy with 30k RAC from the machine. I almost wish I had just bought another 280x instead of the Vega!
3) Message boards : Number crunching : It only took about 7 months (old cruncher thread)... (Message 1891781)
Posted 25 Sep 2017 by Profile Sean
Post:
But my Pentium 3 rig has now completed the amount of work that my main rig does in one day!
Too bad those SLI Voodoo2's can't help crunch :)

https://setiathome.berkeley.edu/show_host_detail.php?hostid=8203828
4) Questions and Answers : GPU applications : AMD Rx Vega 56 issues (Message 1891420)
Posted 22 Sep 2017 by Profile Sean
Post:
The one error was a (Unknown error) - exit code -6 (0xfffffffa) error

I guess I should have mentioned that the error(s) occurred when I crashed the system while tinkering, not during normal operation.
I will take a look at different drivers though!
5) Questions and Answers : GPU applications : AMD Rx Vega 56 issues (Message 1891401)
Posted 22 Sep 2017 by Profile Sean
Post:
I've recently added a Vega 56 to my main SETI crunching box:
https://setiathome.berkeley.edu/show_host_detail.php?hostid=6993114

Despite what the details say, the system has a Vega 56 and a R9 280x running together (not 2 Vega cards). I initially ran just the Vega card to try to work out any problems and it seemed like after updating to the newest AMD HD5 SoG app and setting the Vega to run at its base GPU clock speed, tasks were (mostly) validating. At that point, I added the 280x back in and now both cards are running tasks.

So the main issue is that the Vega is still producing quite a few invalids and inconclusives (which I suspect will turn invalid). It also seems that when the Vega and 280x run similar point value tasks, the 280x consistently finishes them slightly faster (and with no errors).

Some potentially important details:
I'm using a 1000 watt platinum power supply (I can mine 24/7 on both cards with no problems)
I am running the AMD HD5 app with settings tuned for the 280x. Maybe Vega doesn't work well with those settings? I'll try reverting to the default settings this evening.
I wiped the old drivers with 'Display Driver Uninstaller' and did a clean install of Crimson 17.9.2.
I run 2 tasks on both cards simultaneously and keep 1 CPU core free for each GPU task.

I would really appreciate feedback from anyone else running a Vega.

Thanks!
6) Questions and Answers : Unix/Linux : Dual Radeon HD7770, Ubuntu 12.04, Not showing full GPU memory? (Message 1868906)
Posted 23 May 2017 by Profile Sean
Post:
Question pertains to this host: https://setiathome.berkeley.edu/show_host_detail.php?hostid=7184655

This system has been running mostly trouble free (which is why I hesitate to 'upgrade' to a newer Ubuntu) for a few years now with a single HD7770.
To boost its RAC, I decided to add a second and identical Radeon. It took some fiddling to get it to work, but once it was working, the S@H host information indicated that the cards were something like 930mb each whereas before it showed the full 1024mb on the single card. Well, that didn't bother me too much but then I goofed it up and had to reinitialize the driver for a second time. Now it is saying the cards are 899mb... what's happening?

Thanks in advance!
7) Questions and Answers : Unix/Linux : Trying to move boinc-client folder to a USB flash drive (Message 1839678)
Posted 2 Jan 2017 by Profile Sean
Post:
It's working now! For anyone who might be wondering:
    sudo gedit /usr/share/applications/boinc-manager.desktop
    Changed the Exec= line to "Exec=boincmgr -d /media/me/usbdrive/boinc-client"
    Made sure the boinc-client folder permissions were set "sudo chown -R boinc:boinc /media/me/usbdrive/boinc-client"
    Start the boinc service, launch the manager, and it's working with the data directory on the usb drive


For whatever reason, I still have to manually restart the boinc-client service after rebooting, but I can live with that. I should only be rebooting for updates and power outages.

8) Questions and Answers : Unix/Linux : Trying to move boinc-client folder to a USB flash drive (Message 1839650)
Posted 2 Jan 2017 by Profile Sean
Post:
Perhaps making a little progress figuring this out, but I'm still not sure why it isn't working the way I think it should...

I went back to trying the symlink method:
    boinc-client folder moved to the flash drive
    symlink created pointing /var/lib/boinc-client to the boinc-client directory on the flash drive
    restart the boinc service and launch the manager
    result = boinc manager stuck on trying to connect to the client



Now I discovered a bit by accident running "boinc" from the command line (with or without sudo) gives me some useful information about the data directory:

    02-Jan-2017 10:53:07 [---] Data directory: /home/me
    Why is it trying to use my home directory? (edit: figured out it's just going with the current active directory)



OK, so now if I run "sudo boinc --dir /media/me/usbdrive/boinc-client" I get:

    02-Jan-2017 11:13:05 [---] Data directory: /media/me/usbdrive/boinc-client
    Looking good so far...
    Launch the manager and success!



Maybe I can just change the command line for the launcher icon to force the directory?

9) Questions and Answers : Unix/Linux : Trying to move boinc-client folder to a USB flash drive (Message 1839299)
Posted 1 Jan 2017 by Profile Sean
Post:
So I've tried a different approach:
    1. Stop the boinc client = sudo service boinc-client stop
    2. Copy the existing data to the new location = sudo cp -R /var/lib/boinc-client /media/me/usbdrive/boinc-client
    3. Make sure permissions are set correctly = sudo chown boinc:boinc -R /media/me/usbdrive/boinc-client/*
    4. Edit the /etc/default/boinc-client file to point to the new data directory = sudo gedit /etc/default/boinc-client and look for the entry that starts with # This is the data directory of the BOINC core client. Add an # in front of the BOINC_DIR="/var/lib/boinc-client" line and then add the following below that line BOINC_DIR="/media/me/usbdrive/boinc-client" and save the file then exit the text editor.
    5. Start the boinc client = sudo service boinc-client start



I also edited the init.d file similar to what was done in step 4.
BOINC manager starts up fine now and appears to be working, but the event log is still saying "Data directory: /var/lib/boinc-client"

Is there something else I'm missing?

10) Questions and Answers : Unix/Linux : Trying to move boinc-client folder to a USB flash drive (Message 1839214)
Posted 31 Dec 2016 by Profile Sean
Post:
I've recently assembled a new system (i7-6700T, 32GB ram, M.2 SSD, Ubuntu 16.04 LTS) and I'm trying to relocate the boinc-client folder to a USB flash drive to reduce write activity on the SSD (I'd rather let it chew on an $8 flash drive than a $200 SSD).

I have successfully moved the boinc-client folder to my home directory via the following:
    sudo mv /var/lib/boinc-client /home/me/boinc-client
    sudo ln -s /home/me/boinc-client /var/lib/boinc-client


And that is what I have running at the moment.

However, when I try (from a fresh install):

    sudo mv /var/lib/boinc-client /media/me/usbdrive/boinc-client
    sudo ln -s /media/me/usbdrive/boinc-client /var/lib/boinc-client


Boinc manager is unable to connect to the client. The manager screen will sit on "Connecting to client" for minute or so before popping up an error along the lines of "The client has closed unexpectedly several times in the last several minutes, try again?"
The USB drive is formatted to EXT4 and it does appear that the move command and link command are working correctly. Permissions also appear to be correct for the client folder on the flash drive.

At the risk of making this more complicated, I do have BOINC running on two other older Ubuntu systems. On the older machines, I just downloaded the .sh file and then extracted it and ran BOINC out of my home folder. I tried this on the new system, but unfortunately it looks like version 7.4 wants libraries that are not present in Ubuntu 16 and I could not find anywhere to download a version 7.6 .sh directly (had to install thru Ubuntu software center).

11) Message boards : Number crunching : v8 issues on very old cruncher (Message 1773505)
Posted 23 Mar 2016 by Profile Sean
Post:
If you're looking to throw the processing muscle of that Pentium MMX behind another project, PrimeGrid seems to be running fine so far on my K6-2+.

I also tried einstein@home and lhc@home. Neither worked.
12) Message boards : Number crunching : v8 issues on very old cruncher (Message 1772718)
Posted 20 Mar 2016 by Profile Sean
Post:
I've also done some seti v7 crunching on a 586 class chip (AMD K6-2+ in my case. See thread: "Retro PC seti crunching"). Using the older .dll file, v7 worked fine. But, like your rig, mine just throws an error trying to run v8 (old .dll, new .dll, doesn't matter). I think the likely scenario is that the new v8 app is trying to use CPU instructions that just aren't present in the 586.
13) Message boards : Number crunching : Retro PC seti crunching (Message 1772478)
Posted 18 Mar 2016 by Profile Sean
Post:
Completed! =)

https://setiathome.berkeley.edu/result.php?resultid=4778204546

Only about 1900 years before the K6-2+ matches what my i7/280x system has done to date!

Now, about that version 8!
14) Message boards : Number crunching : Question About RAC and v8 (Message 1771381)
Posted 13 Mar 2016 by Profile Sean
Post:
Having been running the optimized apps (AVX, SSE4, and AMD HD5) on my top two crunchers for a while now, my RAC actually seems higher than it was with the optimized apps under version 7. Most notably on my 4790/R9280x box, I don't think RAC ever hit 30k before and now it hasn't dipped below that for at least a couple weeks. Granted, I do understand that AP7 units give a lot more credit than MB8. Has there been more AP7 work than usual going out lately?
15) Message boards : Number crunching : Retro PC seti crunching (Message 1770902)
Posted 11 Mar 2016 by Profile Sean
Post:
So I now have the version 7 work unit making progress! I dug out a spare hard drive, installed Windows XP, copied my data folder over, and we're up and running! It must be some Windows 2000 issue not letting things run right. Estimated time to completion: 137 hrs :).

I'll revisit the version 8 issue when the time comes that I can't get any more version 7 work.
16) Message boards : Number crunching : Retro PC seti crunching (Message 1770518)
Posted 9 Mar 2016 by Profile Sean
Post:
This is a follow up to a previous thread: https://setiathome.berkeley.edu/forum_thread.php?id=78526

Referencing this computer: https://setiathome.berkeley.edu/show_host_detail.php?hostid=7830684

I definitely appreciate the help I got in the original thread. I was able to get the K6-2+ to successfully complete a work unit under Windows 7. Since then, I've ditched Windows 7 and gone with Windows 2000 SP4 (much more usable on the old system - posting from it now, in fact). I have two questions/problems.

First, I managed to get one of the remaining version 7 work units, but the progress is stuck at 0.0% even after 13+ hours of run time. I'm using the exact same app_info, executable, and dll that I did under Windows 7 (see previous thread). The elapsed time continues to run but interestingly, if I exit and restart BOINC, the time starts back at 00:00.
stderr.txt shows:
setiathome_v7 7.00 DevC++/MinGW/g++ 4.5.2
libboinc: 7.1.0

Work Unit Info:
...............
WU true angle range is : 0.107183
Optimal function choices:
--------------------------------------------------------
name timing error
--------------------------------------------------------
setiathome_v7 7.00 DevC++/MinGW/g++ 4.5.2
libboinc: 7.1.0

Work Unit Info:
...............
WU true angle range is : 0.107183
Optimal function choices:
--------------------------------------------------------
name timing error
--------------------------------------------------------
setiathome_v7 7.00 DevC++/MinGW/g++ 4.5.2
libboinc: 7.1.0

Work Unit Info:
...............
WU true angle range is : 0.107183
Optimal function choices:
--------------------------------------------------------
name timing error
--------------------------------------------------------

I don't recall if the previous unit that processed successfully was a vlar, but I'm wondering if that has something to do with it.

Second question is if there is any way to get the version 8 app to process on the 586 architecture? I tried modifying app_info to use the version 8 app with the older dll to no avail. It outputs the following error:
<core_client_version>6.6.38</core_client_version>
<![CDATA[
<message>
There are no child processes to wait for. (0x80) - exit code 128 (0x80)
</message>
]]>


Thanks!
17) Message boards : Number crunching : Noctua Coolers (Best) (Message 1768940)
Posted 2 Mar 2016 by Profile Sean
Post:
My E8500 is cooled by a Noctua 92mm HSF in push/pull. My 4790 and Pentium D systems are both cooled by 120mm Noctua fans mounted to Cooler Master Hyper 212 heatsinks... the Cooler Master fans died.
18) Message boards : Number crunching : AMD 280x cards on i7 920 (Message 1765324)
Posted 16 Feb 2016 by Profile Sean
Post:
I've been crunching on a HIS 280X 24/7 for well over a year now and have been pleased with the results. RAC (paired with an i7 4790) is around 27,000. The switch from MB v7 to v8 didn't seem to make much difference. I'm also using the HD5 app downloaded from Mike's World.
19) Questions and Answers : Windows : Latest Windows 10 update broke AMD GPU app on FirePro M3900 (Message 1750670)
Posted 20 Dec 2015 by Profile Sean
Post:
As the title says, I've been crunching on an EliteBook 8640W. Certainly not the fastest GPU in the world, but it has been working without any issues using the Lunatics HD5 app on Windows 10. With the latest Windows 10 update, the app is now crashing the display driver which causes the task to hang. I un-did the update and now the FirePro is happily crunching away again. Anyone else run into the same or similar issue?

https://setiathome.berkeley.edu/results.php?hostid=7720726
20) Message boards : Number crunching : Retro PC seti crunching (Message 1744040)
Posted 22 Nov 2015 by Profile Sean
Post:
With only 32mb ram, BOINC would not download seti tasks. I tried it on one of my 486's a while back. It gave some message about insufficient memory.

Does anyone know where I might find a copy of libfftw3f-3-1-1a_upx.dll? I've been searching with no luck for a few hours now.


Next 20


 
©2018 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.