Posts by Werinbert
log in
1) Message boards : Number crunching : Clock speed dropping (Message 20444)
Posted 953 days ago by Werinbert
I believe that due to Collatz running as integer math it is pushing the card in ways not used by other programs (single/double precision floating point).

I am not sure, but I suspect that with the quicker gating operations of integer math the chip is pushing through higher voltages (or rather less electrical resistance than float operations). To compensate the card reduces the voltage to keep from over powering the circuitry.
2) Message boards : News : Corrupt Index Caused Authentication Errors (Message 20335)
Posted 1009 days ago by Werinbert
Hey, if you've got an M240 machine gun, I would RTFM on how to disassemble and reassemble it just so I could experience firing it once!


You can always go to Las Vegas and shoot one I think, they have almost every gun available at their public shooting ranges. And yes even machine guns! It is a big draw for some tourists, go there and shoot up some sand.


I would prefer that than to shoot up some servers.
3) Message boards : Number crunching : Odd gpu behavior (Message 20167)
Posted 1073 days ago by Werinbert
I have recently started oc'ing my 750TI. Afterburner settings coreclock +124, memory +300. While Collatz was down I was using PrimeGrid PPS sieve to do my testing. With the PG WUs GPU-Z sensors report 1300 clock, 1500 memory with 1.2V. These are reasonable without really pushing the card. The clock speeds and voltage don't fluctuate.

However when I switched over to Collatz WUs the GPU-Z sensors report a fluctuating core clock and VDDC generally between 1240Hz/1.1180V to 1267Hz/1.1430V. Anyone have an idea why the Collatz WUs would cause a decrease in clock speed/voltage?

I only noticed this because my mouse is having drop outs (the mouse works fine when Collatz is not running), yet my screen is not having noticeable lag. None the less I've tried changing some of the settings in the .config file but that doesn't seem to affect the mouse problems.
4) Message boards : Number crunching : Umm... So... What happened over the holidays? (Message 20165)
Posted 1074 days ago by Werinbert
As far as I can see the servers were on winter holidays. People and servers deserve a break every so often. ;-)
5) Message boards : Number crunching : Happy New Year! (Message 20164)
Posted 1074 days ago by Werinbert
what was happened?

The servers were on winter holidays.
6) Message boards : Number crunching : Happy New Year! (Message 20156)
Posted 1075 days ago by Werinbert
It feels good to be back up and crunching!
7) Message boards : Number crunching : Differences between WUs and their credits (Message 20036)
Posted 1136 days ago by Werinbert
It looks like you are getting a mix of mostly solos with a few larges thrown in.
Check your Collatz preferences to select only the WUs that you want.

very roughly:
micros are around 55cr/wu
minis are around 900cr/wu
solos are around 15kcr/wu
larges are around 240kcr/wu

Per day depends on you computer, how often you are running boinc, and what type of GPU you are running. My 3+ year-old laptop {540M) which I turn off at night does around 30kcr/day, my new desktop (750Ti) which runs 24/7 does around 300kcr/day. Both computers run solos only.

See this thread http://boinc.thesonntags.com/collatz/forum_thread.php?id=2&start=0 for a more thorough discussion of credits.
8) Message boards : Number crunching : using both nvidia-gpu and intel-gpu (Message 20035)
Posted 1136 days ago by Werinbert
I stand corrected, the app info is working almost like expected. The igpu WU is in fact using 1 cpu thread for processing however it still shows up on the Boinc Manager as 0.538.


Have you restarted BOINC since you edited your app_config, BOINC will show the old values until you either get new work units or restart the client.


Yeah, I realized that the old values were being shown. I also found out by inadvertent testing that just changing the app_info file's name (so it would not load) and then reloading the app_info files via the Advance menu will not reset the app_info values to default. I am happy to say everything is working as I want it in any of various scenarios.

Now my current endeavor is to run 8 PrimeGrid 321 (llr) tasks along with solos on my GPU and iGPU and seeing how far it heats up as I have installed a new cooler on my machine. [OC'ing will come sooner or later.]
9) Message boards : Number crunching : using both nvidia-gpu and intel-gpu (Message 20031)
Posted 1137 days ago by Werinbert
I stand corrected, the app info is working almost like expected. The igpu WU is in fact using 1 cpu thread for processing however it still shows up on the Boinc Manager as 0.538.
10) Message boards : Number crunching : using both nvidia-gpu and intel-gpu (Message 20030)
Posted 1137 days ago by Werinbert
using app_config with the following:
<app_config>
<app_version>
<app_name>solo_collatz</app_name>
<plan_class>opencl_intel_gpu</plan_class>
<avg_ncpus>1.0</avg_ncpus>
<ngpus>1.0</ngpus>
</app_version>
</app_config>

I am still getting 0.538 cpus not 1.0 as hoped for the intel gpu. I am probably missing something, I just don't know what.
11) Message boards : Number crunching : using both nvidia-gpu and intel-gpu (Message 20026)
Posted 1140 days ago by Werinbert
I have been testing the values given by FalconFly.

GTX 750Ti:
verbose=1
threads=10 (previously I had 8)
items_per_kernel=21 (previously I had 20)
kernels_per_reduction=9
sleep=1

I have seen no change in time nor credits per time. So this is a wash.

HD4000:
(previously I used the defaults)
verbose=1
threads=8
items_per_kernel=20
kernels_per_reduction=9
sleep=1

I haven't run enough tests yet to be assured of my answer... the run time has gone down by about 10-15% (this is good) but I now have noticeable (if small) amount of lag (not so good). As I am often using this computer for other things I not sure I like the lag and will probably go back to the defaults. It is nice to have these numbers but, they do not solve my problem.

As to my original question I will shortly test the solution given by Slicker. I read a while ago about the new app_version tag in the app_config file, and promptly forgot all about it as I never expected to use it. Thanks for the reminder. I fully expect this to solve my problem (so long as I can find the right class names and keep myself from entering any typos).
12) Message boards : Number crunching : using both nvidia-gpu and intel-gpu (Message 20004)
Posted 1144 days ago by Werinbert
My main crunching computer is running solos on my nvidia card (750Ti) and on my intel igpu (HD4000). The problem is that the igpu WU in reality needs 1 cpu core not .538. I thought about modifying the app_config file but this would also affect the nvidia work units (which are running just fine with minimal cpu...I am running cuda).

The only solution I have come up with is to manually set bionc to run only 90% of my cores. Has anyone come up with a better solution?
13) Message boards : Number crunching : I am aborting a Large because.... (Message 19933)
Posted 1163 days ago by Werinbert
I thought that it would be better to give my GPU work rather than run dry, so I added Mini to the list of acceptable apps. And then turned off "If no work..." Running Minis for two days on my CPU is an acceptable concession. At least those will be returned within the deadline. I just never thought the scheduler would send a large to that computer.
14) Message boards : Number crunching : I am aborting a Large because.... (Message 19928)
Posted 1164 days ago by Werinbert
It was sent to a computer running Intel Atom/Nvidia Ion.

Prefs were set to get Micros and the if not available accept something else. It takes 4 hours to do a Mini on the Ion, 3 hours for a Micro on the Atom. As luck would have it, when the scheduler wouldn't give the Ion a Micro it gave it a Large instead...estimated run time of 1000 hours. Perhaps this is the great flood.

Not complaining, just laughing. :-)
15) Message boards : Number crunching : More QQ and less pewpew (intel-GPU) (Message 19905)
Posted 1173 days ago by Werinbert
An update, sort of...My GTX660Ti died a while ago so I moved processing over to the iGPU. And it was running cool enough.

Then in early August I got a GTX750Ti as a replacement (1/2 the cost, 1/2 the processing, and 1/3 (or so) the heat output). I have since been running PrimeGrid and other apps on my CPUs, Collatz on the iGPU, and Collatz on the GPU. With the 660Ti I didn't see much improvement giving it a free cpu core so didn't try it. But with the slower 750Ti giving it a free core has helped around 10% so I have set Boinc to run 90% or 7 threads.

While running 7 threads, the iGPU and a GPU my computer is, none the less, running within reasonable temps: 70-80deg CPU/iGPU, and 40-50deg GPU. I wonder if there was some sort of weird interaction between the iGPU and the 660TI.
16) Message boards : Number crunching : Badges (Message 19604)
Posted 1290 days ago by Werinbert
WCG has expanded their badges to cover 100 years of PC crunching time.
I wonder if anyone already has a 100 year badge?


Go to http://www.wcgsig.com/ look at the top people and click on the "badges" column to view their WCG page. So yes, there are people with multiple 100+ year sub-projects.
17) Message boards : News : Variable Sized Wokunits & New CPU Applications (Message 18909)
Posted 1364 days ago by Werinbert
never mind...
this thread covered it: http://boinc.thesonntags.com/collatz/forum_thread.php?id=4&postid=18881#18881
18) Message boards : News : Variable Sized Wokunits & New CPU Applications (Message 18908)
Posted 1364 days ago by Werinbert
Is there a way to set it so I don't get the multi core variant of the cpu app? Rather I would prefer to run the single core app if possible.
19) Message boards : Number crunching : Solo_Collatz : waiting validation ??? (Message 17933)
Posted 1480 days ago by Werinbert
Still better than SETI validation times which can take months..... And the WUs are downloaded and uploaded with out problems so the GPU isn't running dry or anything.

Crunch on!
20) Message boards : Number crunching : High Step questions (Message 17876)
Posted 1485 days ago by Werinbert
http://boinc.thesonntags.com/collatz/high_steppers.php
The page shows Today's top High Steps and other recent high steps.

The lists always has a pair of finders, is the solo_collatz not shown on this page? Or is the solo paired with someone?


Also, from my accounts page I have:
Best Result 2240 steps for 2382393916683840078879 on 2013-11-08 21:40:28

After looking at the high step page I find that there is an entry with the same high step value on the given day as mine, but no entry for mine. Mine occurred at a later time if that makes a difference.


Next 20

Main page · Your account · Message boards


Copyright © 2017 Jon Sonntag; All rights reserved.