BOINC

brilong said:
NFS@Home Challenge


I received this challenge on mersenne forum (GIMPS, PSP) and was asked to share with [H] Team Members.
Click to expand...

I will also point out that this will overlap the PrimeGrid Sovereign Challenge for those that are interested. 25-28 June 18:00 SR5-LLR Sovereign Challenge 3 days
 
Good news - problem solved, easiest solution!

Step 1: Blow dust bunnies out of CPU cooler
Step 2: Lower 3.2 GHz winter OC to 3.0 GHz summer OC
Step 3: Confirm CPU temps are running 6-8* cooler
Step 4: Profit!

Doesn't fully explain why it wouldn't even boot to the BIOS screen this morning, but it fired right up when I got home and reloaded the PROPER BIOS... (CMOS Battery is dead on the Board, so every time I have a power failure, I have to reload the proper OC settings... Yes, I'm too lazy to replace the battery...)
 
Jathanis said:
Yes, I'm too lazy to replace the battery...)
Click to expand...

Here's a battery, go replace it! :cool:

311c419328a02d615040f110.L.jpg
 
Jathanis, I'm glad it was an easy and cheap fix. I would send you an extra battery but I tested several recently from system pulls and most of them had died or were really low. LOL :eek:
 
@brilong
WCG team output tripled the last few days ... nice ramp up!!!

What are you running to get that kind of production?
 
Friendly tip for the WCG crunchers having problems running multiple CEP2 units ...
I've been running 4 units simultaneously on both dedicated Ubuntu Server rigs for 6-9 months with no issues. Both boxes have old WD/Seagate SATA drives. File uploads are 20MB right now. My trusty WRT54G is handling them fine. CEP2 looks to be much more stable on Linux than Windows. I had nothing but problems running multiple units on my Windows boxes. Just my experience. YMMV
 
Last edited:
AgrFan said:
@brilong
WCG team output tripled the last few days ... nice ramp up!!!

What are you running to get that kind of production?
Click to expand...

Look at my sig. :D

I had set WCG to "0" in boincstats so it would only pull WUs when nothing else was running. I was pushing SIMAP since we were easily making DC-Vault points a little at a time. Now it appears SIMAP has not provided WUs in a while so my projects set to "0" are picking up WUs.

I may disable BOINC on a few more clients and put more weight on "Prime Sierpinski Problem - PRP" since I finally got added to our team. We already gained 1 vault position (333 points) because of this.
 
POEM has their test server back up if anyone is interested in helping them with the testing. I will just wait for the day they get a working app ready. I just don't have the time to mess with it right now.
 
It's only the OpenCL app that's having such issues isn't it? Either way, I'm good trying to run other things too...
 
Kai said:
It's only the OpenCL app that's having such issues isn't it? Either way, I'm good trying to run other things too...
Click to expand...

Yes...this is the long promised GPU app. From what little I have read up on it, it sounds like they are having problems cross platform. They start with Linux and go from there. I haven't bothered asking about whether they could at least release a working Linux version until they can figure the rest out. I'm assuming they have already thought of that and are doing what they need.
 
Hey all.

A little confused what team to change to if I wanted to start throwing points for HardOCP.

There are two teams "HardOCP" and "[H]ard|OCP" is there a difference?
 
Thanassos said:
Hey all.

A little confused what team to change to if I wanted to start throwing points for HardOCP.

There are two teams "HardOCP" and "[H]ard|OCP" is there a difference?
Click to expand...

Thanossos: First of all, welcome. Glad to potentially have you with us. Second, we are trying to rectify the dual team name issue and I have not heard back from the captains on that progress yet. Thanks for reminding me. I have a thread that helps point out which to use. It looks daunting at first, but I think it covers several questions.

I have a thread in our Guides section that covers most projects both BOINC and non-BOINC. http://hardforum.com/showthread.php?t=1801908

If you scroll down to post 5 http://hardforum.com/showpost.php?p=1040548916&postcount=5 You will see which projects use which name. You didn't say which project you were referring too. Again, we are working on getting things adjusted to eliminate some of the confusion.

If you have further questions or this wasn't clear enough, feel free to let me know. I'm always willing to help however I can.

Thanks ChelseaOilman, you beat me to it. LOL
 
brilong said:
Here's a battery, go replace it! :cool:

311c419328a02d615040f110.L.jpg
Click to expand...

Okay, I printed it out, put it on the motherboard, and it still isn't working to save the time... I'll wait for Gilthanis to do a guide for me! :D
 
FYI, I'm in the process of shutting down for the summer. Set my clients to not get any new work yesterday. One machine has a few E@H wus left to finish. Heading for the much cooler temps of the Colorado Rockies. Not sticking around for the 100 degree temps down here in TX. CO has expensive summer electric rates so not doing any crunching while I'm there. Probably start things back up again late September.
 
qdemn7 said:
I recommend that Free Rainbow tables be considered a Dead Project. Especially the Open CL WU.

No communication from the Admins.

https://www.freerainbowtables.com/phpBB3/viewtopic.php?f=2&t=4659&p=23290#p23290
Click to expand...

Has it been just the OpenCL app that was stopped or did they stop all work units? I know the server status doesn't show anything ready to send. If there is still CPU work, then the project isn't dead. Just the high earning GPU work is in a spell. Since it was announced to be turned off April 22nd, It would be premature to really say "no Open CL WU for months".... We are noq going on 1.5'ish months. However, if no work at all is available, then I would agree it should be put on hold. Any insight on those CPU work units? http://boinc.freerainbowtables.com/distrrtgen/apps.php

And if the admins are looking at Crunch3r's app, they need to test it internally for a bit to make sure it has proper results. We have waited longer for POGS to get their shit together for the optimized client that we had months ago as well as the OpenCL GPU app...

Edit: however, I will mention it to the DC-Vault admins in case they have an opinion in regards to adding it to the Vault. I will also mention that I can't post in their forums because when I tried to sign up months ago I got blacklisted due to the ISP I was on at the time. Now my login name and email address is associated with a Spamming list.
 
Last edited:
So a while back, I mentioned that I fired up NativeBOINC on Bluestacks on my i7 (@ 3.8 GHz) to play with, and after a week or so of running it, here are some rough Enigma results - comparison is from my full time crunching Tablet, a first gen Galaxy Tab 7 OCed from 1GHz to 1.4GHz:

PPD, 7 day average:
Galaxy Tab: 483 ppd
i7 core: 392 ppd

Average run time, 30 point WUs:
Galaxy Tab: 5,022 sec over last 11 WU
i7 core: 4,333 sec over last 12 WU

Add to those numbers the fact that the i7 is also running BOINC on hyperthreading (and Dimes, and M-12, and lately Batman Arkham City GOTY 3-5 hours in the evenings), so the Android emulator is actually using a portion of of a thread, not even a full core, so it would likely do better (as the WU times show) if using a dedicated core... However, I'm not going to suspend a core and lose 2 pogs WU @ >1,000 ppd to gain what I figure would be 500 more PPD at best... :D

But if an Android only BOINC project ever hits the vaults, it's safe to say an OCed i7 is worth about one Arm core per hyperthread, or about 4 dual core Android phones!
 
You also have to look at the fact that the big gains from Linux OS's aren't in the 32 bit versions...so until Google releases 64bit Android, I wouldn't expect it to out perform. What you should also point out is the power draw between that ARM core vs. the i7 core. When looking at the cost of producing, that is the big consideration.

And don't forget, that the official BOINC version has the option to auto update from the Play store. That in itself is a huge perk for many DC'ers.
 
I had not checked their page for a while, but saw this for those running ARM devices...

June 3, 2014
We are happy to announce that today we added the android arm platform minirosetta_beta version. If you have an android arm device/phone that supports android-9, please download and install the boinc client and attach to our project. Please continue to post bugs/issues in the 3.50 thread.
Click to expand...
 
I'm starting a list in the All inclusive DC list that points out which projects our team has confirmed prefers HT on or off. If you have any input to add, let me know and I will get it added so others will know.


http://hardforum.com/showthread.php?t=1801908
 
Last edited:
Gilthanis said:
I'm starting a list in the All inclusive DC list that points out which projects our team has confirmed prefers HT on or off. If you have any input to add, let me know and I will get it added so others will know.
Click to expand...

Awesome idea!
 
Gilthanis said:
I'm starting a list in the All inclusive DC list that points out which projects our team has confirmed prefers HT on or off. If you have any input to add, let me know and I will get it added so others will know.
Click to expand...

Doh, just saw this after asking Grandpa for something similar. :D
 
brilong said:
Doh, just saw this after asking Grandpa for something similar. :D
Click to expand...

Thats OK I already knew what Gilthanis was doing when I read your post. I am just being lazy and posting my test results in the appropiate threads so Gilthanis has them for his guide. :D
 
DooKey why are there 2 DooKey's on the team. Hmmm :eek:

1 399 DooKey 622,906 239,363 -3,006,605 6,851,953 1,295,355 7,609,678 63,447,289 2,178,736 686,808,226

53 28169 DooKey 152,640 152,640 291,360 0 166,766 1,320,000 1,453,539 89,138 5,918,880
 
Grandpa_01 said:
DooKey why are there 2 DooKey's on the team. Hmmm :eek:

1 399 DooKey 622,906 239,363 -3,006,605 6,851,953 1,295,355 7,609,678 63,447,289 2,178,736 686,808,226

53 28169 DooKey 152,640 152,640 291,360 0 166,766 1,320,000 1,453,539 89,138 5,918,880
Click to expand...

Asteroids keeps jumping from one CPID to another back and forth for some reason.
 
http://boinc.fzk.de/poem/forum_thread.php?id=1069#9682

As many of you have noticed, our GPU project has been re-enabled after the update last week.
With the apparent error of our last release being fixed, I'm still not satisfied with the failing rate of GPU tasks, especially on NVIDIA cards.

I think there are currently three main issues:

1. Incompatibility to old AMD GPUs.
We're suspecting our latest version of poemcl won't run properly on cards of the Radeon 4000 series, which only have a limited OpenCL support.
If you are using such a GPU, and it works however, please let me know which OS and driver version you are using.
If there is no working setup, we will exclude the Radeon 4000 series in future releases.

2. Multi-GPU NVIDIA hosts still fail when using more than one graphics card for poemcl tasks.
If you have attached a host with multiple NVIDIA cards to POEM@HOME, please configure your client to use only one of them for our project, see http://boinc.berkeley.edu/wiki/client_configuration headword exclude_gpu.
This is a well known problem which should have been fixed with the last release, but obviously it is a sticky one :(

3. There is a new error, which may be related to incompatible OpenCL library versions.
If your host uses only a single GPU, but still fails with poemcl tasks, please download one of the following files (depending on your OS):
boinc.fzk.de/userdocs/temp/pcl_201_test_linux.tar.gz
boinc.fzk.de/userdocs/temp/pcl_201_test_win.zip
Extract the files to a new folder, run the executable (if it works, this may take some minutes), and post the content of stderr.txt in this thread.

Thanks for your help!
I hope we can improve the application soon with your feedback.
Click to expand...

Looks like I need to make some config changes since I'm running mostly multi-GPU rigs now.

Also, read the full thread because there are still some issues.
 
Last edited:
My nVidia GT430 returned its first work unit.
Run Time (akd Elapsed): 48,961.77 CPU (aka GPU) Time: 48,175.01 Points: 6,500.00

nVidia GTX650 returned its first work unit.
Run Time: 51,315.25 CPU Time: 42,831.65 Points: 6,500.00
 
Ironically, I sat down to order some stuff off Amazon last night and my video drivers kept locking up/recovering to the point I did a clean wipe and reinstall. Then I read this and realized it was just a rogue POEM WU that had slipped through and was playing havoc with my 9600GSO!
 
good to know. My 9600GSO is not currently turned on...so I will have to watch it. I also haven't tried running it on anything less than the two mentioned above yet.
 
So I think I'm getting the i7-950 swap to x5650 bug... Two more cores, lower power equal or better overclocking... And $88 on ebay - am I missing something?

Then to swap out the old q6600 system for another x5650/x58 system... and 12 threads become 24!

must. stay. off. ebay... (of course, the Steam sale starts tomorrow, so I might get distracted!)
 
Yeah, those hex cores are great value as organisations clear out their old hardware. I've picked up several L56xx CPUs myself, for the lower power consumption, but those X5650s look awesome as well. No reason not to, really!

I've had to cut down on my DC drastically...rising costs of, well, everything, plus lack of income due to my gf's situation is biting hard right now so we're cutting back everything we possibly can. Sold almost all my GPUs (keeping some as spares and one for light gaming) and I'm in the process of stripping my rigs to sell on. I'm still running the NCI BOINC projects and DIMES, and MJ-12 is still running on my Atom which is hosted elsewhere, but other than that I'm out of the game for the foreseeable future. I'll return as soon as I'm able to but things like keeping my car running so I can work, and paying the rent, have to come first :(
 
Last edited:
CaptainUnlikely said:
I've had to cut down on my DC drastically...rising costs of, well, everything, plus lack of income due to my gf's situation is biting hard right now so we're cutting back everything we possibly can. Sold almost all my GPUs (keeping some as spares and one for light gaming) and I'm in the process of stripping my rigs to sell on. I'm still running the NCI BOINC projects and DIMES, and MJU-12 is still running on my Atom which is hosted elsewhere, but other than that I'm out of the game for the foreseeable future. I'll return as soon as I'm able to but things like keeping my car running so I can work, and paying the rent, have to come first :(
Click to expand...

Hope situation will improve soon CaptainUnlikely...
 
Gilthanis said:
My nVidia GT430 returned its first work unit.
Run Time (akd Elapsed): 48,961.77 CPU (aka GPU) Time: 48,175.01 Points: 6,500.00

nVidia GTX650 returned its first work unit.
Run Time: 51,315.25 CPU Time: 42,831.65 Points: 6,500.00
Click to expand...

I refuse to crunch those POEM gpu work units. My 290s take 6 hours to do one and they give out 6500 points. Not worth the effort.
 
Gilthanis said:
My nVidia GT430 returned its first work unit.
Run Time (akd Elapsed): 48,961.77 CPU (aka GPU) Time: 48,175.01 Points: 6,500.00

nVidia GTX650 returned its first work unit.
Run Time: 51,315.25 CPU Time: 42,831.65 Points: 6,500.00
Click to expand...

MSI AMD 7970 Twin Frozr 3 running stock and 1 wu at a time
Run Time: 34,374.65 CPU Time: 26,940.17 Points: 6,500.00
 
Some people may be affected by POEM's multi GPU issue. Here is how you fix the problem.

In boxes where you have more than one nVidia GPU, you will want to create a cc_config.xml if you haven't already. If you already have one made, you can just edit the one you already have. Below is an example of what will need added:

<cc_config>
<options>
<use_all_gpus>1</use_all_gpus>
<exclude_gpu>
<url>http://boinc.fzk.de/poem/</url>
<device_num>0</device_num>
</exclude_gpu>
</options>
</cc_config>

The device_num should be the number assigned to the gpu that you don't want to be used by POEM. In this case, it is device 0. Now if you have multiple cards beyond 2, you just copy the text multiple times. Each time you just change the device number. For example, if I have a box with 3 GPU's in it and you only want GPU1 crunching POEM. It would look like this:

<cc_config>
<options>
<use_all_gpus>1</use_all_gpus>
<exclude_gpu>
<url>http://boinc.fzk.de/poem/</url>
<device_num>0</device_num>
</exclude_gpu>
<exclude_gpu>
<url>http://boinc.fzk.de/poem/</url>
<device_num>2</device_num>
</exclude_gpu>
</options>
</cc_config>

Additional configuration tips can be found here: http://hardforum.com/showthread.php?t=1729016
 
Last edited:
You must log in or register to reply here.
Back
Top