BOINC

Doesn't seem to make much difference if I do or don't leave a core free. I just suspended NFS for one Milkyway unit start to finish to test it - no problem while the unit was running, then as soon as it finished and there was a few seconds of low/no GPU activity, as I was scrolling a web page I got crazy flickering again, as soon as the unit started and the GPU load went back up it went away. Just paused Milkyway and without any GPU load, the flickering is still present when scrolling or changing window focus, too.
GPU isn't overclocked, this Win7 install is pretty recent, and I've only ever had this one driver version on here.

After a bit more Googling it seems like the issue is related to the idle/2D clock speeds which I could fix with a BIOS flash - basically edit the 2D clocks to be the same as the 3D clocks. I might have to do that because this is silly, and incredibly annoying how it's only come up again on this card after changing GPU projects, looks like just a coincidence though.
 
Glad you have a clue to go off of. I just recently swapped my living room TV (Optiplex 360) computer with my i7 920 using the same video card. My dual core ran flawless while crunching full bore on both cores when I streamed Netflix in HD. However, the 920 wasn't has happy. I had to free up one thread and I'm not positive that I don't need to open another one.
 
I've had no problems running Netflix with 12 threads of NFS and Collatz on my L5640 ES, going to watch a film shortly so I'll see if it's still happy with Milkyway, or even without anything running now it's doing this flickering thing. I think it's just encouraging me to crunch more - "not running anything? 0% GPU load? What's that about? Here, have some terrible flicker until you start running again, that'll show you!"
 
48 hour NOTICE: FMA3 LLR going live.
Sometime during Thursday, April 17th (UTC) we are going to install the new FMA3 (aka AVX2) version of LLR on the PPS project, in preparation for the Shakespeare's Birthday Challenge that starts on Sunday.

Especially on Intel Haswell CPUs, but to some extent also on Ivy Bridge and Sandy Bridge CPUs, this new app is significantly faster. It also consumes more power, and generates more heat. Please make sure your computer is up to running the new app.

Information and discussion about the new LLR app can be found in this thread.

I expect to install the new app for all of the other LLR projects sometime after it goes live on PPS, depending on how well the PPS deployment goes. 14 Apr 2014 | 20:52:00 UTC · Comment

http://www.primegrid.com/forum_thread.php?id=5642#75523
 
While we're on Primegrid, congrats to Grandpa for hitting 100,000,000 points on the project! He joins Kenlo as the only two with two separate BOINC projects over that big 100M mark...

(And a week before I hit 100M TOTAL BOINC points... :rolleyes: I need a server or two... :D)
 
I don't agree with that for a variety of reasons, though I suppose it does cover operating costs.
 
I would be ok with say 5% for operating costs...but 15% doesn't sit well.
 
Guys, I've had problems with BoincStats/BAM! over the last couple of days. I've specifically changed/reduced my WCG resource share on individual hosts. When I tell boinc to resync to BAM!, I see the following in stdoutdae.txt:
Code:
16-Apr-2014 07:06:00 [---] Contacting account manager at http://bam.boincstats.com/
16-Apr-2014 07:06:02 [---] Account manager: BAM! User: 159023, brilong
16-Apr-2014 07:06:02 [---] Account manager: BAM! Host: 494879
16-Apr-2014 07:06:02 [---] Account manager: Number of BAM! connections for this host: 185
16-Apr-2014 07:06:02 [---] Can't parse account in account manager reply: unexpected XML tag or syntax
16-Apr-2014 07:06:02 [---] Account manager contact succeeded

Searching for this error turns up posts from 2006. Any ideas what might be taking place or how I fix it? Thanks.
 

Not really. I have 39 clients. All of them have the XML error, but some of them sync properly to BAM! and others do not. I took one of them and rebooted it. Now it's syncing without the XML errors, but the resource sharing that it shows for WCG is different from what BAM! shows. In the past, I've changed BAM! settings from 0 to 50, update, then 50 back to 0, update and it finally works. Just wondering if other folks have had problems with host resource sharing changes not properly syncing.
 
I have had issues with BAM!'s resource share not sticking at times. Sometimes a sync with BAM! fixes it and sometimes it requires messing with the resource share option again. I have also noticed that I can not change more than one resource share per sync either. That is the only part of BAM! that I really hate messing with.
 
The Einsteinium Foundation http://einsteinium.org/is a crypto currency that 2% of all coins generated go to science projects nominated and voted for by the public. Many of the people who mine this currency also make large donations to the winner of the vote.

The last winner was a PHd working on a link between EB virus and cancer. The project funded his needed equipment.

I nominated GPUGrid for a possible recipient. I am sure when it comes time to vote the voters will want to know what would be done with the money if GPUGrid wins.

I do not know the details as to how the EMC coins are converted to cash or many other details. But this is if nothing else more exposure for this project. People who are interested in this coin are very science aware and interested in science projects. Maybe there is a particular project GPUGrid would like to try but lack funds.

Anyway I will post when the voting starts so everyone can go and vote. You can also make a nomination as well.

Check it out!
http://www.gpugrid.net/forum_thread.php?id=3720#36270
 
More suitable than einsteinium is gridcoin.
It rewards users for their contribution to boinc, this will hopefully drive a serious amount of computing power towards boinc.

www.reddit.com/r/gridcoin

And yet again as I have pointed out multiple times, this also requires you to join their team instead of ours....

I will also point out that it isn't more suitable since Gridcoin probably isn't trading for all that much and goes directly to the user instead of the project.... Pretty big difference.

Edit: I will also point out that if someone was really interested in Gridcoin, they could actually do both of these, so again I will say your statement is wrong because it is apples to oranges. The link I provided doesn't require mining at all but rather a simple vote...
 
Last edited:
Another 100,000,000 points congrats, this time to Brilong! You beat me to it by about 2.5M points - well done!

We need a BOINC equivalent of the [H]orde's pimp hat over here... :D
 
And yet again as I have pointed out multiple times, this also requires you to join their team instead of ours....

I will also point out that it isn't more suitable since Gridcoin probably isn't trading for all that much and goes directly to the user instead of the project.... Pretty big difference.

Edit: I will also point out that if someone was really interested in Gridcoin, they could actually do both of these, so again I will say your statement is wrong because it is apples to oranges. The link I provided doesn't require mining at all but rather a simple vote...
Yeah, it does require you to join the gridcoin team that's true.

The value of gridcoin is not always going to stay at it's current low value, just like bitcoin/any other crypto it'll fluctuate in value.

The fact that it goes to the user intead of the project it a good point, however a couple percent going towards the project being able to divvy this out to scientists is just their selling point for giving their cryptocurrency value. I'm sure gridcoin could have a volunteer donation system where the grc are given to scientists.
 
Yeah, it does require you to join the gridcoin team that's true.

The value of gridcoin is not always going to stay at it's current low value, just like bitcoin/any other crypto it'll fluctuate in value.

The fact that it goes to the user intead of the project it a good point, however a couple percent going towards the project being able to divvy this out to scientists is just their selling point for giving their cryptocurrency value. I'm sure gridcoin could have a volunteer donation system where the grc are given to scientists.

Point is this is [H]ardOCP team forum not the Gridcoin forum, it appears you are recruiting for the Gridcoin team and that is generally not much appreciated on a team forum page. We all know about Gridcoin and most of us are not interested in getting paid we are here as a donation to humanity we are not here for profit, nor do most of us give a rats behind about it. The projects I and most of the others here fold / crunch get 100% of our donation not 2% or whatever % they are giving.

Anyway it would probably be a good idea to quit promoting the Gridcoin thing here before Gilthanis looses his temper. :)
 
Just for sake of history, it was first brought up here by me and mkrohn discussing Ripple Labs before there was a separate forum to discuss crypto currencies: http://hardforum.com/showthread.php?t=1795940&highlight=gridcoin

Then you tried to recruit 402blownstroker here: http://hardforum.com/showthread.php?t=1812735&highlight=gridcoin which I reminded you of typical etiquette in a team forum and even gave musky a link when he questioned whether I was right about switching teams as being mandatory.

And now you bring your pitch again into our team's forums within this thread. If you want to recruit for Gridcoin's team, please go to the crypto currency threads where you will find more people who have those interests. Here is a handy link in case you forgot about it: http://hardforum.com/forumdisplay.php?f=137 I know this is the BOINC thread and Gridcoin is related to BOINC now, but this is [H]ard|OCP team forums. Not Gridcoins. Go there or tailor to those not already on a DC team here. There are plenty of other forums you can recruit from I would suggest going to the various projects where they actually have team recruitment threads.
 
Just for sake of history, it was first brought up here by me and mkrohn discussing Ripple Labs before there was a separate forum to discuss crypto currencies: http://hardforum.com/showthread.php?t=1795940&highlight=gridcoin

Then you tried to recruit 402blownstroker here: http://hardforum.com/showthread.php?t=1812735&highlight=gridcoin which I reminded you of typical etiquette in a team forum and even gave musky a link when he questioned whether I was right about switching teams as being mandatory.

And now you bring your pitch again into our team's forums within this thread. If you want to recruit for Gridcoin's team, please go to the crypto currency threads where you will find more people who have those interests. Here is a handy link in case you forgot about it: http://hardforum.com/forumdisplay.php?f=137 I know this is the BOINC thread and Gridcoin is related to BOINC now, but this is [H]ard|OCP team forums. Not Gridcoins. Go there or tailor to those not already on a DC team here. There are plenty of other forums you can recruit from I would suggest going to the various projects where they actually have team recruitment threads.

I was under the impression this was a distributed computing section of your forum, but I understand that the cryptocurrency section is more suitable. I brought it up in here because someone brought up einsteinium.
 
Last edited:
I didn't bring up einsteinium directly but rather a way to vote for a project to receive a free donation of it that didn't require actually mining at all. It was a simple vote click. There was also no need to switch teams...
 
We need a BOINC equivalent of the [H]orde's pimp hat over here... :D

I will second that :D

How about this? I got it from the tamale van guys.

7iaMREaiA.jpeg
 
Free tamale's?

For the power conscious of us here (not many) A 4770S with a 750 Ti with 7 threads running near 100% consumes 125W of power. Just thought I'd share now that I'm down to very few rigs (power concerns, can't upgrade power service!).
 
Dears, we have updated the badges to reflect two recent publications performed with your help:

* S. Doerr and G. De Fabritiis, On-the-fly learning and sampling of ligand binding by high-throughput molecular simulations, submitted (2013).

and

* P. Bisignano, S. Doerr, M. J. Harvey, A. Favia, A. Cavalli and G. De Fabritiis, Kinetic characterization of fragment binding in AmpC β-lactamase by high-throughput molecular simulations, J. Chem. Inf. Model., 2014, 54 (2), pp 362–366.

You should be able to see them in your user profile, as usual.

http://www.gpugrid.net/forum_thread.php?id=3735#36535
 
Does GPUGRID have long run wus that will run on my 750Ti cards yet? I'm still running Einstein on them.
 
Does GPUGRID have long run wus that will run on my 750Ti cards yet? I'm still running Einstein on them.

Yes they do. But keep an eye on it.

Edit: 5.5 issue I guess isn't an issue now, but some are noticing that the work units are back to using a full thread/core per GPU.
 
Last edited:
If anyone is remotely interested in BitcoinUtopia, they are testing a CUDA app. http://www.bitcoinutopia.net/bitcoinutopia/forum_thread.php?id=200#743

CudaMiner for 32-bit Windows (should work also in 64-bit) is available for testing. You need to enable test applications to get it.

The current workunits should run slightly over 7 minutes.

Henri.

And for the Maxwell users, beware:

Thanks!

It looks really interesting. I guess the current version of CUDAMiner is using CUDA 5.5.

Henri.

http://www.bitcoinutopia.net/bitcoinutopia/forum_thread.php?id=172#708
 
WCG has BETA work for Android devices. Get them while you can. :D
 
Last edited:
Has anyone experienced instability with the BURP boinc project?
I've had a few BSOD's whilst running boinc recently.
 
I've never had issues with it. It can be very heavy on RAM at times, I would do a diagnostic there to be sure. Any BSOD codes to reference?
 
I asked Carle over at QCN how long they expected to survive since he last messaged me and told me they probably only had a few months left. Here is his follow up response.

I think it will hang on at least a year and probably two, but future development and "survival" depends on getting funding past that.
 
Anyone here running GPUGRID on a 750Ti card? The long run wus seem to be taking twice as long to crunch on my 750Ti cards as my 560Ti cards. 21 hours versus 11 hours. Doesn't seem right.
 
At the moment, your long runs are in line with others with 750 Ti's, though it seems there are still some teething issues with getting it optimized for Maxwell. Hmmm, I just attached it to mine to see how they run. I'll let ya know!
 
Back
Top