|
21)
Message boards :
Number crunching :
Current release BOINC version is 6.10.56
(Message 16961)
Posted 5117 days ago by  Paul D. Buck
53 is now out with a few fixes to the download quota stuff. It seems this may be a release candidate as UCB are pushing to get a 6.10 client out of the door. Yes they are... sadly though it is going to flop worse than the .43 version because there is an unfound issue which causes the late .4x versions to suddenly stop assigning GPUs to work. The messages in the logs are uninformative though the indications point to a flaw in the routine that determines the memory size of the card ... I got hit with it multiple times with .46 and .48 during testing... very depressing to find your GPUs idle for hours while you slept ... I even saw it once on .45 though I am still using that version on all my systems (but the Mac which is using .43) ... to date there has been no change set to address this issue ... so, I worry that once again UCB is rushing things without looking very hard for a significant bug ... though at least there is acknowledgement that an issue is there ... |
|
22)
Message boards :
Graphics cards (GPUs) :
GPUGRID and ATI
(Message 16960)
Posted 5117 days ago by Paul D. Buck
The SDK2.1 is better but still has several bugs in parts which are required to run the code reliably. We must wait for the next release. Thats is good news once again ... and if they are helping, maybe they will see first hand the problems with the SDK and that will improve the chances that that will be fixed as well ... Thanks for the update ... |
|
23)
Message boards :
Number crunching :
BOINC 6.10.43/44 pulled
(Message 16943)
Posted 5118 days ago by Paul D. Buck
You can still download BOINC 6.10.43 64bit There are no fixes yet to the issues that drove them to remove the builds... to put it another way, .53 solves none of the issues that supposedly made .43 bad ... at least not yet... this was merely a build to fix issues with the download / upload limiter code that they added post .45 ... This is why, when you settle on a version, keep a copy of the installer so that if you do want to reinstall a version you can use your own copy ... I keep upwards of 20 copies "just in case" ... heck, I just recently (in the last month or so) dropped 6.5.0 ... :) |
|
24)
Message boards :
Graphics cards (GPUs) :
GTX 295 Errors
(Message 16848)
Posted 5122 days ago by Paul D. Buck
THe most common problem with the second card not being picked up can be one of two things, one is the need for a dummy plug (should not be needed with a 295, I did not use one on mine, but it can be needed, or a second monitor) the other is not extending the desktop to the second card. Or, as already noted a bad driver install ... usually you should see the install actually do the same actions two times ... Oh, and if you are going beyond 6.10.18, the current recommended, I suggest only 6.10.45 which is the one I am using ... .46 and later have severe GPU problems that have not been addressed in any changeset I have seen to date ... so, .51 has no corrections I know of to the issues I reported (as did others) with the post .45 versions ... |
|
25)
Message boards :
Graphics cards (GPUs) :
GPUGRID and ATI
(Message 16788)
Posted 5125 days ago by Paul D. Buck
One ATI HD5850 and a HD5670. Ugh, I would expect based on other projects that those numbers would be more like reversed ... :( Still, as you note, still room for improvement... Looking forward to being able to try it ... I seem to be down to only one card on GPU Grid in that I put the GTX295 card back into rotation but all the tasks seem to be failing ... :( So, down to only one GTX280 card ... which on DNETC, Collatz and MW is considerably slower than the HD5870s which are my current backbone... |
|
26)
Message boards :
Graphics cards (GPUs) :
Slower GPUGRID processing in Windows 7
(Message 16754)
Posted 5126 days ago by Paul D. Buck
Please, no long rambling replies that neither address the problem, deny the problem or that try to mask it with irrelevance. How about short replies that do any of the above? |
|
27)
Message boards :
Graphics cards (GPUs) :
GPUGRID and Fermi
(Message 16750)
Posted 5126 days ago by Paul D. Buck
The task on GPU Grid don't use the CPU to do useful work, increasing this percentage will not substantially make the task run faster... mostly the CPU side is a "helper" thread to load data onto the GPU and remove results ... unlike SaH's ATI application or Einstein's CUDA application where substantial processing takes place on the CPU side ... |
|
28)
Message boards :
Graphics cards (GPUs) :
GT240x2
(Message 16747)
Posted 5126 days ago by Paul D. Buck
Assuming I am the Paul referenced...:) None of my GPUs are over-clocked beyond what they were set to by the MFGR. Part of that is philosophy (since I am interested in accurate science I have trouble with the idea of running my system's components in potentially unstable portions of the "flight-envelope" of that component), and the other is quite simply heat ... I already have a situation where the room gets too hot as it is ... OC would just make a bad situation worse ... So, it is quite probable, as ETA notes/implies, that I could get higher performance as well had I chosen to do so ... So, the comparison is of OC vs. stock ... let the argument continue ... :) |
|
29)
Message boards :
Graphics cards (GPUs) :
Power Consumption?
(Message 16738)
Posted 5126 days ago by Paul D. Buck
Direct but not necessarily linear ... :) In some cases the speed goes up by two, power by four ... It also depends on how you get there from here ... Adding a GPU project will just about double your computer's power draw by a mid to high end card, say a GTX260 (or better, rated at about 170W pull, GTX295 at 312W, and so on), but the CS production will go up by a factor of 4-10 (or more) ... as the other side of the coin ... For example I probably make from 2-40K a day from my CPUs running a lot of NCI projects for which the pay is great for the CPU load involved ... but I also run standard CPU projects at the same time earning about 1-5K on the CPU side per computer... but, I am making about 700K a day total because of the 7 GPUs I run ... It has been so long since I have been running without a GPU in the mix I have no idea what I make on pure CPU though I could make an estimate ... Lets see, for the numbers in BS as of right now, NCI Projects: 13,630 (FreeHAL and WUProj) GPU Projects: 668,081 Collatz, DNETC, MW ----------------- Total: 681,711 Daily Total: 705,173 Delta = 23,462 So, I have 5 computers 3 i7s, one quad and a 8 core mac earning about 5K each ... all total, about 3.3% of my daily earnings comes from CPU work ... 96% comes from the 7 GPUs ... Of course the flip side is not all project's programs are amenable to what we call vectorization which is what GPUs do best, process vectors of numbers ... if the programs are, then, well, Bobs your uncle and you can run the application 60-100 times faster (or more) on a GPU than on the CPU ... but, most problems are not vector problems and so, GPU processing is out ... this is one of the reasons the Cray computers, as wonderful as they were did not monopolize the super-computer marketplace in their day ... there are some problems that would not run efficiently on the architecture ... but for those that did ... they ran like a scalded cat ... On the other hand, the more projects that we can get to use GPUs frees the same cores now doing the work on the CPU when it is shifted to the GPU ... personal example, I don't run MW or Collatz on the CPU at all ... why bother running a task for hours on the CPU when I have GPUs that do that same task in a minute and change (MW) or a few minutes (less than ten mostly, Collatz)... so, those CPU cores run other projects that don't have GPU applications yet, or never will ... Oh, and I have GPUs that with about the same power draw do work at a rate about 4x apart, a GTX280 for example against a HD5870 are about the same power draw but the ATI card is about 4 times faster in doing the work ... of course the ATI card does not work at all here (yet) so, faster yes, always usable, no ... tradeoffs abound ... Heck if it was simple or easy, anyone could play ... :) |
|
30)
Message boards :
Graphics cards (GPUs) :
GPUGRID and ATI
(Message 16735)
Posted 5126 days ago by Paul D. Buck
A big Plus, i'd say! Been running not quite 24 hours ... going to take a couple of days to see a difference in stats I know ... and with the "pulsing" in the numbers it is hard to know where you are till you get at least a weeks data or better a month ... then you can average it out ... today's update had me at 705K for the day, not bad, but less than yesterday's 720, or the day before's 790 ... My pending on Collatz is at an all time high though ... so ... that is good ... making money on DNETC and they seem to be making small changes over the batches so maybe we can close in on the problems they are seeing there ... I have been getting errors and hard to know why, asked for more logging and hours later I have a bunch of results with nice logs ... Not sure if it is going to help with the "hang" problem though, if the task hangs and gets aborted if they just opened the log once it might be still open when the kill comes in and the log data will be lost ... if they do an open, append, close for each statement we might get some where ... though I saw a synch line in there that worries me ... it waits for the GPUs to end ... what if the GPU does not know it has ended, or sent the signal already? |
Previous 10 | Next 10
