Posts by Zydor

41) Message boards : Number crunching : 7970s Dont Run (Message 2337) Posted 20 Jan 2012 by Zydor Post: Nice One - that'll be interesting having another mainstream app for the 7970, any idea what its written in etc ? Needless to say ... but if you need some pre-alpha testing PM me, I'll help. Thats welcome news - Mucho Gracias :) Regards Zy
42) Message boards : Number crunching : DNET Fragmentation - When is it likely to End? (Message 2321) Posted 20 Jan 2012 by Zydor Post: ...... I couldn't easily find any stats page listing those keyspaces, so the information is probably something just an D.Net admin can check. Good Luck with that ...... they are not renowned for being the most responsive of organisations. They seemingly believe that emails are for decoration, and user questions a distraction from the job in hand .... try as I have, I have never succeeded in getting any kind of reply to any topic, never seemed to have latched onto any system of communication where I can get any response, let alone answers :) I cant even find any kind of (active) Forum to post on. Wierd .... Regards Zy
43) Message boards : Number crunching : App_info..xml (Message 2319) Posted 20 Jan 2012 by Zydor Post: The application number is shown inside the Stderr for completed WUs as: Distributed.net Client v1.03 (ati14) Try changing: <version_num>130</version_num> To: <version_num>103</version_num> (dnetc_wrapper_1.3_windows_intelx86__ati14.exe is correct, dont change that) Regards Zy
44) Message boards : Number crunching : 7970s Dont Run (Message 2289) Posted 17 Jan 2012 by Zydor Post: He very well may do .... Collatz has always been the premier sea of calm inside the BOINC maelstrom. Nice Project to go to if you want a respite to recharge the 'ol BOINC batteries before launching back into the tempest of BOINCLand :) He runs a good Project, so guaranteed if he goes for one, it will be a solid good performer. He's a stand up guy who calls it how it is .... that defenitely works for me ... I detest the ego driven hype syndromes that fly around. Regards Zy
45) Message boards : Number crunching : Project Reset still giving out 1.02 (Message 2280) Posted 16 Jan 2012 by Zydor Post: Okie Doke ... Thanks Teemu Regards Zy
46) Message boards : Number crunching : App_info..xml (Message 2279) Posted 16 Jan 2012 by Zydor Post: I was wanting to increase the "<count>1.0</count>" to 2.0 because at 1 @ 4gpu's I am running 4 units at once, If it would work at 2 @ 4gpu's I should be running 2 units. On Poem I run .2 to have 5 at a time per gpu, and only use one gpu per card, the other 2 gpu's either run collatz or milkyway. It seems the cound it the numer of gpu's each task uses (as far as boinc knows), so telling boinc moo uses 2 gpus - in a 4 gpu system it should start 2 units. The Moo app is designed to be multi threaded, and will use all gpus it finds - designed to do that. If you want to try one of more gpus from Moo per core, then the app_info switch is the mechanism as outlined above. It is not designed to mix and match WUs from different projects at the same time, if it does so, great all is well, if it doesnt, thats that, wasnt designed to do it the first place. Any other useage or perception or wish list will have unpredictable effect as its not designed to do it. Pays your money and takes your choice as they say :) Regards Zy
47) Message boards : Number crunching : 7970s Dont Run (Message 2272) Posted 16 Jan 2012 by Zydor Post: I would say external to Moo & Milkyway, they will all be AMD/OpenCL with all the current speed issues of the new OpenCL implementation at AMD, see what GPUGRID come up with. Of all the Projects - from a development view - GPUGRID will I suspect come up with an AMD winner as they have been planning and aiming at the GNU application for many months. The unknown is how that GNU architecture will perform in the very complex GPUGRID environemnt given the shift AMD are making - could be GPUGRID end up shelving it until software matures more, way too early for earnest specualtion, too much water to go under the bridge. I dont think they will, they have planned on a native OpenCL app from the start, see what happens I guess. Regards Zy
48) Message boards : Number crunching : App_info..xml (Message 2270) Posted 16 Jan 2012 by Zydor Post: Suspending the test of this until I see what 1.03 can do, no point going ahead if 1.03 solves the ills, albeit the change log doesnt seem to get to these issues (not surprisingly). Regards Zy
49) Message boards : Number crunching : Project Reset still giving out 1.02 (Message 2269) Posted 16 Jan 2012 by Zydor Post: With the release of 1.03, I reset the project, and the server was still sending down 1.02 files. Have I jumped the gun a bit, or is there a hitch on 1.03 files coming from the server ? Regards Zy
50) Message boards : Number crunching : 7970s Dont Run (Message 2268) Posted 16 Jan 2012 by Zydor Post: Its difficult to know precisely the situation, but its fluid enough for me to rethink 7990 if the OpenCL and CAL question has not settled down by then. The 2x7970s I have now were for more than BOINC reasons, so I'm fine with that, but my main box is pure BOINC, different ball game and the reason I did my second box first. The new GPUGRID AMD app is a particular focus of mine, if that seems ok, and hope sounds are starting to be heard in these (very) early days on that one, then I'll probably stay AMD/7990 So we'll see how it pans out, but if the software side of 79XX relating to BOINC still looks problematic end of March, then I am heading for Kepler as long as NVidia come up with the goods ..... 2012 gonna be fun ..... :) Regards Zy
51) Message boards : Number crunching : 7970s Dont Run (Message 2264) Posted 16 Jan 2012 by Zydor Post: We may find that newer driver revisions include CAL support as a legacy app measure, and this is just deployment driver (they are after all focused on gamers and OpenCL/CAL is not an issue in their Prime Market). Somehow I reckon thats more hope than reality :) I think your right about upstream, they will push back on an OpenCL app until they see AMD clients going walkies, which will take a while - circa 6 months, but walkies will happen if this CAL situation remains extant as existing apps will not have CAL support at AMD developer level, and inevitably degrade over the long term. A possible bright spot on the horizon is (maybe?) all the additional tools and software tools that GCN allows to be brought into play .... maybe .... that will help, I have no idea as I am not a programmer at that level. The brave new world of OpenCL rofl :) Regards Zy
52) Message boards : Number crunching : App_info..xml (Message 2263) Posted 16 Jan 2012 by Zydor Post: Its looking like a runner ..... its very hard to quantify as there are so many factors. However I reckon overall its about a 10% gain as an overall average. Still doubtfull in some respects, however it does look ok .... so its worth trying for longer. I'll give it an extended go for another 2 or 3 hours and see how it goes. For those running twin 5970s, watch the fourth GPU like a hawk. It is running about 4 degrees or so above norm, in the end I downclocked the cards GPUs by 10Mhz to aim off for it. Its fine, but just be wary of the fourth GPU as always, the heat design bug is still lurking as an issue to bare in mind. Dont let that put you off, just keep an eye open for the first hour or so until you are happy its settled down. Regards Zy
53) Message boards : Number crunching : App_info..xml (Message 2259) Posted 16 Jan 2012 by Zydor Post: ..... and the result of the Hampshire Jury is ..... a rematch :) Total recorded cpu time was 8856, equates to 2214 for the batch of four, which makes it 36.9mins, or an average of 9.22 mins per WU. Thats pretty well what I am doing them at present with fragmented WUs. Usually I am at between 8mins 45 and 9 mins 20 fragmented (7mins 45secs unfragmented with 12 Stat Unit Groups) - mega fragmented and lots of stats units going to 10mins plus. So the figure to hang the hat off was an average of 9.22 mins, little change to now. All had slightly varying stats units, and differing fragmentation levels, so averaging is as good a way as any. I'm doing the next batch separated by 5 mins per WU, that will take time to settle down, so probably will not know the results of that for about an hour realistically. So far its no great increase, but lets see what the staggered start does. Meanwhile anyone wanting to try it, you will not lose anything by doing so. Keep your eye on temperatures, mine behaved ... but you never know, not run many like this so far. When the app_info is made it has to go inside the moo project directory, close down BAM completely stopping running apps, then reopen it and it should run. Make sure cache is run down as an app_info, will at startup, clear out all existing pre-app_info WUs without fear or favour and mark them as communication errors - its a bit pocessive rofl :) Back in an hour or so :) Regards Zy
54) Message boards : Number crunching : App_info..xml (Message 2258) Posted 16 Jan 2012 by Zydor Post: Usually that switch in the app_info controls the number of gpus running. Normally .... its used like this .... 0.5 would equal 2 per gpu 0.33 would equal 3 per gpu 0.25 would equal 4 per gpu Do the maths - you'll get the idea Usually no point whatsoever going beyond 2 ...... extreme circumstances maybe three, but very rare, above three is just silly. Usually 2 does the job. I also found it would not load two for some reason, threw http error back at me, maybe co-incidence, dont know. Anyway its running with one per gpu at present - about halfway through, so far so good. I'll post when done. Regards Zy
55) Message boards : Number crunching : Moo doesn't start tasks with BM 7.0.8 (Message 2254) Posted 16 Jan 2012 by Zydor Post: Try a Project dettatch/reattatch (not reset), that will send down a freash load of Project files for the current configuration. It should work ok then. Post again if it doesnt. Bare in mind that you are running BAM 7.XX which an Alpha release, so there is a real possibility that wacky things will happen. With Alpha releases life is rarely what it may look like on the surface :) Regards Zy
56) Message boards : Number crunching : App_info..xml (Message 2252) Posted 16 Jan 2012 by Zydor Post: Interesting result :) I suspect where the improvement is coming from is the last stage of a WU where its at around 99.7% complete, then takes time to collate the result - on mine that takes 2 mins or so. If I pause all the moo units, and let what is running finish, I can run one at a time, with many large drops below 70%, So I can see where running 2 at a time would help The large drops below 70% are as a result of the Stat Unit Fragmentation, the standard Moo WU with 12 Stat Unit Groups runs at around 98-99%. That changed when the fragmented units started coming through, its still the same multi thread, multi-gpu application, but it struggles with fragmented files. In the normal course of events there is very minor effect using an app_info as the Moo WU has very high utilisation. Its looking like an app_info is taking up the slack caused by fragmentation. Your result of 97-99% is back to what the Moo app always provided on multiple gpu's. So step slowly putting the next bit into play .... If you want more than one WU per GPU change the lines .... <type>ATI</type> <count>1.0</count> to read ... <type>ATI</type> <count>0.5</count> and that should startup 2 per GPU. Watch your heat though, Moo is normaly above average in heat output, albeit in acceptable levels. Its looking like the app_info with one per gpu is getting round the fragmentation effect, and returns to normal behaviour of 98-99% utilisation, so its possible two per gpu is going to make the gpu work real hard for little gain and much increased heat. Will only know by trying two per gpu, but watch temps, would hate to see 6990s burn out ... I'll run down my cache, and give it whirl on one per gpu, see how goes, I have only got a 2.5hr cache set as I was running it down getting ready for the new Moo App, so I should get a result by midday at the latest with one per gpu. Regards Zy
57) Message boards : Number crunching : App_info..xml (Message 2250) Posted 15 Jan 2012 by Zydor Post: Set your cache to zero, then do a project detatch/reattatch (not reset - must be attatch/reattatch) which will download a fresh set of Project files. See how the first ones run, if ok, set your cache to your normal value, and you should be fine. Downloaded WUs seem to be running fine, and the server is up and solid. It can happen there are database issues on the server, but more likely something has gone wrong with your Project files, the attatch / reattatch should resolve that. If it doesnt post again. Regards Zy
58) Message boards : Number crunching : App_info..xml (Message 2246) Posted 15 Jan 2012 by Zydor Post: Not sure of my ground here, but Moo!, like DNETC, only allows 1 WU to crunch at a time using all available GPUs simultaneously. Is that correct Zydor? Yup - thats by design as the core DNETC app is mult-threaded. Not much point trying to negate the benefits of muti-threading by doing it yourself with multiple apps. Ends up a heath robinson affair for little gain. Sometimes it will work forcing (say) one app per gpu and you get an increase .... well, fine ... but in the long run, with multi-threaded apps, leave it to the application - in the long run its much better at it than you or I :) Regards Zy
59) Message boards : Number crunching : Size matters? (Message 2242) Posted 15 Jan 2012 by Zydor Post: Latest "informed souce" (define informed source rofl) puts 7950 release at 31 Jan. I would expect it to go for around $300-$350 on first release, price will come down from that within around 2 months depending on sales an NVidia moves. Maybe cheaper on release depending on AMD perception of NVidia moves to market with their 28nm cards. If you still want to buy now, this may be a good source for you (had a search around, it was the cheapest I could find) $210 after redeaming the mail in $30 rebate ticket. From the ubiquitous Newegg :) http://www.newegg.com/Product/Product.aspx?Item=N82E16814150523&nm_mc=AFC-C8Junction&cm_mmc=AFC-C8Junction-_-na-_-na-_-na&AID=10440897&PID=3891137&SID=rewrite Regards Zy
60) Message boards : Number crunching : Size matters? (Message 2238) Posted 15 Jan 2012 by Zydor Post: Hi there. I'm probably going to pick an HD6950 in the coming days to replace my old 4850. Is there an advantage to pick a card with 2Gb of memory rather than 1Gb? There can be - it depends on your monitor setup. If you do - or aim to do - gaming at the extreme end aka triple monitors 2550x1440, then lots of card memory is the order of the day (hence the 3Gb on the new 7XXX range of cards). If you dont have such needs (or only one monitor at such high res), its - 99% of time - not an issue at all staying at 1Gb card memory Currently I can only run Moo! in the 4850 (get errors in the other projects), but I intend to switch the other projects (Prime, MilkyWay, Collatz) back on. Does larger memory size help with the "Leave applications in memory while suspended" option on? Possibly ..... hard to be specific as the devil is in the detail ...... but again in 99% of situations 1Gb is fine and you will not notice the difference with current applications. Could it happen there is an issue and 1Gb gets crowded out .... yes it could, but at present you would look to the sky to find the new Star that has appeared. As applications change memory needs change, so who knows re the furture. For known forseeable needs, I would be amazed if you went over 1Gb in such scenarios. If you do, its not the end of the world and other things can alleviate the situation. By in large, grahpics memory is there for screen resolution and manipulation, its only niche lunatics like us at BOINC that use it for other things, and the applications dont take up nearly as much card memory as high end screen resolutions. IGb will be fine. Dont forget 7XXX is deploying - real hardware not vapour wear, and its a very much faster than 6XXX. Some trips and traps to the cards and useage, but worth waiting for a few weeks util the 79XX and 78XX range becomes clear in spec terms. Guaranteed its way way faster than 6XXX, and far more power efficient, just need to confirm specs and software aspects, and its then just which do you buy. Regards Zy

Previous 20 · Next 20