Posts by th3

1) Message boards : Number crunching : do you NOT overclock? (and crunch 24/7) (Message 63951) Posted 4 Nov 2009 by th3 Post: I'm only doing Einstein at his time so i cant contribute much here but still: Core i7 920, stepping D0, great at stock and can go really high if you one day decide that overclocking is wise (coz it is), even without voltage increase D0 is great if you dont want to overvolt.... Overclockable i7 mainboards and DDR3 isnt expensive anymore and even cheaper if you are sure you will never OC. And dont listen to that crap lots of ppl like to repeat like parrots about "max 1.65v vDIMM", i never had my stepping C0 920 that low on the vDIMM and now its just a few weeks left till it been crunching various projects 24/7 for a year. I guess i5 is nice too, but 2x2GB RAM isnt much for Rosetta, so for the memory its cheaper to go i7 with 3x2GB than 4x2GB for i5.
2) Message boards : Number crunching : Memory timings.... (Message 61203) Posted 15 May 2009 by th3 Post: [quote]Does RAM timings affect R@H much? I lowered mine to 4 clocks (was on 5, which is a bit high) It helps for most things, just dont expect too much, 2-3% is a good improvement when talking timings. 4-4-4-18 is quite loose timings, try 4-4-4-12, or 4-4-4-10, or even better if you get it even lower. Cant expect the 3 first ones to go lower than 4 without bumping up the voltage, so just try to lower tRAS. Can even try as low as 4 without going to extremes, but then you are probably looking at more voltage, depending on which chips your RAM sticks uses. With Micron DDR2 chips i always managed stable at 4-4-4-4 or 4-4-4-5 at 900MHz and above. Atlon64 doesnt always respond much to tighter timings, you may have to look at subtimings.
3) Message boards : Number crunching : Error code 193 for nearly all WUs on Linux 2.6 (Message 60030) Posted 8 Mar 2009 by th3 Post: Nobody has this problem? Not on rosetta and not that error code. I did however have all WUs fail on 2 other projects with newer kernels. I think the problem was nvidia chipset for Intel processors, the other PCs, with Intel chipsets and the same kernel, had no problems. Maybe yours are not nvidia but thats all i got, sorry. Edit: to be more specific, it was nvidia chipsets with integrated graphics (nf630i/gf7100)
4) Message boards : Number crunching : Low granted credit (Message 60002) Posted 6 Mar 2009 by th3 Post: Rather interestingly, my RAC is picking up now even though I'm using the old client. Looks like whatever I did, it's working :) Or you didnt have to do anything? :) My Core i7 had problems overtaking another Core i7 in the top hosts list a few weeks ago, it stayed behind for days, RAC was then around 3900, before overtaking and continued to climb to above 4100 before leveling off, the other i7 stayed at around 3900 and is still there. I believe its like mod.sense said here (and dcdc), "credit is not an absolute thing. It is based on how difficult all the other machines out there are finding it to produce models for a given task" - and thats always changing and the RAC change with it. R@H credit system is imho inaccurate but its about as good as it gets without quorums,
5) Message boards : Number crunching : A SINGLE Graphics Card = 1 teraFLOPS (Message 59875) Posted 28 Feb 2009 by th3 Post: you ARE a Penguin after all NOT a bat!! He is a very special penguin, a Windows penguin: http://boinc.bakerlab.org/rosetta/hosts_user.php?userid=89694&show_all=1
6) Message boards : Number crunching : Enlarge Boinc manager window at start-up (Message 59771) Posted 24 Feb 2009 by th3 Post: I tried it again, and the setting does not stick if the manager is open while editing the file. Close the manager, edit and save the file and it works. Using Debian Lenny with KDE. To install boinc as a daemon on debian based distros: "sudo apt-get install boinc-client boinc-manager" or use the package manager if you prefer GUI. It will probably downgrade your boinc-version so if you need the latest client for GPU crunching and so on it requires you to manually copy a newer binary to /usr/bin
7) Message boards : Number crunching : Enlarge Boinc manager window at start-up (Message 59733) Posted 22 Feb 2009 by th3 Post: Theres a config file for boinc manager in your users home folder, unhide hidden files and edit the file ".BOINC Manager", edit the values Width and Height, you'll find them close to the top of the file. You have to open the Boinc Manager to start crunching? From your hosts list it looks like a Ubuntu kernel and since that is Debian based its better to just use the repos to install Boinc as a daemon, it will then start crunching at boot, no need to open boincmgr or even log in.
8) Message boards : Number crunching : Intel i7 CPU (Message 59085) Posted 28 Jan 2009 by th3 Post: That do make sense, and i noticed just that yesterday when crunching some Rosetta Beta 5.98 alongside Rosetta Mini 1.48/1.54, coretemp was higher and the Mini1.48 tasks ran ~20 minutes faster. Another thing i noticed is that Rosetta is considerable more memory intensive than E@H and QMC, that also helps explaining why it doesnt stress the execution cores as much.
9) Message boards : Number crunching : Intel i7 CPU (Message 59026) Posted 25 Jan 2009 by th3 Post: Moved my i7 to Rosetta now recently, somewhat disappointed with the Linux app. It doesnt stress the i7 anywhere near what QMC and Einstein does, dont know if its because Rosettas Linux x86 app is less efficient or if it is less suited for hyperthreading, but i do know the i7 is running significantly cooler now. With Einstein the tubes for the WC system used to be quite warm, almost hot, now barely warm.
10) Message boards : Number crunching : Discussion of the new credit system (Message 24775) Posted 24 Aug 2006 by th3 Post: The new system is Lotto. My last 2 results: CPU-time: 10,061.41 GC: 68.19 CPU-time: 9,951.78 GC: 105.76 So much more points for less work? This been discussed up and down already but users need credits based on how much CPU-time they contribute, when the variation can be as big for same CPU as what i just experienced then something is seriously wrong. "Even out over time", wish those who wrote that also would realize how mongoloid that sounds. I understand it wasnt possible to stop the cheating with the previous credit system but this is a poor substitute. Sorry guys, i cant accept that someone will decide how much work my PC has done, this IS the computing power i contribute and i expect to receive credit based on that, not based on what some other computer estimates, then that computer can do my WUs instead and see how much that will benefit the project.
11) Message boards : Cafe Rosetta : Lets put Philippines on the Rosetta map! (Message 22464) Posted 15 Aug 2006 by th3 Post: Team Philippines is reaching for RAC top20, so if you are Filipino or have Filipino blood or Filipino relatives/friends, or if you visted Philippines (like me) and fell in love with the beauty of the islands, or if you would like to go there some day or whatever your connection to Philippines is: Join us in Philippines I'm throwing in one Core2Duo and one Pentium D (overclocked to the max ofcourse) to set the wheels of Philippines in motion, but we need help to keep rolling. Please leave a short reply here if you join the team.