Message boards :
Number crunching :
Do the WU get less credits now since thanksgiving crash
Message board moderation
Author | Message |
---|---|
Send message Joined: 30 Mar 09 Posts: 15 Credit: 15,856,582 RAC: 0 |
Hi, I am asking the question about the credits that are issue to WU sice November Crash Before the November crash I was stable at about 32,000 RAC, then after the thanksgiving crash it dropped directly to 18,000 and I slowly worked up to 28,000 and then there was the downtime for the new server and upon completion I had dropped down to 13,000 and so far I have climbed back up but only to 24,000 and it's dropping again. My PC's doing the crunching are still on for the same time periods as before the november crash but I just can not seem to crunch and get the same RAC. Do the WU now collect a smaller credit for completion. |
Send message Joined: 1 Sep 08 Posts: 204 Credit: 219,354,537 RAC: 0 |
Do the WU now collect a smaller credit for completion. They don't. Long term RAC is highly dependent upon the project being available all the time. RAC is averaged over.. I think about a month. So any downtime will stay in the stat for that long. In order actually reach your maximum RAC you'd need about a month of undisturbed crunching (on your side and on MWs). You can check the speed yourself if you devide the seconds per day by your time per WU. Multiply by the credits per WU and you get your maximum RAC. Mine hasn't changed since a long time. MrS Scanning for our furry friends since Jan 2002 |
Send message Joined: 30 Mar 09 Posts: 15 Credit: 15,856,582 RAC: 0 |
Thanks, but could you help me understand a bit more where to find the credits per work unit from. I have had a look in my MW account under one of my PC's that runs 24/7 and look at the tasks. I noticed several MilkyWay@Home N-Body Simulation v0.80 (mt) tasks and they on average ran for.. Run Time 34secs CPU Time 44secs and I got an average of 0.73 credits. See link below, I am host 299165 http://milkyway.cs.rpi.edu/milkyway/workunit.php?wuid=88896154 So from this data I calculate as you say.... 86400 secs per day divide by 44 (is it 44 or would it be 34?) = 1963 Then multply 1963 by credits per work unit which is 0.73 = 1432 RAC OK, so it my max RAC for my host 299165 is 1432 but it's actual RAC is 749. So for host 299165 it's RAC before the Thanksgiving crash was at 1396, after the crash it dropped to 615 and then it slowly built up to 1226 by 11 December when the Server was replaced then the RAC dropped down to 425 and since then it has only managed to creep up to 749 over the last 7 weeks. This PC is on 24/7 and does nothing apart from MW cruching. It is a 32but Win7 PC on a Core Duo T2080 @ 1.73Ghz So you can see why I am asking about the RAC dropping that or my PC is shagged :-) |
Send message Joined: 8 May 09 Posts: 3319 Credit: 520,313,538 RAC: 20,660 |
Thanks, but could you help me understand a bit more where to find the credits per work unit from. Actually if you look here it might help: http://milkyway.cs.rpi.edu/milkyway/host_app_versions.php?hostid=299165 That is YOUR pc and how it is doing, it shows you doing multiple KINDS of units per day. Some units will give more credits, some will give fewer credits. To just use a simple formula like you did will never give you an accurate picture of what your pc is doing day to day, or even over the long term. What if today you run 10 higher credit units and only 1 lower credit unit, and you continue that for the next 3 days. Then you start getting a 50/50 mix of the the two kinds for 1 week, then the pattern changes, your average is so messed up it would take a pc just to try and figure it out! The key to a regular per pc rac is to run only one kind of unit and to keep the pc from running anything else at the same time. Essentially a Boinc only machine, but then you get into cost effectiveness and what are your goals. Me personally I have several Boinc only machines, but I like the results! |
Send message Joined: 1 Sep 08 Posts: 204 Credit: 219,354,537 RAC: 0 |
Sorry, I assumed you were talking about the "separation" type tasks. I'm not running the nbody-stuff, so I can't tell whether credits have changed there. Otherwise your calculation is in principle correct. You'd want to use "run time" rather than "cpu time", as nbody uses both of your CPU cores. And to a varying extend. For the longer WUs "run time" approaches half of "cpu time" - which would be perfect scaling on a dual core. For the smaller WUs (which you used for the estimate) the speed-up due to the 2nd core is less. So, as Mikey said, your actual RAC will depend on the mix of WUs you're getting. And your CPU is also running separation tasks. Averaging over 4 of them yields 31500 s/WU. That's 24*3600/31500 = 2.7 WUs/day per core, or 5.4 WUs per day for your CPU. Each is worth 160 credits, so you'd get an RAC of 877 running only these. The more of them you get, the lower your actual RAC will be. By the way: my GPU crunches through these separation tasks in 50s, the current top cards achieve ~30s. Your CPU could be much more useful elsewhere. And your GTX460s need 8000 - 9000s for these WUs - not exactly efficient either. If I were you I'd opt out of separation tasks and use the GPUs elsewhere (GPU-Grid, Einstein, SETI, POEM etc.). MrS Scanning for our furry friends since Jan 2002 |
©2024 Astroinformatics Group