Message boards : Number crunching : PROTEIN-PROTEIN DOCKING MODELS
Author | Message |
---|---|
PUDDIN TAME Send message Joined: 3 Oct 06 Posts: 13 Credit: 53,998 RAC: 0 |
What is the skinny on the new Protein-Protein docking W.U. I just ran one and it seemed to use a lot more of my computers resources than the regular single protein W.U. Does it take more RAM to run the docking models? I have an older computer. My machine only has 256MB of RAM, but, it runs the single protein models fine. When running the docking model it seemed to constantly be going to virtual memory. The disc drive light was on almost the entire time the W.U. was running. I would hate to think that I will have to drop out of rosetta because my computer is no longer adequate. PUDDIN TAME |
Feet1st Send message Joined: 30 Dec 05 Posts: 1755 Credit: 4,690,520 RAC: 0 |
I just checked, I've got a dual core, and so have two WUs running. One is a DOC and the other is a huge single protein (FRA_2rio_CASP7...) turns out the DOC task running on my machine is using about 114MB, whereas the FRA_ task is using about 225MB. The memory usage will vary by task, and within a task depending on the course your search takes you, it may vary a little between different models. The DOC work units are the latest science improvement, just released last week. So, these run the newest code. As they gain more experience with it, they may be able to find some ways to reduce the memory requirement. The project's original requirement was 512MB, but they later were able to make improvements and dramatically reduce memory requirements and reduced the minimum to 256MB. I guess what I'm saying is that it's not just the DOC WUs. And I agree, if they are going to take this much memory to run, then the project's minimum requirements may need to be revisited. I see you also have a laptop with more memory. Laptops typically have to throttle up the fan to keep cool enough when running Rosetta. Help is on the way there, BOINC is adding a feature which will let you decide more much of your idle CPU goes to BOINC projects (rather than 100% of the idle time). This will allow you to set it to run at say 70% or whatever you find keeps the fan at a reasonable noise level for how you use your machine. Add this signature to your EMail: Running Microsoft's "System Idle Process" will never help cure cancer, AIDS nor Alzheimer's. But running Rosetta@home just might! https://boinc.bakerlab.org/rosetta/ |
R.L. Casey Send message Joined: 7 Jun 06 Posts: 91 Credit: 2,728,885 RAC: 0 |
What is the skinny on the new Protein-Protein docking W.U. I just ran one and it seemed to use a lot more of my computers resources than the regular single protein W.U. Does it take more RAM to run the docking models? I have an older computer. My machine only has 256MB of RAM, but, it runs the single protein models fine. When running the docking model it seemed to constantly be going to virtual memory. The disc drive light was on almost the entire time the W.U. was running. I would hate to think that I will have to drop out of rosetta because my computer Jim, hang in there! I have an old Compaq laptop with an AMD K6 processor at ~400 MHz with only 160 MB of RAM, less than 900 MB of free disk space and ~900 MB swap file that has run its only (so far) DOC WU successfully. Perhaps you could add some swap space to increase virtual memory (VM) to see if the performance is improved. The Windows Task Manager can be used to see how much VM is in use. Good luck, and happy crunching! |
David Baker Volunteer moderator Project administrator Project developer Project scientist Send message Joined: 17 Sep 05 Posts: 705 Credit: 559,847 RAC: 0 |
we will increase the minimum memory requirement for larger jobs to avoid causing problems with people with smaller memory machines |
Astro Send message Joined: 2 Oct 05 Posts: 987 Credit: 500,253 RAC: 0 |
My Celeron 500, 256Mram, Win98 just completed one model ofWU 37086996 in 38,965.00 seconds(10.82 hours) and did so successfully with Valid results. If was an FRA_2rio_CASP7_hom001_1_2rio_1_1a06__IGNORE_THE_REST_197_1281_17 wu. It's also done 1 model on DOC_2VIR_pose_u_pert_with_bbmin_1282_544_0 Basically no wu have errored on this sytem. Yep, getting proud of this old chugger, Where the AMD Athlon(tm) 64 Processor 3200+ failed, mine succeeded. :) It might be helped along by the fact it sits under a stand all by itself with no keyboard, no mouse, no monitor, and otherwise not bothered by human operators. LOL |
Message boards :
Number crunching :
PROTEIN-PROTEIN DOCKING MODELS
©2024 University of Washington
https://www.bakerlab.org