1) Message boards : GPUs : I keep seeing threads warning people about CPU/GPU damage... (Message 54325) Posted 3 Jun 2014 by TallBearNC Post: As long as your aren't OCing your GPU and CPU as well as overvolting them.. it's pretty hard to fry them... it's rare but possible If you want to take the UTMOST PRECAUTIONS.. then keep reading 1) Proper Case fan control that increase case fan speeds as the ambient air inside the case heats up 2) Proper CPU fan control that automatically increases as heat increases (this is the same for water cooling as water cooling uses a fan(s) to cool the water). Some h2o systems allow for pump speed control that speed up the pump as heat increases (I keep my case air between 55-70F at all times. I have central AC, but I installed a Window AC right above my PC and run the h20 tubing and the put the h20 cooling systems condenser right behind the window units evap core. This blows 45F air through the h2o cooling systems evap core vs room air... which results in 45-50F degree water being returned to my CPU & GPUs. Standard room air of 65-72 is fine for most ppl. I don't recommend doing boinc applications with room air much over 75F 3) Proper GPU fan control. For ANY nvidia GTX cards, EVGA Precision X works best. There's another brand that has a simular product, but I'm not too fond of it. I DO NOT LET THE GPU DRIVER CONTROL THE FANS, CLOCKS, VOLTS, ETC Even if you lack ALL of the above, you can get by with ONLY TThtrottle that will slow your BOINC CPU/GPU projects down so that they don't cross a certain temp threshold - even if it has to run your projects at 1% (100Xs slower) to keep your computer cool enough Lastly... most any, modern computer (even if home built) will simply just "turn off" (cut the power to the mobo)if the CPU nears critical temps. Although this is NOT ideal as you can corrupt your HDD or SDD file system if write back cache is enabled. Most ATI and NV cards have the same safeguards built it.. once you hit a preset temp (usually 1-2C away from doing permanent dmg to your card), either the driver will just shut the GPU down and/or the GPU will shut itself down (computer will just lock or freeze or you will come to a black screen) The last part is NOT ideal.. as there is some risk that the BIOS for the mobo won't shut down in time to stop CPU dmg, and the same goes for the GPU But as long as you use TThrottle, that will keep you safe no matter how many work units, from a single project, that you pile on your GPU. If you pile too many, the GPU can get so busy that the computer will appear to lock up. Just make sure you know the max temp for your cpu and gpu.. and not JUST the max temp!! The max temp is the ULTIMATE MAX... before the ultimate max (it's usually 10-15C below the ultimate max is what I call the life reduction max. Once you go over that number, you start taking months and years off your gpu and cpu depending how much time you spend in this "yellow zoe( If you are lucky, and have another computer on your network that shares at least one admin account in common you are lucky: 1) log into the other computer with the account common between both 2) open an elevated command prompt (Personally, I find CMD.EXE and permanently mark it run as admin) 3) enter: shutdown /r /f /t 0 /m \\computer-name This says reboot the computer named computer-name (/r), don't wait for apps to close, FORCE terminate them (/f), don't give any time to the shutdown/reboot (/t 0), and lastly the name of the computer (/m \\) .. the \\ is very important If you want to just force a shutdown vs reboot, change the /r to a /s, or if you want the shutdown to happen in 5 mintues , change /t 0 to /t 300 This will usually shutdown a system with an overloaded GPU. It will also work with an overloaded CPU, but you may have to issue the command multiple times until you get a shutdown already in progress message

2) Message boards : GPUs : Seperate GPU tasks (Message 40842) Posted 25 Oct 2011 by TallBearNC Post: Yeah you can separate things. Get Boink Tasks one of the columns is called "Use" it will show if something is using just CPU or CPU+GPU (all GPU takes use some CPU), and you can click the bar to sort by type :) BT is much better than boincview imo as you can also hook it into TTthrottle and keep an eye on the temps of all your systems Stats since 10-5-2011:

3) Message boards : GPUs : The ultimate high-end GPU-driven cruncher that excels at gaming and multi-monitor setup (Message 40822) Posted 24 Oct 2011 by TallBearNC Post: Tesla's and Quadros can run in one box, and neither are good for gaming. Geforce cards can't be run with Teslas from what I understand. I see no need for 8 monitors, unless you REALLY need them for something. Windows Vista/7, and soon 8, and current drivers do not need a monitor attached or a "dummy" dongle attached to use a GPU any more with boinc. Linux is a great OS and good for "some" games. I say some because many games aren't made for Linux. IMO if you want a well rounded crunching/gaming, go with a Win 7 box and use linux for crunching only. If you want a system that does BOTH crunching and gaming, I'd get a nice mobo with 8PCIe 2.0 slots... and shove them full of dual GPU cards like the nv590.. you won't get the power of teslas, but 8 590s will do quite a bit of work :) Figure a TOP end nv gpu like a 590 costs about $600 and, it's a x2, so you get about 2Glops or so from it (SP math, about hald that for DP).... take a low end Tesla for $2000 and get 1Tflop to 700Glops from it.. and well.. you can see a Telsa card is a MUCH better value if you are JUST running science, etc apps... plus Teslas work with Quadros as well. now 8 teslas (and 16+K USD).. will scream and give you ~8TFlops of GPU power. no gaming video card can even come close to what a tesla can do. the BEST gaming gpu, right now, does, I think about 1-1.5Glfops tops. and a low end Tesla does 1TFlops single precision and 700GFlops double I have a Telsa card, and it can do things in a FRACTION of a time a 580, 590, etc can push out WUs. Just 1 of those beasts is better than a room full of nvidia and ati GPUs.

4) Message boards : BOINC Manager : feature request - limit CPU temperature (Message 40821) Posted 24 Oct 2011 by TallBearNC Post: No need if you use windows... get TThrottle v5.10. It can interface with boinc and slow things down if your CPU(s) and/or GPU(s) get hot. You can set the max temp for all the GPUs and the max Tjunction temp that a "core" can hit. Example an i7-2600K has a Tcase max of 72.xc, but a Tjunction max of 98. Since Tthrottle measures, tjunction, I set it to about 80c on my i7. I'm OCed up to 5.8Ghz on water cooling. My GPUs are right next to each other and can EASILY go over 90c (especially the one with the fan right at the back of the other card). So I crank their fans to max. and put a temp limit in tthrottle to 85c (the second most NV GPUs touch the 100c mark, their life is cut short). One thing that helps is I put my computers near the 2 windows in my home office and put in 2 15,000 BTU window AC units. I then rigged it so they would force their cold air right into the cases. Once that was done, heat was no longer a concern. I still keep Tthrottle running in case someone in the house turns off the AC units :)

---

Copyright © 2024 University of California.
Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.