Hi everybody,
As my lack of former posts would indicate, I'm somewhat new-ish around the computer chess scene. And as someone who knows more about the "chess" aspect rather than "computer" aspect of "computer chess," I've been wondering:
1) Is it unsafe to run chess engines at 100% CPU usage on a standard laptop (Core i3-330m, 2.13 GHz for example, is what I have) for extended periods if it's only cooled by the built-in fan? Is 5 minutes, 30 minutes, an hour, a day , a week etc, enough to result in negative consequences?
2) If it is harmful, what consequences would be expected? Is it just a matter of the thermal compound degrading more quickly over time? Maybe CPU components failing? Something else?
I've seen pics of elaborate cooling mechanisms for desktops here and on the Rybka forum, and I understand those are primarily intended for over-clocking, but how well do ordinary laptop computers tolerate excessive "non-overclocked heat?" I guess I would just be surprised if manufacturers designed average computers with non-stop 100% CPU usage in mind. Any input would be appreciated.
Thanks
-Justin
General computer heating/cooling question
Moderator: Ras
-
- Posts: 29
- Joined: Thu Mar 04, 2010 11:35 pm
-
- Posts: 20943
- Joined: Mon Feb 27, 2006 7:30 pm
- Location: Birmingham, AL
Re: General computer heating/cooling question
Ignore it. Good laptops have built-in protective mechanisms for overheating cases. I've run my core-2 for hours. My older core-2 has run for days, non-stop, with nary a problem...Justin Sherron wrote:Hi everybody,
As my lack of former posts would indicate, I'm somewhat new-ish around the computer chess scene. And as someone who knows more about the "chess" aspect rather than "computer" aspect of "computer chess," I've been wondering:
1) Is it unsafe to run chess engines at 100% CPU usage on a standard laptop (Core i3-330m, 2.13 GHz for example, is what I have) for extended periods if it's only cooled by the built-in fan? Is 5 minutes, 30 minutes, an hour, a day , a week etc, enough to result in negative consequences?
2) If it is harmful, what consequences would be expected? Is it just a matter of the thermal compound degrading more quickly over time? Maybe CPU components failing? Something else?
I've seen pics of elaborate cooling mechanisms for desktops here and on the Rybka forum, and I understand those are primarily intended for over-clocking, but how well do ordinary laptop computers tolerate excessive "non-overclocked heat?" I guess I would just be surprised if manufacturers designed average computers with non-stop 100% CPU usage in mind. Any input would be appreciated.
Thanks
-Justin
Just don't block the air intake or exhaust or it may well throttle back or even shut down if temps go too high.
-
- Posts: 454
- Joined: Sat Apr 04, 2009 6:44 pm
- Location: Bulgaria
Re: General computer heating/cooling question
Totally agreed with Bob! Also, don't forget to pull out the battery, don't run it for extended period with it. I even have an old laptop P mobile on 1.6, that runs 24/7 on 100 % load (and never idles) without any problems, but again
don't forget to pull out the battery!
don't forget to pull out the battery!
-
- Posts: 29
- Joined: Thu Mar 04, 2010 11:35 pm
Re: General computer heating/cooling question
Thanks for the input Bob and Mincho, it's nice to have some re-assurance from more experienced members of the computer chess community!
-
- Posts: 2094
- Joined: Mon Mar 13, 2006 2:31 am
- Location: North Carolina, USA
Re: General computer heating/cooling question
I agree with the others but I would add one simple thing.
My laptop got rather hot and I was concerned enough to purchase one of the cooling pads that you set the laptop on and plug into a USB port. The result was a much cooler laptop. I've run it continuously for weeks with one of these and it stays cool.
My market observations are that some of them may not be so good. Some of them pull the air from under the pad with little clearance while others pull the air from a large rectangular hole in the back. I like the latter ones, but the former should work it you prop the front up an inch or two.
My laptop got rather hot and I was concerned enough to purchase one of the cooling pads that you set the laptop on and plug into a USB port. The result was a much cooler laptop. I've run it continuously for weeks with one of these and it stays cool.
My market observations are that some of them may not be so good. Some of them pull the air from under the pad with little clearance while others pull the air from a large rectangular hole in the back. I like the latter ones, but the former should work it you prop the front up an inch or two.
Re: General computer heating/cooling question
Well, before everyone says "yeah, go ahead!", do you know what temperature your CPU is reaching? It could be a problem if it gets too high. Some poorly made laptops will overheat, as the cooling systems aren't very good.
-
- Posts: 3721
- Joined: Thu Mar 16, 2006 7:10 pm
Re: General computer heating/cooling question
Laptops are generally designed for portability and battery efficiency, not for full cycle 24/7 usage. I really don't see any harm though as good laptops usually are built with a lot of tolerances. Of course if you live in a very warm area then there might be a problem if the air coming in via fan cannot cool the CPU fast enough, but the worst that I think will happen will be a throttle down of the CPU.
For long term analysis 24/7 at full cycles, you are probably better with a small desktop solution. Those things are so cheap now that if you stay away from the top end laptop models you could get a pretty powerful desktop with the money saved from not going with the top of the line laptop. Also with the proper desktop CPU and motherboard setup you can easily overclock it to get a huge performance boost. Best of all is that you can then setup your laptop to run with minimal CPU load and with client remote software you can have the extensive CPU cycles run on the desktop instead and output will come to your laptop. This way you have the best of both worlds.
For long term analysis 24/7 at full cycles, you are probably better with a small desktop solution. Those things are so cheap now that if you stay away from the top end laptop models you could get a pretty powerful desktop with the money saved from not going with the top of the line laptop. Also with the proper desktop CPU and motherboard setup you can easily overclock it to get a huge performance boost. Best of all is that you can then setup your laptop to run with minimal CPU load and with client remote software you can have the extensive CPU cycles run on the desktop instead and output will come to your laptop. This way you have the best of both worlds.
Re: General computer heating/cooling question
Heat is the killer of all things, shortening the lifespan of most electronic components.
I still have and operate the following:
1. A 1989 Macintosh SE/30 accelerated to 32 MHz from 16 MHz and cooled with copper heat sinks and thermoelectric (Peltier) devices I built.
2. A Macintosh G4 Cube at 450 MHz I bought in September of 2000, air-cooled with a high speed fan mounted vertically above the exhaust port. My Dual G4 @ 500 MHz (from 2001) just died a while ago, when its cooling fan shut off and I didn't realize it!
So I have 1 system that is 21 years old, and another going to turn 10 years old, and my 9 year old just died.
All of them were cooled, and that's why I expect they had such long lives.

I still have and operate the following:
1. A 1989 Macintosh SE/30 accelerated to 32 MHz from 16 MHz and cooled with copper heat sinks and thermoelectric (Peltier) devices I built.
2. A Macintosh G4 Cube at 450 MHz I bought in September of 2000, air-cooled with a high speed fan mounted vertically above the exhaust port. My Dual G4 @ 500 MHz (from 2001) just died a while ago, when its cooling fan shut off and I didn't realize it!
So I have 1 system that is 21 years old, and another going to turn 10 years old, and my 9 year old just died.
All of them were cooled, and that's why I expect they had such long lives.
