Computational Friction

Tonight, I’ve been transferring some videocassettes to DVD. It’s a time consuming process – more alchemy than anything else. This is well out of the realm of standard procedure. I’ve learned it all through trial and error and more error.

I capture the video as avi files and then convert them to mpeg. The specifics of those two aren’t important, except to say they’re different. Same with the resolution. My video capture card wants to give me 640×480 while DVD’s play video that’s 720 pixels wide. Again, all that’s important are they’re different.

In order to go from one format to another, one resolution to another, my computer has to do massive amounts of number crunching. I just popped open my Task Manager a few seconds ago. The computer’s CPU, its brain, is running at 100%.

There’s nothing wrong with that. I want my computer to work as fast as it can.

However, something else has changed while this crunching has been underway. The temperature of the CPU has gone from 143&#176 to 154&#176. That’s not critical, but it certainly gives me a little pause. I believe the CPU can hit about 176&#176 before it starts smoking and breaks down (Actually, the number is probably higher than that. 176&#176 is just the specification).

So, what’s going on? I don’t really know, and a search of Google hasn’t been very fruitful. I think this might frictional heat caused by the additional electrons necessary to really number crunch. But, I could be very wrong.

No matter what, it is very curious that computers heat up when they work hard, as humans do. It is something I would have never expected.

2 thoughts on “Computational Friction”

  1. Friction — an interesting and original thought. Electons do suffer collisions (well, I’m not sure there’s really any suffering going on). On the great scale of how many electons there are nobody can say what any one electron is doing at any time, but the average difficulty they are all having moving in a given direction is known as mobility (try searching on electron mobility). This affects speed, and therefore current, but I’ve never thought of it as friction. Friction as I remember is more of a macro-scale problem, perhaps with the roughness of surfaces in contact?

    Anyways, your CPU simply has more nodes changing voltage per second when its actually doing something than when its idle. It takes power to switch nodes and dissipating power heats it. A google search on thermal impedance might bring up something useful.

  2. Thanks – I really didn’t know. But, it is interesting to me to see that computers are doing more work when they’re doing more work… if you know what I mean.

    All the best,

    Geoff Fox

Leave a Reply

Your email address will not be published. Required fields are marked *