At 10:45 PM 11/6/98 +0000, Robert Paris wrote: >It occurs to me that when a chip is running, it generates heat, thus heating >its surroundings. When its surroundings reach a certain temperature, it >doesn't get any hotter. (the energy dissipated is less than or equal to >the amount required to sustain the temperature) My question is.... Is >this temperature really too hot for the chip? (I assume a computer that >runs almost constantly - the heating/cooling expanding/contracting in >getting to and from such a temperature would very soon destroy a chip, >I'd imagine) Generally, your primary concern with overheated chips is not chip damage, but chip unreliability. The properties of semiconductors change with temperature, and this is taken into account during the design of semiconductor products. The logic elements on a chip are designed to operate properly within a certain temperature range. If the chip gets too hot, the properties of the semiconductors change sufficiently such that the logic elements of the chip start to give incorrect results. You start to get memory corruption, program crashes, and other nastiness. There are things that chip designers can do to make a chip more heat-tolerant, but there are all kinds of tradeoffs inherent there. You might be able to run a chip past its heat tolerances for quite a long period of time, with reasonably good results. Overclockers do this all the time with Pentiums. If the chip is cool enough to be at least reasonably operational, who cares if the chip is slowly damaged over the course of a year or two? The chip would be pretty out of date by then anyway. Anyway, to really directly answer your question: Yes, heat is a major concern for many modern computers. You didn't think that those big honkin' heatsinks that they put on Pentiums and Pentium IIs were decorative, did you? They're there to try to conduct heat away from the CPU as fast as possible, and lacking that heatsink, *plus* sufficient airflow to carry heat away from the heatsink itself, the computer will crash after just a couple minutes. If you don't have good airflow, the heat will pretty much just stay in the case. -- Subcription/unsubscription/info requests: send e-mail with subject of "subscribe", "unsubscribe", or "info" toWear-Hard Mailing List Archive (searchable): http://wearables.ml.org
From Wear-Hard Mailing list Archive (WH)
Maintained by R. Paul McCarty
Archive created with babymail