> > My colleagues indicated that not all the power consumed by the > > processors and other hardware is converted into heat. The > energy is > > converted into work done by the processor. > > You colleagues are right, all the power consumed by the > processor is not converted into heat. If this was true we > would be calling it a radiator instead of a processor. Sorry, no, absolutely wrong. You are confusing electronics with data processing. The transistors, resistors, etc... In the equipment converts all the power it consumes into heat or output signals which are then converted into heat in turn. The fact that it accomplishes data processing work is unrelated. Real world physics are unaffected by Cyber space physics which do not actually exist. All wattage consumed is expressed as either heat or is distributed out of the room to other places where it is eventually turned into heat. The wattage consumed is a very good indicator of the heat produced by the electronics in the room. Outgoing network signals are very low power. Other sources of heat are transfer from outside the room and human or other biological occupants. In my experience, the heat loads (or cooling) from the outside of a building is much more than that produced by the humans or machines. It doesn't really matter where the room is placed; interior rooms are heated less directly by the rooms around them, but they heat none the less. If the surrounding rooms are cooled, then the room in question will be as well. Insulation in office buildings is (usually) non-existent. If the room has an exterior wall and especially a window, that will be the bulk of the load even for a server room. For example, my little office and the server rack are in a room with two very small windows on the exterior wall which is thick, and apparently solid, concrete. I have 2 routers, a hub and a switch, 5 towers and a laptop with 4 old 21" CRT monitors and 2 LCDs along with 2 laser printers. If I work late, past the point where the thermostat is programmed to regulate to a wider range (74-76F becomes 65-80F), the temperature change is amazingly quick. During the winter, the room temperature will drop 10 degrees in less than 10 minutes. In the summer, even at night, the room temperature will rise 5 degrees in less than 5 minutes. My bosses office, which has large windows covering about half of the walls, with about 3/4's of those facing directly south, and no awnings, will remain hot in the summer even with the HVAC system on full all the time. He now has 3 inlets and one return plus an open door with the common return just outside it. The baffles are adjusted to send more than half the systems air through his office. The windows have had a heat blocking coating applied. The effect was minor. The air between his blinds and the window glass (before the coating) would literally burn your hand during the peak summer months. If the room already has A/C, I would increase the capacity by the wattage being put in the room, but if it is not already being air conditioned, thinking that the 'trons and humans are the only thing to worry about would be a major mistake. --- James. -- http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist