Extreem koelen met Fluorinert deel 2

Na deel 1 waar wij twee nieuwsposts aan hadden gewijd, namelijk deze: nieuwsposting 1 en nieuwsposting 2. Nu weer een nieuw vervolg. Hier een hap uit de tekst:

WHAT LN2 WILL DO TO YOUR OVERCLOCK?

So what happened? Did it boot? Yeah it did. Booted easily with the CPU at about -150C(block temp), but you cannot understand the language anymore. The screen suddenly turned alien into us. We cannot understand a damn thing! Checksum error was up. We could go into the BIOS but everything was different. The keyboard types different letters. Pressing letter 'a' gives you a different letter and same goes with all other keys. Tried reflashing the BIOS but floppy drive won't work either whenwe attached one. Besides you cannot type the right command as it gives you different keys. In other words the motherboard was stuffed. Somehow the CMOS was damaged by the extreme temperature. Everything seems to work fine. Video card, RAM, CPU all worked properly after drying and testing them in a normal 'unsubmerged' system. Although the HD was corrupted. The only question though is...was it the extreme board temp that stuffed board? Or was it the CPU's extreme low temperature that caused it? Could the CPU's sudden temp spike corrupted the CMOS? Funny thing though the CPU is 100% working fine up to now. No cracks whatsoever caused by the LN2. So goes our dream of total LN2 immersion.

CONCLUSION

Supercooling an immersed rig can almost double your CPU's rated speed. Considering the 566 we used was a very bad overclocker. Before immersion, it would only do 807mhz (FSB 95) with normal cooling and get to 850mhz with a 52w peltier. After immersion and some subzero temps later, this 566 is stable at 1057mhz. I know we haven't hit the CPU's limit yet but unfortunately we have to stop early because the motherboard got stuffed. But hey 500 mhz overclock is very good achievement for this project. We were actually more interested in how extreme, with regards to temp, the motherboard can take rather than how much we can overclock. By doing this experiment, a lot of questions were answered and actual facts were established based on the results we got.

Het is weer een leuk verhaaltje geworden met een hoop foto's. Het is op zich natuurlijk absoluut belachelijk om met dit soort methodes te gaan koelen, maar het is wel koel. Beetje jammer voor de jongens dat ze niet verder kwamen dan 1057Mhz, want Jasper Janssen kreeg zijn Celeron 566 wel tot 1133Mhz.

Bedankt, Sir Abarth voor het submitten van de link.