Over the next few months I played with getting the Celery to run as cool as possible in my previously altered Inwin A500 case.  This involved modifying a Vantech PII CPU Cooler heat sink into the popular "sandwich" configuration.  I made the observation then that I didn't consider the Celeron a very warm chip and I still don't - at anywhere near the stock voltage of 2 volts plus or minus a tenth or two.  With the Celery now running at about 4°F over room temp while at idle with the case on, in 80° temperatures, I again tried to reach the 500 MHz barrier.  At 2.3 volts, it would enter windows and would promptly lock up as soon as any strain was put on it.  I had my best results using Win98 as the operating system which sometimes allowed me to play with it for fifteen minutes or more before locking up.  Win95 and NT4 didn't seem to get along with the processor quite so well.  With the exception of the motherboard, Hitachi PC100 RAM, and the G200 AGP video card, all of the other components in the case had successfully run with my K6 at an 41 MHz PCI bus clock, so I doubted that the small increase of 4+ MHz to the PCI bus had any impact on the Celeron's instability at 112 MHz.

I shelved my hopes of making 504 MHz until I finally decided to go for broke and crank up the voltage until the chip either fried or attained 504.  With the introduction of a .18 micron offering from Intel seemingly coming soon and a spare CPU to fall back on if I only succeeded in producing that culinary delight, toasted celery, I set out to see if pushing the voltage past the limits set by the BIOS would allow me to run at 504.

Now before I proceed, this is the point in the article that one usually inserts such pearls of wisdom as: "Proceed at your own risk." or "Hey, it worked for me, but you might not be so lucky." or some other liability limiting phrase.  I will refrain from doing so on this occasion, 'cause if you bump the voltage and the chip evaporates in a cloud of acrid smelling smoke, you got what you half-expected would happen anyway.  'nuff said.

Intel's Specs
The following are Intel's published limits for the Celeron processors.

Maximum heat - 185°F (85°C) measured at the CPU's case (slug).
Maximum voltage - 3.0 volts (2.0 + 1.0 volt)

Yeah, I know, it seems like a misprint, but those are the specs from the PDF document, section 2.9, page 19, available from: ftp://download.intel.com/design/celeron/datashts/24365806.pdf

Now, bear in mind that these are maximum figures, and as Intel puts it:

"Functional operation at the absolute maximum or minimum is not implied or guaranteed.  The processor should not receive a clock while subjected to these conditions."

So, you can apply 3 volts to it while cooking it at 185°F as long as you don't try to run it.  Gee, that's helpful information... for someone....