After testing, re-testing, and then testing some more, the final overclocking results for our particular Celeron(tm) 433 example were atrociously bad. In fact, they're the worst we've ever recorded for a Celeron(tm) since the line was introduced.

Here's the breakdown:

6.5 x 66MHz	=	433MHz	=	100%, Stable
6.5 x 68MHz	=	442MHz	=	100%, Stable
6.5 x 75MHz	=	488MHz	=	100%, Stable
6.5 x 83MHz	=	541MHz	=	0%, Fail
6.5 x 100MHz	=	650MHz	=	0%, Fail

Ouch.

This is the first time in recent memory that Sharky Extreme failed to obtain a Celeron(tm) overclock at the "gravy" 83MHz FSB speed. Honestly speaking, it wasn't even close. We experienced lockups right after powerup, never once achieving a full Win98 boot. That's disappointing, especially to the same group of people who were able to take two Celeron(tm) 366 CPUs to 550MHz (5.5 x 100MHz) an almost impossible task according to the industry people we've talked to at the retail level.

For whatever reason, our Celeron(tm) 433 was miserable at anything higher than 488MHz, which it handled without incident.

Perhaps the .25 micron Celeron(tm) core really is limited at a certain finite MHz level which some in the industry have previously speculated, and only a die-shrink will allow speeds above 550MHz. Our experiences with a number of Celeron(tm) 400 CPUs would seem to indicate that this is very possible as we've never come close to hitting the mythical 6.0 x 100MHz (600MHz) overclock with any of them.

In any case our Celeron(tm) 433 is just a single CPU, and an early one at that. Hopefully we just got a bad apple here, but we can't know for sure until more OEM distributors obtain large quantities of the part next week.