The Intel Celeron 533A & GlobalWin FKP-32

Elation cannot begin to describe the feeling that this particular processor chip gave me a few seconds after installing it into the motherboard and booting up for the first time. You see, it wasn't supposed to work. But I'm here to tell, no SHOUT, that the Celeron 2 series of processors work naked (sans any adapters) in the venerable Abit BM6 and probably the ZM6 boards. And what a pleasant surprise! Without much effort or expense, the diminutive (pricewise) 533A provided by Azzo, achieved a too good to be believed 300+ MHz overclock and has become the best overclocking value to date.

Factory Specifications	The 533A
Pentium III Coppermine Core MMX and SSE support 32KB L1 & 128KB L2 Cache 1.50 Core Voltage 0.18u die, 90C max temp. Socket-370 FC-PGA 17.1 Watt Dissipation Azzo price: $139 (OEM)	Intel
Factory Specifications	The GW FKP-32
Rated DC 12V @ 0.18A Dual Ball Bearings 4200 RPM fan speed 26 CFM output 36 dBA noise level Therm. Cond. 0.0015K Azzo price: $22	GlobalWin

Pins & Needles

I was sweating bullets when the 533A arrived via FedEx a few days ago. Although reading various BBS and newsgroup posts provides some measure of believability, 100% assurance was not bouncing around in my head. Risk involves some heartache and loss, but this time, neither reared their ugly heads.

After benchmarking the 1999 Combo Of The Year one last time, I pulled the 366@550 out and blew off the dust that the more than capable GlobalWin CPM-32 spread liberally everywhere. Just prior to removing the chip, I clocked the FSB down to 66Mhz and 1.9 volts. Rumors have been that if the BIOS of your MB doesn't support the newer 0.18u Coppermine chips (P3 or C2), disappointment and stomach turning can happen when booting to a significantly higher voltage than factory specs allow. This C2 has a default voltage of 1.50, while some speed equivalent FC-PGA P3's are labeled as 1.60 volts. If the BM6's PQ BIOS didn't "get" the C2's core voltage right, I might have heard that dreaded "popping" sound followed by my own gut wrenching hurling sound.

I downloaded the spec sheet from Intel that had the C2 533-600 particulars and glanced at all the numbers I could. The max (before damage) static voltage for the 533-600 C2 series is a high 2.1 volts. The maximum recommended operating voltage was harder to determine as a listing with that type of wording was not found. What I did spy was a Vtt and Vcc spec which ranged from 1.545 to 1.635 volts. I'd venture to guess that 1.60 volts would be the maximum I'd run at long term and Abit's PQ BIOS only allowed a maximum of 1.70 volts. Please don't subject the little marvel to 2.0 volts, boot up at 1+ GHz, and write to tell me as you suddenly smell something akin to burning silicon.

When the white BIOS screen faded in for the first time, I laughed and cried. The laugh was the "HA" kind that only happens when you've done something that many say can't be done. The crying was the good kind too, because the PQ BIOS has hidden in it nearly full support for the C2 chips. I saw 1.50 volts and the default highlighted after entering the BIOS that claimed the 533A was a 533E. Since the ZM6 is a video castrated version of the BM6, I could easily believe that similar results would be obtained although I have no ZM6 available for evaluation.

By this time, I did the "happy dance" and ran around the room hollering "YES, YES" for about 2 minutes. That was just for a lousy DOS boot at 533 MHz. Now, I love overclocking and have even squeezed a P120 up to 133, but I wasn't fully prepared for the overclocking nirvana that the C2 so graciously provided.