Hi. I don't if this is a news or not but I still want to post here. There are two different cores of Celeron C-128.
0x683: 533A-700 v:1.5~1.65 Back of the chip: only two resistors with no other contact plate. The size of the die is bigger.
0x686: 533A-700 v:1.7 Back of the chip: Just look like a P3 but some open connector( fewer resister than P3). The size of the die is smaller. CHECK the datasheet on Intel's web.
Overclocking: 0x686 is better for overclock. I got two 0x686 600s and they will work beautifully @ 945 with default 1.7V. And one is now working @ 990 with 1.9V. I also got a retail boxed 0x683 600 which refused to work at 900 with even 1.8V (default 1.5V) . FYI, from Intel, the MAX V for FC-PGA is 2.1V.
Also from Intel's datasheet. 1.13G and 1.0G P3s used only 0x686 and with 1.8V and 1.7V. That means 0x686 is the newest core design and was designed to reach 900 to 1G+ range with only 1.7V. So, my conclusion is, go for a 0x686 chip for overclocking.
0x683: 533A-700 v:1.5~1.65 Back of the chip: only two resistors with no other contact plate. The size of the die is bigger.
0x686: 533A-700 v:1.7 Back of the chip: Just look like a P3 but some open connector( fewer resister than P3). The size of the die is smaller. CHECK the datasheet on Intel's web.
Overclocking: 0x686 is better for overclock. I got two 0x686 600s and they will work beautifully @ 945 with default 1.7V. And one is now working @ 990 with 1.9V. I also got a retail boxed 0x683 600 which refused to work at 900 with even 1.8V (default 1.5V) . FYI, from Intel, the MAX V for FC-PGA is 2.1V.
Also from Intel's datasheet. 1.13G and 1.0G P3s used only 0x686 and with 1.8V and 1.7V. That means 0x686 is the newest core design and was designed to reach 900 to 1G+ range with only 1.7V. So, my conclusion is, go for a 0x686 chip for overclocking.