|
|
After noting the limits of its .25-micron fabrication process, Intel shifted to a new .18-micron core for the Pentium III Coppermine. Intel also made its Level 2 cache internal to the core die, thus improving performance compared to the Katmai's external cache architecture. The "Coppermine" name created confusion, however, and many users upgraded to this architecture believing that its internal circuitry featured a copper interconnect design. In truth, the Coppermine still used the aluminum standard. Even so, upon its release the Coppermine became the best overclocking option available.
The desktop Coppermine is available in two distinct forms. Overclocking techniques remained consistent for the Slot 1 interface (a long, skinny card-edge socket) for Coppermine. The newer model, featuring a 370-pin FCPGA socket (a flat, square socket), extended overclocking potential by allowing the use of vertical heatsink coolers for greater thermal dissipation and efficiency. Combined with a new .18-micron core and lower operating voltage, the Coppermine allows overclocks in the hundreds of MHz, some even exceeding default operating frequencies by more than 50%!
Those possessing type eb processors will have less headroom for overclocking. The 133-MHz front-side bus of this series, combined with motherboard chipset limitations, prevents achieving higher speeds. A higher multiplier value is desirable in this series, especially for those interested in overclocking to 1+ GHz for maximum performance.
The most promising models are in the e series. These chips feature a base 100-MHz front-side bus design. The lower multiplier ranges (5.5× to 7.5×) offer overclocking returns of 200 to 300 MHz, requiring only a slight bump in core voltage levels and the addition of an aftermarket cooler. Even the retail cooler provided by Intel can often provide satisfactory cooling for overclocking by 100 to 150 MHz.
|
|