Intel Core micro architecture

Performance

In a nutshell, the performance of a processor can be described as the product of the clock frequency and the so called IPC (Instructions per Clock cycle), which is the average amount of instructions that a processor can finish each clock cycle. This IPC depends on the micro architecture: some architectures are more efficient than others and can finish instructions faster. While developing the NetBurst micro architecture Intel mainly focused on increasing the clock frequency and this way increasing the performance. In a way they succeeded: where the first Pentium 4 was launched on 1.5 GHz, the newest models are working on frequencies up to 3.8 GHz. But during the years Intel found out that the implementation of the Pentium 4 used far too much power on high clock frequencies, with extreme operation temperatures as a result. Instead of reaching the 10 GHz barrier - which was promised off the record during the Pentium 4 launch - the micro architecture didn't even seem to hit the 4 GHz mark. Technologies like HyperThreading and dual-core made sure that the performance of the Pentium 4 was increased quite a bit, but it was definitely time for something brand new. The new micro architecture features a much higher IPC and therefore better performance, while at the same time keeping the power usage on an extremely low level.

Intel’s CTO Pat Gelsinger shows- Intel’s newest dual-core processors during IDF. Conroe (Core 2 Duo for desktops), Merom (Core 2 Duo for laptops) and Woodcrest (Xeon 5100 series) are based on the new micro architecture.