One of Intel’s major marketing pushes for the Pentium 4 processor has been its ability to scale to incredibly high clock speeds. Architecturally we’ve explained countless times how Intel is able to accomplish this through the lengthening of the CPU’s primary pipeline and the use of low latency caches. To the end user however, all that is understood in the end is a higher clock speed.
A side effect of Intel’s intense marketing focus on clock speed (and AMD’s marketing focus on IPC) is that many users seem to devalue the clock speed bumps that the Pentium 4 has seen. This may be partially due to the success of AMD’s new modeling system or simply because a good portion of the market realizes that AMD CPUs are offering equal or greater performance at noticeably lower clock speeds.
It’s difficult to explain over and over again the two different approaches that AMD and Intel take to microprocessor design which result in the difference in the way the Athlon XPs and Pentium 4s are marketed today.
In the end what truly matters is the overall performance of the CPUs and the price they’re going for. If AMD accomplishes great performance through high IPC rates and lower costs through small die sizes, or if Intel does the same through high clock speeds and larger Silicon wafers, so long as the end result is a high-performance, low-cost CPU the end user should be satisfied.
Just last month we took a look at the AMD Athlon XP 2100+ which typifies AMD’s approach to microprocessor design; the CPU was clocked almost 25% lower than the fastest Pentium 4 CPU but kept up quite well through the use of shorter pipelines and raw execution power. Today we’ll be looking at the latest in Intel’s aggressive roadmap for the Pentium 4. Now clocked at 2.4GHz the Pentium 4 is far from the CPU we scorned back in November of 2000.