Megahertz myth

From Wikipedia, the free encyclopedia

The megahertz myth, or less commonly the gigahertz myth, refers to the error of using clock rate to compare the performance of different microprocessors. While clock rates are a valid way of comparing the performance of different speeds of the same model and type of processor, other factors such as pipelines and instruction sets can greatly affect the performance when considering different processors. For example, one processor may take one clock cycle to add two numbers and another clock cycle to multiply by a third number, whereas another processor may do the same calculation in one clock cycle. Comparisons between different types of processors are difficult because performance varies depending on the type of task.

Contents

[edit] Basis of the myth

The myth arose because the clock speed was commonly taken as a simple measure of processor performance, and was promoted in advertising and by enthusiasts without taking into account other factors. The term was originally coined (or at least came into widespread use) in the context of comparing PowerPC-based Apple Macintosh computers with Intel-based PCs. Marketing based on the myth led to the clock speed being given higher priority than actual performance, and led to AMD introducing model numbers giving a notional clock speed based on comparative performance to overcome a perceived deficiency in their actual clock speed.

For example, a processor at twice the clock speed may only accomplish half the number of instructions per cycle, thereby offering no more performance than the slower-clocked alternative.

[edit] Historical development

[edit] Background

The x86 CISC based CPU architecture which Intel introduced in 1978 was used as the standard for the DOS based IBM PC, and developments of it still continue to dominate the Microsoft Windows market. An advanced IBM RISC based architecture was used for the PowerPC CPU which was released in 1992. In 1994 Apple Computer introduced Mac computers using these PPC CPUs, but IBM's intention to produce its own desktop computers using these chips was thwarted by delays in Windows NT and a falling out with Microsoft. Initially this architecture met hopes for performance, and different ranges of PPC CPUs were developed, often delivering different performances at the same clock speed. Similarly, at this time the Intel 80486 was selling alongside the Pentium which delivered almost twice the performance of the 80486 at the same clock speed.

[edit] Rise of the myth

Computer advertising emphasised processor megahertz, and by late 1997 rapidly increasing clock speeds enabled the Pentium II to surpass the PowerPC in performance. Apple then introduced Macs using the PowerPC G3 which outperformed Pentium IIs while consuming far less power. They proclaimed the advantage in performance with commercials showing a Pentium II on a snail, and a "Toasted Bunny"-suited character parodying Intel's commercials. Intel continued to promote their higher clock speed, and the Mac press frequently used the "megahertz myth" term to emphasise claims that Macs had the advantage in certain real world uses, particularly in laptops.

In November 2000 Intel's heavily advertised advances in clock speed reached an extreme with the release of the Pentium 4 which sacrificed per-cycle performance and used a deep instruction pipeline to gain very high clock speeds, ignoring problems that this introduced of heat production and power consumption.

Showdowns between PowerPC and Pentium had become a staple of Apple presentations. At the New York Macworld Expo Keynote on July 18, 2001, Steve Jobs described an 867 MHz G4 as completing a task in 45 seconds while a 1.7 GHz Pentium 4 took 82 seconds for the same task, saying that "the name that we've given it is the megahertz myth". He then introduced senior hardware VP Jon Rubinstein who gave a tutorial describing how shorter pipelines gave better performance at half the clock speed.

AMD was now producing innovative x86 designs which competed with Intel on performance rather than price. In January 2002 the Pentium 4 gained a lead in sales and AMD marketing responded by giving their processors numeric suffixes approximating the clock rate that an AMD Thunderbird (and by inference a Pentium processor) would need to give matching performance, openly undermining the "megahertz myth".

[edit] The myth becomes counterproductive

The power-hungry hot-running Pentium 4 was unsuitable for laptops, and in March 2003 Intel overcame these difficulties with the successful Pentium M which proved capable of matching the Pentium 4 on performance at much lower clock rates. In 2004 problems of overheating led Intel to abandon further development of its Pentium 4 experiment in high clock speeds. Instead, Intel focused its future plans on the Pentium M architecture which by then incorporated RISC techniques to the point that the distinction between RISC and CISC was blurred.[1]

The IBM G5 also proved unsuitable for laptops, and in 2005 Apple announced that over the following year Macintosh computers would switch to Intel CPUs developed from the Pentium M.

[edit] The future of the myth

Ironically, Intel is now having to dig itself out of a marketing hole it created for itself when it released the Pentium 4. Their new generation of chips, the Intel Core 2, has clock speeds of around 2–3 GHz. While the Core line is a breakthrough in terms of performance-per-watt, its low clock speed when compared to late generation Pentium 4s (rated at upwards of 3.8 GHz) is likely to cause some marketing confusion. Intel is now in the position of trying to sell consumers processors with lower gigahertz ratings, having spent the better part of the last five years telling consumers that slower clock speed denotes inferiority.

This can also cause problems for third party manufacturers. For example, Panasonic lists a Pentium 4-based machine running at 3 GHz as the minimum system requirement for their soon to be released Blu-ray Disc drives. A 1.83 GHz Core 2 Duo is significantly faster than the 3 GHz Pentium 4, but to some consumers reading specifications on the side of a box this statement can be completely confusing.

[edit] References

  1. ^ Crystal Chen; Greg Novick and Kirk Shimano (2000). Recent Developments. RISC architecture. Stanford.edu. Retrieved on 2007-11-27.

[edit] External links

Languages