Megahertz myth

Lore & Background

The myth began around 1984 when comparing the Apple II with the IBM PC. The IBM computer was argued to be five times faster because its Intel 8088 processor had a clock speed roughly 4.7 times that of the MOS Technology 6502 in the Apple II. However, the LDA # instruction on the 6502 required two clock cycles (2 μs at 1 MHz), while the same instruction on the 8088 needed at least four cycles (0.84 μs at 4.77 MHz), making the IBM PC only a little more than twice as fast for that instruction.

The term "megahertz myth" came into use when comparing PowerPC-based Apple Macintosh computers with Intel-based PCs. At the New York City Macworld Expo on July 18, 2001, Steve Jobs described an 867 MHz PowerPC G4 completing a task in 45 seconds while a 1.7 GHz Pentium 4 took 82 seconds, stating that "the name that we've given it is the megahertz myth." Senior hardware VP Jon Rubinstein then gave a tutorial explaining how shorter pipelines gave better performance at half the clock rate. The online cartoon The Joy of Tech subsequently presented a series of cartoons inspired by Rubinstein's tutorial.

From approximately 1995 to 2005, Intel advertised its Pentium processors primarily on clock speed alone. This continued until about 2005, when the Pentium Extreme Edition reached thermal dissipation limits at nearly 4 GHz. The introduction of the Core 2 desktop processor in 2006 allowed nearly a 50% decrease in clock speed while retaining the same performance, having its beginnings in the Pentium M mobile processor where energy efficiency was more important.

Reader's Guide

The megahertz myth's significance lies in how it shaped marketing, consumer perception, and processor design for over a decade. By promoting clock rate as the primary measure of performance, Intel's advertising from approximately 1995 to 2005 led competitors like AMD to introduce model numbers giving a notional clock rate based on comparative performance to overcome a perceived deficiency in their actual clock rate. The myth obscured the fact that factors such as pipeline depth, cache hierarchy, and instruction sets can cause different processors to deliver vastly different performance at the same clock speed.

The myth's legacy is evident in the shift toward more balanced processor designs. The thermal limits reached by the Pentium Extreme Edition at nearly 4 GHz around 2005 forced a rethinking of design priorities. The Core 2 processor in 2006, derived from the energy-efficient Pentium M, demonstrated that reducing clock speed by nearly half could maintain performance while improving efficiency. This marked a departure from the clock-speed race and a return to architectural improvements. In subsequent years, advances in manufacturing and power management allowed clock speeds to rise again—some Intel processors exceeded 4 GHz by 2018, and AMD's FX-9590 reached 5.0 GHz in 2013—but with much higher efficiency than the earlier NetBurst designs. The myth remains a cautionary example that benchmarks, not clock rates, provide a more thorough measure of computer performance.

Did You Know?

• The term 'megahertz myth' was coined by Steve Jobs at the New York City Macworld Expo on July 18, 2001.
• The IBM z10 achieved 4.4 GHz in 2008, and the IBM z196 reached 5.2 GHz in 2010, before any x86 processor broke those barriers.
• The Core 2 desktop processor introduced in 2006 allowed nearly a 50% decrease in clock speed while retaining the same performance as previous Intel desktop processors.

The Core Misconception: Why a Bigger Number Does Not Mean a Faster Machine

Clock speed, expressed in megahertz or gigahertz, tells you how quickly a processor's internal clock ticks, but it says very little about how much useful work gets done per tick. The megahertz myth is the widespread belief that a higher number on the frequency dial automatically means a faster machine. In reality, factors like the number of execution units, pipeline depth, cache hierarchy, branch prediction quality, and the instruction set architecture all shape real-world throughput. A classic illustration dates to 1984, when the IBM PC's Intel 8088 ran at roughly 4.7 times the clock rate of the Apple II's MOS Technology 6502. Enthusiasts declared the PC five times faster, yet a simple load-immediate instruction that the 6502 completes in two cycles (two microseconds at one megahertz) requires at least four cycles on the 8088, translating to roughly 0.84 microseconds. The actual speedup is barely two-to-one. What ultimately matters is how long a specific task takes to finish, not how finely the machine slices its operations.

The PowerPC Showdown and Naming the Myth

The term megahertz myth entered popular vocabulary during the early-2000s rivalry between Apple's PowerPC-based Macintoshes and Intel Pentium machines. By 1994, Apple had adopted the RISC-based PowerPC architecture, and the 80486 was still selling alongside the Pentium, which delivered nearly double the performance at identical clock rates. Yet marketing campaigns and enthusiast communities kept elevating raw frequency as the headline number. AMD, feeling the sting of a lower actual clock rate, even began assigning model numbers that implied a notional speed based on comparative performance. The moment the myth got its name came on July 18, 2001, at the New York City Macworld Expo. Steve Jobs presented a Stevenote showing an 867 MHz PowerPC G4 finishing a task in 45 seconds while a 1.7 GHz Pentium 4 needed 82 seconds, and he explicitly labeled the confusion the megahertz myth. Senior hardware VP Jon Rubinstein then walked the audience through how shorter pipelines could outperform longer ones at half the clock rate. The online comic The Joy of Tech later turned Rubinstein's tutorial into a series of explanatory cartoons.

Hitting the Thermal Wall

From roughly 1995 through 2005, Intel's mainstream Pentium advertising leaned heavily on clock speed as the primary differentiator against AMD's competing chips, and press coverage routinely predicted that processors would eventually reach ten or twenty gigahertz within a couple of decades. That trajectory crashed into a physical barrier around 2005. The Pentium Extreme Edition was pushing toward nearly four gigahertz when thermal dissipation limits made further frequency increases impractical; going faster would have demanded exotic cooling solutions like microfluidic channels embedded directly in the silicon. The industry's response was the Core 2 desktop processor, launched in 2006, which achieved a landmark shift: it cut the operating clock by nearly fifty percent while preserving the same level of performance. The design philosophy traced back to the Pentium M mobile chip, where energy efficiency had always been prioritized over raw throughput, and it introduced power-saving features that the Pentium 4 and Pentium D lines simply did not offer. The clock-speed arms race, as a marketing strategy, was effectively over.

Who Actually Broke the Speed Barriers

In the years following the retirement of Intel's NetBurst microarchitecture and its three-plus-gigahertz Pentium 4 and Pentium D chips, consumer clock speeds initially fell by about a gigahertz before gradually climbing again. Advances in manufacturing processes and the ability to set clock speeds independently per core eventually allowed Intel to reclaim and surpass those old frequencies with far greater efficiency. By 2018, chips like the Core i7-7700K and i3-7350K were shipping with a 4.20 GHz base clock. On the AMD side, the initial Bulldozer-based FX processors crossed the four-gigahertz threshold for x86 in 2011, and the FX-9590 released in June 2013 reached up to 5.0 GHz, though the familiar problems of power draw and heat returned. Notably, neither Intel nor AMD was first in the industry to cross these milestones. IBM's mainframe line got there ahead of both: the z10 hit 4.4 GHz in 2008, the z196 reached 5.2 GHz in 2010, and the z12 pushed to 5.5 GHz in the autumn of 2012.