What ever happened to this ‘next big thing’?
Random survey: “Hello, Mr. Office User. Could you give me your opinion of the new 64-bit computers?”
“Uh, the what?”
“Hello, Mr. Executive. Are you ready to buy 64-bit machines for your company?”
“We’ll, ah, take it under consideration. What is that, some new kind of computer?”
“Hello, Ms. Geek, got your first 64-bit PC yet?”
“You gotta be kidding! The people around here don’t even know they’ve got 32-bit computers.”
So it goes, computer fans. The mystery of the 64-bit personal computer. Is it a myth or legend? Is it the Next Big Thing coming to your desk (or lap)?
Let us waltz for a moment down memory (non-RAM) lane and recall those days of yesteryear when people yearned for the latest and greatest personal computers. There were idealists in those days, people who said you could never give enough computer power to the people. Let the bits roll–2 bits, 4 bits, 8 bits, a dollar…16 bits, 32 bits, and then the ultimate: 64 bits. Just like the big boys: mainframe computers. They have been 64-bit for a long time, as have many high-end workstations.
“What,” you might ask, “is 64-bit?” Here’s a bit of a primer.
It’s likely you’ve heard of bits and bytes. They’re the fundamental units used to measure the computational capacity of a computer processor. They’re also used to measure computer storage such as RAM and disk space, which confuses some people. Without going too far into Computing 101, bits refer to the original zeros and ones of binary computation, which in turn refer back to the individual electronic “switches” in transistors that are either on (1) or off (0). Get a row of switches, say eight of them, and set them up like this: “on, off, on, on, off, off, off, on.” That’s eight bits, representing 10110001 in binary code. The eight bits make up one “word,” or a byte. Most computer processors do computations with “words” of this kind-adding, subtracting, manipulating text, and so forth.
It doesn’t take much intuition to see that your processor (CPU) would compute faster if it could handle 16 or 32 bits at a time. In fact, the more bits you have, the more you can break them up into separate words and do all kinds of fancy tricks, especially processing different things at the same time (parallel processing). And when you get to 64 bits, well, now you’re cookin’.
Personal-computer evangelists have for some time promulgated the dogma that you can never have enough computing power. We users of personal computers were first promised 64-bithood almost 10 years ago. Enterprise-level business has been running on 64-bit processors for decades. Good, so where are these CPUs? In fact, Intel, AMD and others already make 64-bit processors. So where is the 64-bit revolution for the average user? Stuck in the marketing department, that’s where.
The magic words of economics are supply and demand. These ineffable market forces should operate with an invisible hand. Maybe they do if you take the long view (say, from somewhere around the moon), but up close in the computer industry you also see the hand of market manipulation and the control of supply. We tend to get what we want when the companies can figure out how they want to give it to us. The big chipmakers for personal computers, essentially Intel and AMD, love to spoon-feed incremental increases in CPU speed and capacity. They tend to initiate their products as new and high-end CPUs, usually for servers and professional workstations. Then they do a gradual rollout of less expensive versions to pay back development costs and milk the profits. Nowhere is this strategy more apparent than with 64-bit CPUs; however, in this case there’s more to the story.
We’re accustomed to the idea of processor speed, traditionally measured in megahertz (MHz) and now gigahertz (GHz). This has some value as a relative measure of processor capacity, but it’s misleading with 64-bit processors. The MHz rating of a processor tells you how many cycles (“ticks of the clock”) it can perform in a second, but it doesn’t indicate what it can do in each cycle. That’s where the bit capacity comes in. A 64-bit computer can handle more computations, memory, and input/output with each processor cycle. That’s why current 64-bit processors–such as the Intel Itanium, which runs at only 733 MHz–can be a more powerful processor than the current Pentium IV, which runs at 2.2GHz.
Unfortunately, there’s a catch to using this power–software. Intel and AMD 64-bit processors are backward compatible with current 32-bit software, but using them that way is a waste. Software must be rebuilt with 64-bit compilers and, more realistically, redesigned for 64-bit operations. This includes the operating system. Consequently, the shift to 64-bit processors is a major change, not only in hardware but also in software. You may not recall the shift from 16-bit to 32-bit computers, which occurred about a decade ago, but that was a rugged time, early in the days of Windows. The change to 64-bit is an even bigger step.
The 64-bit mentality
What distinguishes 64-bit computing in the minds of designers, manufacturers, and typical users is its association with large-scale computing (huge databases, corporate applications) or very intensive computing (professional workstations). This level of computing is IT territory with procedures and attitudes that are far more systematic than the chaos of typical consumer use. I have heard IT people express horror at the thought of Joe User working with 64-bit computing. On top of all the other expenses involved with such a major change, this attitude helps to explain why corporations are in no hurry to implement 64-bit computers for personal users.
This time around, I think users are in a different economic frame of mind, and so the demand is different. It’s not that people don’t want faster computers, they just don’t want to pay much for them, which is another way of saying there’s no compelling need. Processor speed and capacity continue to climb but for most applications, the gain in productivity doesn’t increase nearly as much. Many applications targeted for 64-bit processors, such as very large databases, are hardly relevant to desktop computing, and some that might be, such as multimedia, are often handled by specialized processors (DSPs for sound, video, etc.).
Eventually, either a hot application will come along to drive the change to 64-bit computers for the average user, or else the manufacturers of 64-bit hardware and software will simply throw products into the personal-computer market (via computer manufacturers) and let declining prices and the march of progress carry them along. Voice recognition might be the killer application, or maybe it’ll be streaming media–but now nobody knows. For the time being, the mystery of the missing 64-bit computer is relatively simple: You won’t see 64-bit computers on your desktop until somebody’s convinced you’re willing to pay for them.