|
Advice Beginners BIOS Guide CPUs Links Mainboards Memory Network Storage Video/Sound Cards Contact Forum SiteMap Sponsors WebNews Home |
. | . |
Prices: Mainboards ABIT ASUS Chaintech Shuttle Soyo Tyan CPU Intel P4 2.4C-800 P4 2.6C-800 P4 2.8C-800 P4 3.0-800 P4 3.2-800 AMD AthlonXP XP 1700+ XP 2000+ XP 2400+ XP 2500+ XP 2700+ XP 3000+ XP 3200+ Athlon64 Athlon64 3200+ Athlon64 FX-51 Opteron Opteron 240 Opteron 242 Opteron 244 Opteron 246 Memory Corsair Crucial Kingston Mushkin OCZ |
The 64 bit Challenge
Beyond the fading glory of past accomplishments, there are still new challenges coming up that go beyond the virtual needs created by the marketing machines and the brains behind them. We have gotten so used to progress that hardly nobody notices it anymore - the continuous change itself has become a static factor in our lives and the digital age is only beginning to start. Along these lines is the transition from today's 32 bit standard of computing to the next iteration, that is, the 64 bit processors and operating systems.
Before going into the intricacies of 64 bit computing, a brief retrospect of the transition from 16 bit computing to 32 bit is in place since this issue has been brought up over and again with respect to the long time it took to gain acceptance in the market place. In short, the 386 processor introduced 32-bit computing in 1988 but it took another 7 years until the software transitioned to a 32-bit operating system and applications with the migration to Windows95 - in 1995.
There are different ways of looking at this, one of which is to project a one by one comparison and, therefore, to predict a slow and cumbersome acceptance of 64-bit software. The second angle, which we believe is more realistic, is based on the faster turnover and hunger for an exponential increase of digitalization that is no longer bound by the fetters of DOS and the need of being an expert in order to master it - the pushbutton and pull-down age allows almost anybody to hit the ground running. What used to be available only in the book stores is at the fingertips of the internet with simple, or elaborate, tutorials, custom-tailored to anybody's needs. This versatility, however comes at the expense of a sizeable overhead. Which, in turn, needs the extra performance, especially if it is based on x86 ISA (as in Instruction Set Architecture).
X86 ISA: Ballast, Legacy or Inheritance?
Regardless of the insufficiencies of the x86 concept, the one thing it has going for it is the wide acceptance in the software market and the historical evolution from a 16 bit to 32 bit architecture including the addition of the x87 floating point coprocessor and so on and so forth. Like any evolutionary approach, the x86 platform carries with it a huge legacy, call it ballast, redundancy or inheritance but in the end it does not matter because the same redundancy that is conceived as a weight around the neck also warrants its versatility.
In other words, it is huge, it is bloated but at the end of the day it offers any support that the software industry has carried over for the past 20 years, even if it looks like a patch on a patch on a patch. It is Darwinian in the true sense of the word, survival of the fittest where fitness stands for optimal adaptation to the environment and environment encompasses more than just optimal computing strategies.
Everything taken together, including the predictable acceleration of evolutionary processes, we believe that the 32-to 64 bit transition will go much faster and with fewer bumps than the 16-to-32 bit transition, it is only the question of who throweth the first rock to start the avalanche.
64 bit CPU Architecture
64-bit CPU architecture is one of the buzzwords that have been thrown around over and again, unfortunately there is enough confusion about this subject to fill, er, the internet. In general, a 64-bit architecture simply means that the general purpose registers (GPRs) on the processor are capable of holding 64 bits as opposed to the 32 bits of the x86 ISA. To make it simple, for data, that means 64bit data and for instructions, it means instructions that - - operate on 64-bit data.
In summary, the only thing designated by the moniker "64" is the width of the data path and, consequently, the largest data size that can fit into one of the registers. Keep in mind here, however, that not necessarily all data are 64-bit, rather, the most common scenario is a mixture of data of different size. One additional important issue is that the just mentioned data path width only applies for integer and address data. x87 floating point registers are already 80 bits wide and vector data registers are 128 bits. With respect to the two latter, nothing has really changed, at least nothing that would specify the transition from a 32-bit to a 64-bit processor.
next page: => Registers, LOADs and STOREs =>
