|
Thursday, June 03, 2004
Ng: Intel renames chips
By Wilson Ng
Wired Desktop
BACK THEN. Intel, the world’s largest processor company that powers most personal computers, is introducing a new name of its processor. Before we tell you what it means, let us go on a refresher course on what they used to name it. This will be somewhat technical, but if you follow carefully, you just might have a better idea what these numbers are all about.
Intel’s original processor was the 8088 (a hybrid 8 bit/16 bit computer). It ran at 4.77 mhz and contained about 150,000 transistors. That was about 25 years ago. Now, processors are 32 bit or 64 bit, run at around 3,500 mhz, and can contain more than 200 million transistors.
Intel named its first full 16 bit processor the 80286, and then the 32 bit processor the 80386. It used to run at around 10 mhz to 25 mhz. Then came the 80486, and then instead of the 80586, they renamed it the Pentium.
Up to that time, other than the name, there was only one distinguishable characteristic—the clock speed. The faster the clock speed, the better the performance. It made sense that a Pentium 100 mhz was faster and decidedly better (and pricier) than a 66 mhz Pentium.
Then came the Pentium II, and that was when it introduced its first variant—the clock speed would vary, and also the cache memory. There would be the Celeron, the Pentium II, the Xeon and the Itanium (which is the 64 bit).
A Celeron has the least amount of cache memory, so a Celeron 400 mhz was slower than a Pentium II 400 mhz. Alternatively, the Xeon, which has the largest cache memory, would be ideal for servers which can handle larger data transactions faster.
Then came other determinants which can also vary—the bus speed (the speed at which the processor communicates with memory and the other components of the computer), the architecture (the Hyperthreading Architecture or HT allows the computer to think there are two processors instead of one, and thus processes multiple tasks together faster), and also the battery consumption.
This become a little more confusing now. For instance, technically, a Pentium 4 2 gigahertz logically would be slower than a Pentium 4 2 gigahertz who might have faster bus speed, and Hyperthreading Technology. Also, a 1.8 Pentium M is positioned for notebooks, which would mean that it would have longer battery life, but because it has a bigger cache, it might mean it is as fast in most tasks as a 2.8 Pentium 4 machine.
Now comes another change, and this is probably the first change in over 20 years when the clock speed is very much de-emphasized. Basically, what will happen is that the processors will be named according to 300, 500 and 700 series models. If you are familiar with how BMW names its cars, that would be it.
As we know, the BMW 300 series is for smaller cars for the younger single generation which they drive themselves. The 500 series will be mid-range cars for affluent families, while the 700 series would most likely be big cars for the successful company executives.
Intel will be naming it the same way, taking into consideration all the things enumerated above. The naming convention will be Processor family + 3 digit number. So in the future, you will see something like Intel Pentium 4 535, or Intel Pentium M 720, or Intel Celeron M 320 etc.
Intel hopes that it will be easier to name and identify processors based on user experience, performance expectations and value proposition instead of being concerned about specific numbers on bus speed, clock speed or cache size.
More on these soon!
(email: wilson@esprint.com.)
(June 3, 2004 issue)
Write letter to the editor.Click here.
Join the Sun.Star message board.Click here. |
|
[return to top]
[home]
[network page]
|
LOCAL NEWS
BUSINESS
OPINION
SPORTS
LIFESTYLE
FEATURE
SUPERBALITA
WEEKEND
|
