INTEL, old processors:
AMD
| K6 II
(also named K6-3D) |
3Q 1998 | 300 to 350 | 9,300,000
(K6: 8,800,000) |
0.25
(K6: 0.25 or 0.35)) |
| K6 III
(also named K6-3D+ or Sharptooth) |
4Q 1998 | 400... | 21,300,000 ! | 0.25 |
| K 7 | 1st Half 1999 | ??? | ?? | less than 0.25 |
| New 6x86MX
(Cayenne technology) |
98 (2nd half) | 300 to 400 | ??? | 0.25 |
|
|
|
|
|
|
| Pentium Pro
(686 , P6) |
95 | 150 to 200
Maybe 233 ? |
5,500,000 | 0.32 |
| Pentium MMX (586 , P55C ) | 4Q 1996 | 166 to 266 | 4,500,000 |
0.35 |
| Pentium II (686) | 2Q 1997 | 233 to 333 | 7,500,000
(without the L2 cache) |
0.35 |
| Deschutes (686) | 2Q 1998 | 350 to 450 | ??? | 0.25 |
| Xeon (deschutes mobile) | 98 | 300 and more | ??? | ??? |
|
|
|
|
|
|
| Mendocino
(Celeron N°2 , 686) |
4Q 1998 | 300 to 333 | ??? | 0.25 ??? |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
The intel Merced :
200Mhz front bus; 460GXmobo; 4/8 INT, 4 FPU, 4 decoders.
And it will be a professional version used in Servers. A later version
should be adaptated for us.
The K6 III:
100Mhz bus
Cpu freq L1/L2 (cache)
Mobo freq L3 (cache)
Superscalar MMX (like K6 II)
bigger than K6 II but Super7 compatible.
The K7:
??????????
0.08 Microns !
New CPUs:
As you can see each of the three best CPU manufacturers are creating
processors with instructions enhancements.
To produce this new chips they must add lots of transistors, and this
is a problem for the futur size and temperature, because a large processor
is expensive, difficult to produce, and especially it would be too error
prone of calculation because of the temperature (electronigration). Thus
it is necessary to reduce the size of the engraving of the transistors.
A new technology !
The first CPUs for PC had a unit thickness which decreased by half
every two or three years. A new technology is necessary, and it is that
developed at the University of Texas which could be the solution well.
Because it would make it possible to thus engrave until a smoothness of
0.08 microns by using Ultra Violet which are particularly easy to produce
and the not very expensive ones. This price/quality ratio is very
favouring and is likely well to be adopted by the principal leaders of
the production of CPUs.
But we must wait,
This technology will require given it on level of the factories and
will thus ask for a time of adaptation. We will not see processors 0.08
microns on the market before less than one year. It should not be hoped
that the processors listed on this page profit from this evolution.