Summer1 said:
Oops: I meant to write:
To be clear:
A 1995MHz CPU is 1.995GHz which is approximately 2.0GHz and a 233MHz CPU is
.233GHz?
If the 1.995GHz rounds up to 2.0GHz, what does the .233GHZ round up to?
We round them, based on a knowledge of how the clock was generated
in the first place.
On your processor
CPU input clock = 100MHz
FSB = 100 x 4 = FSB400
Core_clock = 100 x 20 = 2000MHz = 2.0GHz
Due to small errors in the generation of the 100MHz signal, and
measurement errors in programs attempting to measure the speed,
the value printed on the screen won't be exactly 2000MHz. So
if we see 1995, we *guess* the correct value is 2000. Based on
knowing it involved FSB400, a 100MHz input clock, and some
multiplier value. The nearest multiplier is 20, so we guess
the correct answer is 2000MHz.
For an example of a 233MHz processor, I can look here. See PDF page 28.
http://download.intel.com/design/PentiumII/datashts/24333503.PDF
BCLK Multiplier Core Clock
66.67 7/2 233.33
In the case of the 233MHz processor, the precise value might be
233.33MHz, but for convenience, we might round that to 233MHz as
it is shorter to write. Apparently, the external clock in that
case is 66MHz, and the multiplier scheme supports half integer
values. Many subsequent processor families have also featured
half integer capability in clock synthesis, so you still run
into it.
When you see someone give an answer here, it is possible they
cheated and looked it up. Or, they may have first hand
experience with the technology when it existed.
Paul