CPU "speeds"?

  • Thread starter Thread starter Xeno Chauvin
  • Start date Start date
X

Xeno Chauvin

Not to get too complicated.
Back in the "old" days we had 286,386,486.
Then we moved to MHz and they progressed
form 75 to whatever, then we went to GHz!
Somewhere in there I bought a 1.8GHz machine.
Now I'm looking at a new machine that is
2400MHz or ,I guess, 2.4GHZ?
Are we now back to MHz's and not GHz's?
Thanks
Xeno
 
Pentium likes to list theirs in GHz, while AMD goes for MHz. Keep in mind
that when AMD list a (for example) "Athlon64 3200+", the actual clock speed
is around 2000 MHz, or 2.0 GHz. The 3200+ is a Performance Rating and not
clock speed ( the PR based on comparison to Intel). Independent benchmarks
have shown the PR comparison to be relatively accurate.

AMD chips do more "work" per clock cycle than Intel, which is the reason for
their Performance Rating naming of processors- it's all about competition
and sales.

And you didn't want to get too complicated!

Fitz
(about to get flamed from the Intel crowd)
 
Xeno said:
Not to get too complicated.
Back in the "old" days we had 286,386,486.
Then we moved to MHz and they progressed
form 75 to whatever, then we went to GHz!
Somewhere in there I bought a 1.8GHz machine.
Now I'm looking at a new machine that is
2400MHz or ,I guess, 2.4GHZ?
Are we now back to MHz's and not GHz's?

Clock speeds are far less important than they used to be. It's not
just a matter of the clock speed, but also of the chip design.

Intel and AMD got fixated on clock speed for years, because they
(Intel in particular) believed that the market judged processors based
solely on the announced clock speeds (and they were probably right,
and that's probably still true). But different processors can execute
differing numbers of instructions at any given clock speed, so clock
speed is only comparable among submodels of a given processor model;
when comparing two processors with different chip architectures, it
means hardly anything at all. And if one is comparing two different
computer architectures, all bets are off.

Remember also that processor power is not the usual limiting factor on
computer performance for ordinary desktop use. Memory and disk speed
are more likely to slow things down. A lot of the time you spend
waiting for an application to start on the PC is time spent doing
hundreds of physical I/Os to your disk. Disks have only increased in
speed by a factor of ten or so in forty years, while processors are
closer to a million times faster.
 
Back
Top