[rc5] Processor performance (was Core Autoselection)

dan carter daniel.carter at stonebow.otago.ac.nz
Mon Nov 3 20:23:58 EST 1997


On Mon,  3 Nov 1997 00:16:01 -0500 (EST), James P Blanding wrote:

>Excerpts from mail: 3-Nov-97 Re: [rc5] Processor Core Au.. by "dan
>carter"@stonebow.ot 
>> If you mean equivalent performance, then look at their P rating. The
>> souped up 486 is equivalent to a Pentium 75 (or a p100 when oc :}) the
>> K5-PR166 is equivalent to a Pentium-166, the 6x86 P150+ is equivalent
>> to a Pentium 150 etc etc
>
>be carefull here...i know for sure that cyrix (and possibly amd) chips
>run at a lower clock speed (maybe its internal clock speed...argh, cant
>remember) than specified by their "p-rating."  the p-rating is more of a
>marketing thing...although they do have to meet some standards in order
>to be able to advertise that a chip has a particular p-rating...just be
>careful with how you use that term basically..heh 

OK, lets be careful how we use the term. I meant, equivalent 'real
world' performance (not equivalent rc5 performance or equivalent quake
performance). Where real world means the speed to perform a wide
variety of common everyday operations.
Of course, one chip might be faster at the RC5 algorithim, another
better at running pentium optimised floating point operations.

MHz does not equal performance measure. My (oc)486 runns at 160MHz,
does that mean its faster than a pentium 133?

Cyrix realised the market would hurt them, because naive purchasers
would look at the MHz rating rather than any sort of benchmarks to
determine performance, so they use the P Rating. The P Rating a chip
gets is not based on a marketing decision, but is determined by the
chips performance running a number of benchmark programs.

P Rating = Performance based on benchmarks
MHz = number of clock cycles per second

Dont' ya just hate myths perpetuated by marketing from monopolistic
companies.... :)
+-------------------------------------------------+
|                  daniel carter                  |
|               phone +64 3 4736986               |
|        daniel.carter at stonebow.otago.ac.nz       |
|       1 crown st, dunedin, new zealand ;-}      |
+-------------------------------------------------+

----
To unsubscribe, send email to majordomo at llamas.net with 'unsubscribe rc5' in the body.



More information about the rc5 mailing list