problem in BOGOmips

problem in BOGOmips

[Ashish Gupta]

Hi,
	I want to use bogomips as the indicator of CPU capability for
different architectures. I have found following values from /proc/cpuinfo
for different CPUs.

	MHz		bogomips version
	233 intel	233	 2.2.9, 2.0.36
	166 intel	331	 2.2.9
	450 AMD-K6	900	 2.2.14
	800 intel	1600	 2.2.16

Why there is a exceptional behaviour of bogomips for 233 intel ?
If there is some patch or changes then please indicate so that i can use
it. 

Thanks in advance
Ashish Gupta




-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/

Re: problem in BOGOmips

[Mathieu Chouquet-Stringer]

gashish@cse.iitk.ac.in (Ashish Gupta) writes:

> Hi,
> 	I want to use bogomips as the indicator of CPU capability for
> different architectures. I have found following values from /proc/cpuinfo
> for different CPUs.

You got your answer. Try pronounce bogomips and you will hear bogo like in
bogus. The Bogomips don't represent anything but a way to have a good delay
mechanism... So you can't compare them between CPUs!
-- 
Mathieu CHOUQUET-STRINGER              E-Mail : mchouque@e-steel.com
     Learning French is trivial: the word for horse is cheval, and
               everything else follows in the same way.
                        -- Alan J. Perlis
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/

Re: problem in BOGOmips

[Richard B. Johnson]

On Sat, 10 Feb 2001, Ashish Gupta wrote:

> Hi,
> 	I want to use bogomips as the indicator of CPU capability for
> different architectures. I have found following values from /proc/cpuinfo
> for different CPUs.
> 

In a nutshell, you can't!

There is the requirement for some small delays when performing
certain kernel functions. Therefore, upon startup, a delay
loop is calibrated by checking how many times the delay-loop
gets executed between two or more hardware interrupts. The
hardware interrupts occur at quite precise intervals so the
calibration can be quite accurate.

A side-effect of the calibration is a loop-count variable.
I should have been called "Loop count", but instead, it
was called "bogomips". Many persons assume that "bogomips"
has some relationship to CPU speed or machine capability.
It doesn't. It can only be claimed that a CPU with a faster
core clock might result in a larger loop-count. In fact
a CPU with a faster core-clock might even show a lesser
loop-count because of differences in internal cache management
or branch-prediction with different CPUs.


Cheers,
Dick Johnson

Penguin : Linux version 2.4.1 on an i686 machine (799.53 BogoMips).

"Memory is like gasoline. You use it up when you are running. Of
course you get it all back when you reboot..."; Actual explanation
obtained from the Micro$oft help desk.


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/

Re: problem in BOGOmips

[Brian Gerst]

Ashish Gupta wrote:
> 
> Hi,
>         I want to use bogomips as the indicator of CPU capability for
> different architectures. I have found following values from /proc/cpuinfo
> for different CPUs.
> 
>         MHz             bogomips version
>         233 intel       233      2.2.9, 2.0.36
>         166 intel       331      2.2.9
>         450 AMD-K6      900      2.2.14
>         800 intel       1600     2.2.16
> 
> Why there is a exceptional behaviour of bogomips for 233 intel ?
> If there is some patch or changes then please indicate so that i can use
> it.
> 
> Thanks in advance
> Ashish Gupta

Bogomips are simply a timing delay loop factor.  They do not reflect CPU
performance at all.  Bogomips is short for bogus mips.

--

				Brian Gerst
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/

Re: problem in BOGOmips

[Prasanna P Subash]

[-- Attachment #1: Type: text/plain, Size: 1358 bytes --]

the bogomips algorithm changed between kernels. Now it uses the tsc register. Your
bogomips should typically be about 2*processor mhz.

On Sat, Feb 10, 2001 at 02:35:05AM +0530, Ashish Gupta wrote:
> Hi,
> 	I want to use bogomips as the indicator of CPU capability for
> different architectures. I have found following values from /proc/cpuinfo
> for different CPUs.
> 
> 	MHz		bogomips version
> 	233 intel	233	 2.2.9, 2.0.36
> 	166 intel	331	 2.2.9
> 	450 AMD-K6	900	 2.2.14
> 	800 intel	1600	 2.2.16
> 
> Why there is a exceptional behaviour of bogomips for 233 intel ?
> If there is some patch or changes then please indicate so that i can use
> it. 
> 
> Thanks in advance
> Ashish Gupta
> 
> 
> 
> 
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> Please read the FAQ at http://www.tux.org/lkml/

-- 
Prasanna Subash   ---   psubash@turbolinux.com   ---     TurboLinux, INC
------------------------------------------------------------------------
Linux, the choice          | I've got a COUSIN who works in the GARMENT 
of a GNU generation   -o)  | DISTRICT ... 
Kernel 2.4.0-ac4      /\\  | 
on a i686            _\\_v | 
                           | 
------------------------------------------------------------------------

[-- Attachment #2: Type: application/pgp-signature, Size: 232 bytes --]