Lahey Support
08-15-2003, 01:33 AM
As a note to these times. P4 vs Athlon times depend largely on the code.
Where the P4 has a higher clock rate, the Athlon is able to perform many
more calculations per clock cycle. Our experience here and with many
customers is that and Athlon at a significantly slower clock rate
outperforms the P4. Due to the processor designs, this will change due
to the way code is designed. I do not doubt these times, though my
experience is this is a rare anomaly rather than the norm.
Damon
Larry Wigton wrote:
>Hi Klaus:
>
>Concerning these CPU times:
>
>************************************************** ******************************
>
>System CPU time
>
>Athlon 800MHz 100 %
>Athlon 1200 MHz ~ 70%
>P4 1800 MHz ~ 32%
>
>************************************************** ******************************
>
>You did not mention the memory being used by these systems. However it is
>reasonable to assume:
>
>Athlon 800Mhz was using 133MHZ SDRAM
>Athlon 1200Mhz was using 200MHZ DDR
>P4 1800Mhz was using 400Mhz Rambus
>
>If all machines were memory bandwidth limited then I would expect:
>
>System CPU time
>
>Athlon 800MHz 100 %
>Athlon 1200 MHz ~ 67%
>P4 1800 MHz ~ 33%
>
>which is close to what was observed.
>
> Larry Wigton
>
>
>----------------------------------------------------------
>To unsubscribe, send to [address removed] the following
>as the first and only line of the message body:
> unsubscribe fortran
>----------------------------------------------------------
>
>.
>
----------------------------------------------------------
To unsubscribe, send to [address removed] the following
as the first and only line of the message body:
unsubscribe fortran
----------------------------------------------------------
Where the P4 has a higher clock rate, the Athlon is able to perform many
more calculations per clock cycle. Our experience here and with many
customers is that and Athlon at a significantly slower clock rate
outperforms the P4. Due to the processor designs, this will change due
to the way code is designed. I do not doubt these times, though my
experience is this is a rare anomaly rather than the norm.
Damon
Larry Wigton wrote:
>Hi Klaus:
>
>Concerning these CPU times:
>
>************************************************** ******************************
>
>System CPU time
>
>Athlon 800MHz 100 %
>Athlon 1200 MHz ~ 70%
>P4 1800 MHz ~ 32%
>
>************************************************** ******************************
>
>You did not mention the memory being used by these systems. However it is
>reasonable to assume:
>
>Athlon 800Mhz was using 133MHZ SDRAM
>Athlon 1200Mhz was using 200MHZ DDR
>P4 1800Mhz was using 400Mhz Rambus
>
>If all machines were memory bandwidth limited then I would expect:
>
>System CPU time
>
>Athlon 800MHz 100 %
>Athlon 1200 MHz ~ 67%
>P4 1800 MHz ~ 33%
>
>which is close to what was observed.
>
> Larry Wigton
>
>
>----------------------------------------------------------
>To unsubscribe, send to [address removed] the following
>as the first and only line of the message body:
> unsubscribe fortran
>----------------------------------------------------------
>
>.
>
----------------------------------------------------------
To unsubscribe, send to [address removed] the following
as the first and only line of the message body:
unsubscribe fortran
----------------------------------------------------------