Assimilator1
I have a P-3 600E overclocked at 800 giving a 133 bus speed and 133 memory Cas 3.
I have a Duron 850 on a K7SEM with 133 memory Cas 3. The speed is listed as supposedly 1200. Now I do not know how the overclock is achieved either by pencil lead trick or by changing the bus speed. If by bus then the bus speed would be 141mhz.
First off I use what I could to get up and running as quickly as possible. This was a diskless Linux used as a router hacked to run SETI and the owner doesn't know lots and says that it is not optomised and may lose 15%-20% cause of hardware.
Now what I can get is percentage completed and how many seconds has been run.
What I get is horrable times.
However the time is near the same for both systems.
Right now the Pentium 3 is ~43% completed at 16046 seconds. So what I do is take 100% or 100/42.766=2.338 and 2.338x16046seconds= 37520/60seconds=625minutes 625minutes/60= ~10.442 hours
Now if I do the same for the Duron 37.2% and 13852 seconds gives ...
10.34 hours
As you can see the times are similar.
Now if I look at a time for a Pentium 3 that is similar it is close or you can X percent difference the amount stated by other posters.
However when you compare the work times for the P-3 and the Duron they are the same.
But when you compare THIS Duron to the Duron time you posted it is like 50% off.
This leaves one of 4 things.
1)My duron is overclocked by pencil instead of bus
2)The cache makes a world of difference cause of this strange hacked FreeSCO Linux router OS
3)The OS has been optimised somehow in favor of the P-3 setup
4)A 5 hour work time is remembered incorrect.
Ok I have a old Duron that came from RC/Roy via Geoffs. I don't know anything about it except it is 700 and the boot sped says 927, I don't know if this is just bus or lead pencil and bus since the items are not in the bios.
(I put this stuff up as soon as possible for the MTR and left well enough alone. Will make alterations after.)
Time for this configuration AND this uses VC RAM that makes it a little faster.
~51% completed 17183 ... ~9.31 hours
Now This time is lower then the K7SEM 1200 soldered in duron.
1) K7SEM soldered in SUCKS or K7SEM MB sucks or this runs a 100 bus and pencil trick. Or BIOS helps lie like a salesman.
2)From my boards in front of me and this Linux OS a 5 hour time is wrong.
3)I'm not running this diskless OS for seti unless I have to.
A)Are cas 2 ram used to get 5 hour work unit. Or VC ram similar enough in speed increase and still 5 hour is wrong.
Also using this page and using ratios the work time comes out to ~5 1/2 hours.
http://www.teamanandtech.com/faqs/seti/WU_times.asp
muttley