Why not a bit more voltage?? You are talking about pretty massive overclocks and without a bit of extra voltage I do not see it working..

With a bit of extra voltage a Tualatin Celeron 1000A should do 1666mhz easy and also A Tualatin P3 1133-S should be able to reach 1400mhz easy at 166fsb. Also the p3-s 1266mhz would be fine but not at stock voltage..

I think I topped out my P3-S 1400mhz around 1700+ mhz on an Asus TUSL2-C so all the lower ones should be easy to reach higher fsb.

meljor wrote:

Why not a bit more voltage??

Because the motherboard that I use doesn't have the option to increase the voltage + I believe that overvolting shortens the life of a CPU 😉 But hopefully that mobo has the right dividers for AGP/PCI so I am just overclocking CPU 😀

I know there is a way to make a volt mod on pins, but I do not want to do that at this point. I think that 1.45v should be enough for P3-S 1133 to run at 1411, because the fastest Tualatins - 1400 MHz use that voltage.

Which motherboard do you own that supports 166MHz with correct AGP divider?

Success!

I found one coppermine that works stable @ 166 MHz with overvolting!!! 😁

It's one of the latest revisions: SL5FQ with cD0 stepping!

This one is pretty odd... At first I thought it's a Tualatin one - it has an IHS just like the one used on Tualatins, but the core is coppermine... and it is not a Coppermine-T. Just coppermine 😉

Take a look:
http://www.chipdb.org/data/media/845/Intel_Pe … INES_pinkie.jpg

What is also surprising for me is the fact that it doesn't get really hot! In BIOS the highest temperature was 51. In Everest the temperatures are wrong... I used simple AIR cooling. Nevertheless I stressed it for more than 1,5 hour! ZERO problems! 😉 Did some benchamrks tried different graphics cards, I even overclocked them - everything just works fine!

The results from 3D Mark 2001SE are astonishing! At default clocks it was 7823!!!!

Here is what I got when I overclocked Ti4600 a little bit

I even managed to run and OVECLOCK Radeon 9800 PRO to the clocks of Radeon 9800XT

Will a Radeon9800 PRO run on an overclocked AGP frequency of 83 MHz and be able to overclock itself so much?

Johnylyr wrote:

Which motherboard do you own that supports 166MHz with correct AGP divider?

I am afraid I can't tell you at this moment... sorry. I have to be 100 % sure that the right dividers are there, but so far I couldn't find any software that would prove it... I also have been working on a project related to this motherboard for a several months now, and I don't want to spoil anything... Just please be patient and wait untill the end of the year, I should be able to prepare all the rusults I got so far! It's going to be big!

There's a hint for those who pondering: few motherboards used ITE Super I/O chips at that time and even less used ITE IT8705F. You can start digging in that direction 😈

Ohhh sorry, I forgot to delete those information 😉 Please do not go into that direction - don't spoil my surprise 😜 I still need more or less two weeks to sum everything up 😉

VooDooMan wrote:

Success! […]
Show full quote

Success!

I found one coppermine that works stable @ 166 MHz with overvolting!!! 😁
It's one of the latest revisions: SL5FQ with cD0 stepping!
This one is pretty odd... At first I thought it's a Tualatin one - it has an IHS just like the one used on Tualatins, but the core is coppermine... and it is not a Coppermine-T. Just coppermine 😉

Nice !! 😀

The distinction between Coppermine-T and Coppermine is not just the metal IHS, the Coppermine-T underwent some changes in, among others, voltage management. Some high-end Coppermines also came with the metal sinks to make it more confusing.
Coppermine-T's can be distinguished by the part numbers that contain the string [80533], like these:
RB80533PZ800256
RK80533PZ866256
RK80533PZ933256
RK80533PZ001256
RK80533PZ006256

Coppermine part numbers contain the string [80526] and Tualatin part numbers contain the string [80530]. The problem is correlating the part numbers with the spec numbers because occasionally there is conflicting info on the web.

Call it luck! I have found another one that works at 166 MHz FSB! ;D This in is also with the IHS... It's SL5B3 - runs as stable as SL5FQ 😀

Here are the pics:

I also tried SL4C8 - this one won't even boot at 166 FSB... but it has different stepping.

PARKE wrote:

Nice !! :-) […]
Show full quote

Nice !! 😀

The distinction between Coppermine-T and Coppermine is not just the metal IHS, the Coppermine-T underwent some changes in, among others, voltage management. Some high-end Coppermines also came with the metal sinks to make it more confusing.
Coppermine-T's can be distinguished by the part numbers that contain the string [80533], like these:
RB80533PZ800256
RK80533PZ866256
RK80533PZ933256
RK80533PZ001256
RK80533PZ006256

Coppermine part numbers contain the string [80526] and Tualatin part numbers contain the string [80530]. The problem is correlating the part numbers with the spec numbers because occasionally there is conflicting info on the web.

Thanks for the info! This is very useful 😉

Just to sum up: 1250 MHz CPU Clock on 166 MHz bus without increasing the voltage achieved 😉 Next step is to find SL4Z2 and try to run it on 1411 MHz clock...

It's funny that Coppermine PCUs run so good without overvolting, while I was not able to get a stable 166 bus on ANY of my tualatins without overvolting....
Let's just hope that I will be luckier with SL6BW once I get one in my hands 😉

Next step is to find SL4Z2 and try to run it on 1411 MHz clock.

Most likely not possible without voltage increase. Heck, even 1250mhz might not be so stable after few hours of Prime95...

VooDooMan wrote:

Next step is to find SL4Z2 and try to run it on 1411 MHz clock...

I picked up two of these a while ago. Didn't think there was anything special about them?

VooDooMan wrote:

Thanks for the info! This is very useful 😉

Just to sum up: 1250 MHz CPU Clock on 166 MHz bus without increasing the voltage achieved 😉 Next step is to find SL4Z2 and try to run it on 1411 MHz clock...

Let us know if it does...

re the info, I found this website useful for research:
http://pcrebuilding.altervista.org/2/INTEL+SS … entium+III.html

You can click on the links in the right column for details.
The whole caboodle is downloadable too via clicking on
[ 244453-060_Intel_Pentium_III_sspec_cpuid (CSV)]
at the top.

The thing with the Coppermine-T is that the term may not be 'official', I have not been able to find any direct references in Intel documentation. Attached what seems to be the latest/last Intel PIII update.

I was able to get my PIII-S 1400 running at 1.68GHz (160FSB) on a TUSL2-C. Not sure if it was the CPU or motherboard that was preventing me from going any higher. Nevertheless, it was stupid fast at that frequency, and managed to pull even with a 2.4A P4 in SuperPi and 3DMark01. But the 512MB memory limit just wasn't enough for my extreme PIII needs, which is why I now run that CPU at "only" 150MHz FSB on a Apollo Pro266T board, with 2 gigs of DDR.