How Well Were Pentium III to OverClock?

[slim142]

Hi

Did anybody ever overclocked Pentium III? Im talking about processors of 1.2Ghz and up. Were they stable? Showed very improved perfomance? Did you have to add voltage?

 

[deweycd]

I got an old pentium 3 for a server, I know that the pentium 3 were really good for not over heating, just make sure that you don't have one of the first pentium 3 they don't cool as well.

 

[TheMaster]

I got an old pentium 3 for a server, I know that the pentium 3 were really good for not over heating, just make sure that you don't have one of the first pentium 3 they don't cool as well.

My old pentium 3 800 didn't even have a fan over the tiny heatsink and it was never hot to the touch.

Over volt that bad boy and slap on a prescott heatsink... me thinks at about 2 ghz you could put a p4 to shame. :twisted:

Show us some benchmarks after you're done.

 

[jimytheassassin]

They never made a p3 over 1.3ghz right? Depending on what P4 you're talking the P3 was already better in some cases. .willamette core was aweful. Coppermine was a good proc, but i'm not sure you can expect much over 1.3 since you're gonna be held back by the boards fsb and multiplier options. I don't believe 2ghz would be possible, even on a great board and arctic cooling.

 

[Gelde]

:arrow:

I had several P3's and (coppermine cores) they all overclocked like mad - esp with a thouch more volts and a nice Alpha PAL - However - you seem to be referring to the Talutin cores (1.2ghz > ) and these were true MONSTERS with nearly all cores hitting 1.6ghz with ease and the best actually seriously getting close to 2ghz with a bit of effort. I had one running at approx 1750mhz and it bettered the AthlonXP at the time whilst killing the P4.

Damn shame intel let that design go in fav of the netslurp - hang on - Pentium M.....whats this under this sticker !........

 

[Multiplectic]

I've read somewhere (sorry but it's kinda late, and my memory starts to fail a little bit... :roll: ) that the "Costa Rica" PIII's were very good overclockers...
I'm not familiar with PIII technology, and I think "Costa Rica" is the codename for one of their cores...

 

[Gelde]

I've read somewhere (sorry but it's kinda late, and my memory starts to fail a little bit... :roll: ) that the "Costa Rica" PIII's were very good overclockers...
I'm not familiar with PIII technology, and I think "Costa Rica" is the codename for one of their cores...

:arrow:

Gotta correct this one - sorry - but the costa rica is a manufacturing place and not a core codename - Im sure you mean what I said earlier - Coppermine.......

 

[Multiplectic]

That's why I said I'm not familiar with the PIII's :lol: :lol:
And, BTW, you posted while I was writing...

 

[Gelde]

Ill be slower next time !.... :arrow:

 

[Herr_Fritz]

Well I can't comment on the 1.2 GHz and up Coppermines, but I do have a 1 GHz Coppermine that I am currently in the process of slowly overclocking. I'm taking this real slow as the AS5 hasn't even "set" yet. Right now its running at 1.08 (heh I know that's a laugh, but it IS 8%) with no heat increase at 100% utilization nonstop. I'll keep you updated as to what I finally reach. It'll definately be a learning experience for me, because I want this bad boy to scream (heh), and I'm wondering what's going to fail first. i.e. RAM or PCI bus speeds.

Specs:
ASUS CUSL2, 512 MB Mushkin CAS2 PC133 RAM (2 x 256), a NIC on the PCI bus and a AGP vid card, 1 HDD...nothing else


EDIT: Damn I love this...this thing keeps getting cooler and cooler...that Arctic Silver 5 "set" time is NO joke!!! 😀

 

[1Tanker]

Hi

Did anybody ever overclocked Pentium III? Im talking about processors of 1.2Ghz and up. Were they stable? Showed very improved perfomance? Did you have to add voltage?

I have a P3 1.2 and the best i've gotten out of it was 1446Mhz(160.68FSB), on an MSI 6309v5..with 512MB of crucial RAM. I'm not sure if the CPU was topping out, or the mobo, as my P3 866 wouldn't hit 160FSB(hit 160 once or twice, but very unstable) with the same RAM,mobo, and HS/F (SLK-900A+92mm Tornado).I had to raise the voltage to 1.65v to stabilize it, but i just run it at ~ 1400 now(155x9),as i had to run the RAM asynchronously(FSB-33) at over 155FSB, and there was a noticeable
performance hit with RAM running slower. Pretty peppy processor though at 150+FSB.

 

[1Tanker]

Well I can't comment on the 1.2 GHz and up Coppermines, but I do have a 1 GHz Coppermine that I am currently in the process of slowly overclocking. I'm taking this real slow as the AS5 hasn't even "set" yet. Right now its running at 1.08 (heh I know that's a laugh, but it IS 8%) with no heat increase at 100% utilization nonstop. I'll keep you updated as to what I finally reach. It'll definately be a learning experience for me, because I want this bad boy to scream (heh), and I'm wondering what's going to fail first. i.e. RAM or PCI bus speeds.

Specs:
ASUS CUSL2, 512 MB Mushkin CAS2 PC133 RAM (2 x 256), a NIC on the PCI bus and a AGP vid card, 1 HDD...nothing else

Unless you have a golden CPU, and some great cooling, you probably won't get a lot more, as 1080 is getting near the limit for Coppermine CPU's. The best i could get out of my 866 was ~ 1044, but not very stable. It didn't like running much over a Gig, stable. You might get ~1125(150X7.5). GL

 

[Herr_Fritz]

Max for the mobo with the 7.5 multiplier is 1245MHz (@166 FSB). I have the distinct feeling that I wont quite reach that. But it sure would be cool if I did. 😀

Do you think there will be heat issues with the RAM at FSB speeds approaching 150? I've never seen SDRAM with stock heatsinks or heatspreaders on them. If so I might try cooling them.

 

[1Tanker]

Max for the mobo with the 7.5 multiplier is 1245MHz (@166 FSB). I have the distinct feeling that I wont quite reach that. But it sure would be cool if I did. 😀

Do you think there will be heat issues with the RAM at FSB speeds approaching 150? I've never seen SDRAM with stock heatsinks or heatspreaders on them. If so I might try cooling them.

I don't think there should be too much heat trouble at 150. I usually have a fan in the case blowing somewhere near the RAM, which helps a little. sdram doesn't get anywhere near as hot as DDR though. IIRC the good Mushkin RAM didn't hit much over 150. They did put out some PC150, or PC166....can't remember off hand, but you'll most likely have to drop to CAS 3 over 150. GL

 

[Herr_Fritz]

Just went about reading some WAY old reviews of the Mushkin Rev.2 DIMMS and it appears I might be able to run them stable CAS2 @ 155ish. That seems like good news...16.25% overclock considering the coppermine can take it, which from the temps its generating now, I don't see any reason why it won't...I can't say I'd be in complete tears about that.

 

[Herr_Fritz]

Update:

Running at 1.17 GHz

Wierd thing is, I have yet to see a temp increase in the CPU or mobo. I figured they would go up a little.

 

[ANIA_fanboy]

I blew up a Coppermine 1.0.

OC'ed to 1.2 with AS3 and Swiftech WC. It was running just fine.

But, no! I needed more!

The increased involtage (don't remember what but it wasn't large) did it in I think.

It could have been my inexperience with WC or Vcore changing.

Anyway...good luck with a solid proc!

That good old days when FSB increases did so little (quad-pumped be damned 😛 )

 

[Gelde]

Well you are reaching the limits of even the BEST coppermine cores there - unless you actually have a TALUTIN core (celeron is 256k and p3 512k caches ?) also by running at - do you mean STABLE at ?.....ie prime95 for hours on end etc.....?......

I want another talutin !.....they were f'in gr8 o/c chips !>......mind you my 2.6c northwood at 306mhz fsb is damn nice too though.....if only amd offered me that kind of cpu id be back with them.....k6 was crap - original slot athlon was poo too.....k75 was ok, thunderbird took off fine and then palimino was better.....barton got the cake at 2.65ghz.....a64's are pretty much process limited tho.....oh well.....conroe next up.....ill lash that to within an inch of its life !........

 

[Herr_Fritz]

It's definately a Coppermine. The ASUS mobo doesn't support the Tulatin cores. Am I wrong, or did the Tulatin cores start at 1.2 GHz?

I'm not running Prime95. I figured Seti@home loading the proc 100% would be a good stress test. So far no problems. I also ran a series of SuperPi calcs to see if it was doing the math right. I think I'll go ahead and get it running Prime95 though. Just to see what results I get.

 

[1Tanker]

It's definately a Coppermine. The ASUS mobo doesn't support the Tulatin cores. Am I wrong, or did the Tulatin cores start at 1.2 GHz?

I'm not running Prime95. I figured Seti@home loading the proc 100% would be a good stress test. So far no problems. I also ran a series of SuperPi calcs to see if it was doing the math right. I think I'll go ahead and get it running Prime95 though. Just to see what results I get.

No, you're right. Tualatin P3's started at 1.13GHz.

 

[Herr_Fritz]

Not that it matters much, but the current speed is 1.18, which has been running for a few days now. Its running Prime95, and so far so good. We'll see if it's still stable in the morning.

Temps are a steady 49-50 / 33 at load.

EDIT:

Just figured out that this chip is a Coppermine-T which means cDO stepping revision. Apparently these were the very last Coppermines to be produced before the switch from 180nm to 130nm (Tulatin). From what I can gather the Cu-T was supposed to incorporate some of the tech that was going into the Tulatins. Intel apparently canned the Cu-T program, but not before they already made some PIII 1GHz chips with the cDO revision. Perhaps this is why I'm getting such good results.

 

[Herr_Fritz]

OK I think I just hit the wall with FSB increases. It runs stable at 1.177, but not at 1.185 running BOINC. Though for some reason it ran Prime95 blend-mode just fine at 1.185.

I'm really happy with this as I never had to increase my volts OR go to a CAS3 on the RAM. I think I'm gonna let it run as-is for a while to see if BOINC will lock it up eventually at 1.177.

I'll have to stress individual components to see what it is that is failing from here on out, which I don't think will be a whole lot of fun. However, I would so like to see 1.2 run stable. 20% OC on a chip that was pretty much at the end of its range has me hyped. We'll just have to wait and see if I can reach it.

P.S. this would be the STOCK MHz that slim142 wanted to talk about in this thread (though he has yet to reply to anything...wierd....) The Tulatin's ability to OC seems to be very good from the replies so far. I wonder if he found his answer elsewhere.

Fritz