P2.5 at 2.97

[Altheus]

I'd thought I'd let you all know how I got on with my new CPU.


P2.5Ghz (c1) Retail
Retail Heatsink fan *using thermal pad*
512MB PC2700 Samsung
Abit BE-7


I've managed to get it up to 2970 at a 119 FSB with a 10% jolt in CPU voltage. (47 idle, 53 max)

I also managed to get it running at 3Ghz with a 120FSB and a 15% jolt, however even though Sandra worked, 3D Mark failed to keep focus. (53 idle, 57 max)

I think even with some Arctic Silver I'll hit well over the 3Ghz mark, and with some proper cooling even more so.

I think this is pretty sweet for a bog standard fan/heatsink. I'll let you know how I get on with some arctic silver too.

[toastedzergling]

Shouldn't the 2.53 P4 has 133FSB?

[hc1001]

there is a p4 2.5 100fsb version and a 2.53 133fsb version.

[Wommalong]

So which is beter and why get the 100MHz version? just curious;)

[Altheus]

Well, I went for the 2.5GHz because it is a C1 Stepping chip, also because it ran at the slower bus speed I can clock it higher, and keep the PCI/DDR bus stable. It's also one of the most overclockable chips available on the market today, people are squeezing out 4.4Ghz with liquid nitrogen, and squeezing 3.9Ghz with phase.

[markodude]

The 2.50 C1 isnt all that, the only reason it gets so high on LN2 is cos it has the highest multiplier and it would be unfeasible to get another P4 that high without your FSB reaching over 200mhz.

Most boards allow PCI/AGP lock nowadays so that shouldnt be an issue. I have this chip too - sure it will run at 3ghz all day, but try Prime95 and see how long it lasts - you will need to go a step up on your cooling and have water/phase change to get stability at that speed I reckon - I have had better luck with my 2.53 which I got cos the 2.50 wasnt great.

Memory bandwidth is also a major issue - even if you get it to 3ghz stable on a 120FSB, your memory bandwidth will be at best DDR320 on a 3:4 divider. In contrast my 2.53 at 3.0ghz runs a 158FSB and DDR420, which will far outperform the 2.50 at the same speed......

Sorry, this chip isnt worth it unless you have major cooling and can reach at least 133FSB....you should have gone for a 2.53 (c1 if possible) or a 2.66 if you can afford it in my opinion. Ive learned from my mistake.

here are my benchies for the 2.50 (http://www.groveindependentschool.co.uk/benchie.htm)

[Wommalong]

Originally posted by Altheus
Well, I went for the 2.5GHz because it is a C1 Stepping chip, also because it ran at the slower bus speed I can clock it higher, and keep the PCI/DDR bus stable. It's also one of the most overclockable chips available on the market today, people are squeezing out 4.4Ghz with liquid nitrogen, and squeezing 3.9Ghz with phase.

WTF??? thats fast very fast...:burn: what a way to 'spend' money;)

[Altheus]

The 2.50 C1 isnt all that, the only reason it gets so high on LN2 is cos it has the highest multiplier and it would be unfeasible to get another P4 that high without your FSB reaching over 200mhz

I think you'll find that's my point exactly. The higher multiplier means that once I've got a decent cooling system I'll reach far higher speeds, with a 133 FSB, I'll be hitting 3.3Ghz, which I feel I can expect, considering the stability and speed I'm getting with it at the moment, (13 hours solid games playing, benchmarking and divx encoding).

While I may not be hitting my memory speeds just yet, once I upgrade the board come January with Dual DDR, I'll be far ahead. Also it's C1 stepping so Multithreading is not too far off either.

I analysed my purchase and feel it's the finest overclocker, and the most future proof CPU on the market at the moment.

Perhaps if something is to go wrong I'll change my mind and learn the hard way but for now I'm flying, 18% increase with bog standard cooling. I've seen 52% increase with phase vapor stable.

Say what you like but those facts speak for themselves

[markodude]

Altheus - I wont argue over this, but the facts you pointed out apply to almost any P4.
This chip is not the finest overclocker, its comparable to other P4's and to be honest I think its worse than a lot of the B0 stepping chips. My own experience tells me this as I have both a 2.50 and a 2.53.
Once you get Dual Channel DDR you will be far ahead of who? Anyone running single channel DDR yes, but the 133FSB chips that can currently get 160ish FSB and DDR420+ will be even faster on Dual Channel.
a 2.50 at 3.3ghz 133FSB and DDR354 (3:4 divider) is nothing like as fast as 3.3ghz on a 2.53 at 3.3ghz with 175FSB and DDR466.....
Dont get me wrong your rig will still be fast, just not that fast.

[markodude]

"Intel is moving all of its desktop P4s that ran using the 400MHz system bus to the newer 533MHz system bus. At the same time, it is rapidly phasing out the Pentium 4s with only 512K of cache"

http://www.theinquirer.net/?article=6194

Still think the 2.50/400 is future proof?

[micamica1217]

Originally posted by markodude
The 2.50 C1 isnt all that, the only reason it gets so high on LN2 is cos it has the highest multiplier and it would be unfeasible to get another P4 that high without your FSB reaching over 200mhz.

Most boards allow PCI/AGP lock nowadays so that shouldnt be an issue. I have this chip too - sure it will run at 3ghz all day, but try Prime95 and see how long it lasts - you will need to go a step up on your cooling and have water/phase change to get stability at that speed I reckon - I have had better luck with my 2.53 which I got cos the 2.50 wasnt great.

Memory bandwidth is also a major issue - even if you get it to 3ghz stable on a 120FSB, your memory bandwidth will be at best DDR320 on a 3:4 divider. In contrast my 2.53 at 3.0ghz runs a 158FSB and DDR420, which will far outperform the 2.50 at the same speed......

Sorry, this chip isnt worth it unless you have major cooling and can reach at least 133FSB....you should have gone for a 2.53 (c1 if possible) or a 2.66 if you can afford it in my opinion. Ive learned from my mistake.

here are my benchies for the 2.50 (http://www.groveindependentschool.co.uk/benchie.htm)

the sad thing is that I agree with you..."nothing is future proof."

yet this would be a great chip for rambus.

the best thing about rambus is that you can use almost any CPU and run at any FSB and still have good if not great bandwith.

don't listen to peeps who say you can't OC easy with rdram, or that mobos can't go to high FSB.....they just don't know what they are talking about.

this is not a bad CPU....maybe not "my" first choice, but with good cooling it should be fine.

markodude, not evan your DDR mobo will be future proof.
and GB may be craped on by a quick release of DDRII mobos.

I hear your points, yet he's happy. so what are you trying to prove???

I say enjoy your fine cpu, and nice OC btw.:)

and please get a better HS(don't forget case cooling)....getting below 50c at full lode will help with the life span as well as stability.

good luck
mica

[MS]

He has a point about Dual Channel DDR and the 2.5 P4. As was already suggested by this website, DCDDR P4s will have bandwidth to spare. DualChannel 2700 (let alone PC3200) would be overkill, so the 2.5 will get as much use out its 4300MB/s as a lower multi p4 with 5500MB/s. It wont be an issue.

I have the 2.5 as well with Vapor Cooling (Prometia). With very little tuning or screwing with settings I was getting 3660MHz out of it with DDR at around 195MHz. Id like a bit more, but quite good results I think. Imagine my setup with a DC DDR board?



Matt

[hipro5]

Originally posted by MS
He has a point about Dual Channel DDR and the 2.5 P4. As was already suggested by this website, DCDDR P4s will have bandwidth to spare. DualChannel 2700 (let alone PC3200) would be overkill, so the 2.5 will get as much use out its 4300MB/s as a lower multi p4 with 5500MB/s. It wont be an issue.

I have the 2.5 as well with Vapor Cooling (Prometia). With very little tuning or screwing with settings I was getting 3660MHz out of it with DDR at around 195MHz. Id like a bit more, but quite good results I think. Imagine my setup with a DC DDR board?



Matt

You've got a 2,5GHz with 100MHz or a 2,53GHz with 133MHz bus???

[MS]

Well, um, a 2.5GHz as I said. The 100mhz FSB one (to avoid any more confusion). You full load temps look better than mine BTW.. I was running at 1.78v core, perhaps that causes the difference?



Matt

[markodude]

"I hear your points, yet he's happy. so what are you trying to prove???"

Yeah Mica you are right - Im sorry if I seemed out of order, I just remember how annoyed I was with myself for buying this chip (2.50/400 C1).

At the end of the day if you are happy thats the main thing.