My Prescott 2.8E overclocks.

T

Tooden

n00b
Joined
Jan 4, 2004
Messages
16
Well, yesterday I went to frys and got myself a retail prescott. I was unsure of how good it might overclock.

System:

P4 2.8E SL79K Retail
Asus P4C800 deluxe bios v. 1015 board revison 2.0
Corsair twinx PC4000 1024MB
BBA 9800P

Water cooling specs:
Danger Den maze 4
Eheim 1250
1/2 ID tubing
Black ice extreme radiator w/ Sunon 12cm fan pushing 98 CFM

I plopped it in and it booted just fine, everything was A OK. Windows loads up, and as soon as I start a program, beep, computer turned off. I boot up again and see that my Asus is only putting like 1.3 volts to it, when on the box it says 1.4 is where the money is at.

I adjust the voltage and see that it aint crashing anymore. I load up prime, do a torture test, and then had a look at my temps...

CPU 45C

Holy jesus, my old P4 2.4C (max overclock was 3.0....) ran at an idle temp of 28, and full temp of 34. I now have this thing and it is facking hot! I guess everyone was right, these things put off HEAT!


I reboot the comp and set the FSB to 250 right off the beat at 1.4v. It loads up windows fine, temps are the same, and I am sitting pretty at 3.5ghz.

Unsatisfied that I havent crashed it yet, i bump up the fsb to 260. Im now at 3.64 ghz 1.4v. I load up windows, run a super PI of 1 million, and I do it IN 39 SECONDS!!

My best before was 48, this thing BLAZED past that. Unfortuantely, at this point it was not 100% stable, and it needed a voltage increase which Im not willing to give it.

So, to conclude upon what everyone has already concluded in other threads is, This thing is a fucking oven, I cooking up some eggs and probably will throw on the bacon next.

I just got this like 4 days ago, so once it runs a while and burns in, ill probably try to hit 3.8 ghz.
 
heheh, come back and tell me how the oc to 3.8 goes. u will screw that p4 up. u should use a water cooling if u want to go that high, and does this p4 have a limit to where they can go?
 
Originally posted by DJ PiLLz
heheh, come back and tell me how the oc to 3.8 goes. u will screw that p4 up. u should use a water cooling if u want to go that high, and does this p4 have a limit to where they can go?
Click to expand...

Next time read the post and you'll see that I do have water cooling...
 
Originally posted by DJ PiLLz
heheh, come back and tell me how the oc to 3.8 goes. u will screw that p4 up. u should use a water cooling if u want to go that high, and does this p4 have a limit to where they can go?
Click to expand...
wow, that is a depressing response...

All i can say is i can't wait to see LGA
 
Next time read the post and you'll see that I do have water cooling...
Click to expand...

Lol...tards will be tards... :). Anyway... I WAS thinking of getting a Prescott to play with. But decided not to. Not because they do not OC, they do. But because of the added stress they put on the rest of the components in the system. ATM the Prescott performance is lacking against Northwood in almost all areas at similar clock rates. Add in the mix the hotter than h3ll temps, strain on current MOBO's and PSU's and the Prescott looses any new hardware appeal to me. I'll probably just pick up a 3.0/3.2 Northwood and OC it to get me over the hump until the new AMD64 sockets release. GL with your new CPU and keep us posted. BTW what were your temps at 3.6??
 
Full temps at 3.6 were 48C. Keep in mind that this is water cooling. These things are very very VERY hot. But I like my prescott just the same :)
 
Originally posted by Tooden
.....so once it runs a while and burns in, ill probably try to hit 3.8 ghz.
Click to expand...
once the thermal paste (as5?) burns in?....
cuz over the last 4 years I have yet to se "burn in" make any diff in hardware overclocking..(quite the op actually...may be the thermal paste between the heat spreader and cpu breraking down/shifting)...except for thermal interface material.

congrats on yer new cpu...
i'd sur like to play around with one of those.
 
congrats man...thats a hell of a temp with watercooling tho...

what was your old cpu? That superpi time of 39 seconds is nice but my 2,4c at the same speed did 37 seconds. I guess its true about the northwood being faster clock for clock. You will smoke me at 4.0ghz tho, I would guess! My ram may also have something to do with that score though...actually, yeah, you know what, looking at your ram and assuming it isnt running 2-2-2-5 like mine is I would think its safe to say that the northwood and prescott clock for clock are about the same at 3.6-37ghz. Very interesting. congrats again.
 
Originally posted by computerpro3
congrats man...thats a hell of a temp with watercooling tho...

what was your old cpu? That superpi time of 39 seconds is nice but my 2,4c at the same speed did 37 seconds. I guess its true about the northwood being faster clock for clock. You will smoke me at 4.0ghz tho, I would guess! My ram may also have something to do with that score though...actually, yeah, you know what, looking at your ram and assuming it isnt running 2-2-2-5 like mine is I would think its safe to say that the northwood and prescott clock for clock are about the same at 3.6-37ghz. Very interesting. congrats again.
Click to expand...



then again, you did have a faster memory bandwidth/higher FSB.


Anand Tech found out that at 3.6ghz and up, Prescott beats Northwoods of the same stock speed, i.e. 2.8c vs. 2.8e.
 
Actually, the fsb and memory bandwith is identical

P4 2.4C

12x200 default, but it was at 12x250 = 1Ghz FSB, or 500 MHZ DDR speed

P4 2.8E

14x200 default, at 14x250, same memory speed as before
 
AnandTech's performance discovery isn't really a discovery at all. It's a fact that as you ramp up clock speeds, one thing stays the same - and that's memory bandwidth and latency. This disparity causes a diminishing return effect that can only be improved by increasing memory bandwidth and decreasing latency - this is what adding 1 MB of L2 gives you - a decrease in average latency to memory. However, there are still going to be cases where the P4C is going to outperform the P4E - the cases where memory latency doesn't matter (tests fit in cache, etc). The only way the P4E will outperform a P4C is when a P4E will reach a higher clockspeed than a P4C ever could =)
 
Nice clock Tooden.
I'm curious as to what voltage your reading under load on the P4C800 ? Mine is dropping pretty good under full load
 
Yea, mine drops a good deal too. Lets say I set vcore to 1.585, under full load it will drop to 1.5. At idle its pretty stable, always at what I set it at, which is nice, because before my prescott my board would overvolt from set vcore of 1.625 up to like 1.7.
 
Thanks, I am seeing the same thing, its running 1.52 idle, 1.46 load (+/-)
I've been running at 250fsb for 2 weeks now checking stablilty,
Seems to be 100% so far(knock on wood)
going to give it a couple more weeks to 'settle in' then I'll kick the crap out of it
 
Originally posted by BoBo 007
Thanks, I am seeing the same thing, its running 1.52 idle, 1.46 load (+/-)
I've been running at 250fsb for 2 weeks now checking stablilty,
Seems to be 100% so far(knock on wood)
going to give it a couple more weeks to 'settle in' then I'll kick the crap out of it
Click to expand...


Yea, Im at that exact point right now too. The more this things runs, the better chance of it overclocking higher? Or is that just wishful thinking on my part?
 
I dont know if it improves the chances, but I can hope:)
Its just establishing a frame of reference for what this core will do.
Then I'll have some compairision point when I start to go on up.
Instead of starting high and working down like I normally would do.
 
Chuck some bacon on for me while your there lmao:p But man these things sound hot.But it is the way chips are gonna go in the next couple of months by the sounds of it retail watercooling anyone;)
 
Any screenies or 3dmarks, im really interested in picking one of these processors up. Thanks in advance, nice Overclocks
 
But dont forget, the Prescott P4E's are first being released in their Socket 478 State. The 0.09 Die size should warrant a New socket which could help the longer pipelines & Latency issues, thus resulting in a much faster all round performance.

Most retailers are selling the Prescotts at a lesser price than the P4C's but I`m sure that will change once we are all ready to be raped for the New Socket design boards.

I really hope they dont use PCI express at the same time. It'll be nice to see one step at a time & be able to evaluate the benefits of PCI Express on its own.
 
I think I might just wait on the new motherboards
with new socket before I buy a PS.
The new mobo could give the PS much
better performance than what we are seeing
now from these older boards.
After all they were not designed from the gound
up for the PS cpu.
These new ones will.
 
well I'll give you some links to my prescott

prescott.jpg


3dMark2003
http://service.futuremark.com/compare?2k3=2149956

3DMark2001
http://service.futuremark.com/compare?2k1=7613414

PCMark2004
http://service.futuremark.com/compare?pcm04=232121

also 35secs in SuperPi 1M, so far 267 is my most stable setting running 2 instances of Prime for 24+ hours, I can boot into windows at 4 + Ghz so I'm still tweaking trying to increase my oc.
 
You must log in or register to reply here.
Back
Top