x2 4400+ Core 0 no like to O/C

[X2beach]

Well I am not sure how everone is getting there 4400+ overclocked to 2.7ghz mine pretty much doesnt even like 2500+ I am watercooled,

I recently removed the heatspreader. then applied AS5 and replied the Heatspreader, that did help with the temps 2-3c. and I just received some OCZ PC-4800 Plat Limted Edt Ram. didnt help. rest of my specs are in sig

[MMIQHW]

Just remove the heatspreader completly and try again.

[highoctane]

Sounds about like my 4400, core 0 is much weaker than core 1 and doesn't like going much over 2.5ghz, at least not prime stable, water cooling here as well.

I can boot into Windows fine at 2.7ghz but it won't even prime for a minute before stopping, not that prime stable is even 100% stable but if it can't pass prime at the very least something isn't kosher, seems to hit a stability wall around 2.5ghz.

I've been very tempted to remove the heat spreader completely but I've also been thinking about selling it on ebay and picking up another 4400.

[X2beach]

going to remove the IHS in a sec.. I will results in a abit

mine runs 2.7ghz in windows but if I primes 5secs crash

[Chris27]

hmm so I'm not the only one one of my cores doesn't like to even go above 2.4ghz

[fine]

here to . core 0 run out at ~2500 mhz and core 1 at 2800 mhz 1.55v vcore
i cant test higher , it get to hot.

EDIT
i have change bios from 711 to 602 .
and is now stable at 2600 mhz 1.55v vcore
http://valid.x86-secret.com/show_oc?id=28885

[NickS]

thats weird. that kinda sux actually, getting a rele good core then the other one being complete junk

[sai]

That doesn't make me wanna get a X2 soon

[X2beach]

ok.. 300htt*9 (2700MHZ) Running Stable with IHS Removed.. Full Load Temps 47C

I think the problem is the IHS doesnt make good contact with both cores.. but at anyrate I AM STABLE 1hr prime on both cores !

[X2beach]

ok here is my CPUZ Validate

http://valid.x86-secret.com/show_oc?id=28887

[highoctane]

Damnit, you're gonna make me pull the IHS!!!!

I've been thinking about the posibility that my IHS isn't making proper contact somewhere. I went from a XP120 air cooler, to a basic Swiftech h20 120 water setup, to now a much more powerfull water setup (Storm block, D5 pump, BIX2 radiator, 1/2 plumbing) with temps STILL staying in the same range, tired AS3, AS5, & ceramique!

The water and radiator temps are hardly even noticably warmer than anything else yet cpu temps under full load ALWAYS get right to 50C+ with all three different cooling setups I've used.

I think its time to pull the IHS after hearing your results! :thumbsup:

[Reflex1]

ok here is my CPUZ Validate

http://valid.x86-secret.com/show_oc?id=28887

so this is without the IHS? ...stable?

[X2beach]

yep.. Stable as a rock.. I havent done the HARd Core 24hr prime.. but all games work pime ran for 1hr. SuperPI running.. anything else you like to me run?

[X2beach]

highoctane: dont be afraid to remove the spreader, its very easy, if you need some help send me a pm. it really helps with the temps.

[fine]

my cpu is 40C idle and 58-60C load at 2600 mhz 1.55v vcore.
and it take 2 sek to go from 40 to 60C
my "water cooler( I dont now to english word for it)" on the cpu is 30C on the outside.
maybe the IHS have to come off.

[X2beach]

fine: yeah remove the ihs spreader, its holding you back

[sai]

Damnit from 2500 to 2700

[highoctane]

You're not going to believe this but after removing my IHS my full load temps, dual priming, dropped from 50-53C to 34-36C, everything else the same other than removing the IHS.

WOW is all I can say, was a little nervous trying to get the razor between the die and ihs but once I got the feel for how deep to cut and how much pressure to use it went alright, a little nerve racking still.

Anyways, hopefully I'll see similar gains as you, I definitely had a contact issue with the IHS!!!!

EDIT: Ok set things to 10x270 and 1.6vcore to start with, been priming for 3 minutes now, thats a record for this core to keep priming at 2.7ghz, max load temp is 37C at 2.7ghz and 1.6vcore, priming each core with affinity set to run on each core independently, up to five minues now from me typing so slow!

[hoop]

ok ima warn yall in advance but im a tard when it comes to this but if i dont ask ill never get it but here it goes.
i order a 4400+ or the 4000+ and its on the way but can you tell me what
(IHS) is because im going to have to do this as well.
maybe a pic of one of yall actually removing it.

p.s. should i have just gotten the 4000+ over the 4400+ im asking because all i do is play cs 1.6 and source but anyways thx.

[MMIQHW]

Removing the ihs was the first thing I've done with all of my 3 a64 cpu's. The temps always dropped around 8-10°C under load. It very easy to remove it, just use a very sharp razorblade. Cut about 2-2,5mm throught siliconglue that is used to fix the ihs on top of core. To remount it perfectly, for example with as5 on the core is very difficult. Just forget about that an use your a64 without that good cooling killer

[X2beach]

4400+ is good if you want to multitask alot


also its really is amazing how much IHS Sucks!

[MMIQHW]

yeah, I've "only" a 4200+ @ 2904 ihs removed
the system is answering to demands much better in general. Really cool. I would have bought a 4400+, but it's very hard to get it in germany. Only the 4200+ and the 4600+ are available in quantities.

[highoctane]

Integrated heat spreader over the cpu die, you might want to leave it on. Depending on your particular heatsink you have a high likely hood of cracking and ruining the core and the mounting height for the heatsink must be lower to accomidate the lack of the heat spreader.

I've been on the fence about removing mine for a while now and finally decided to take the risk, a $600 risk at that.

Anyways I'm happy I did in the end but I accepted the risk, been priming for over 50 minutes now at 2.7ghz 35-37C reported full load temps.

I'm probably going to let it prime over night to see if it keeps going but so far so good. If it makes it through the night I'll start dropping the voltages some as I normally don't run my ldt at 1.4 and it probably doesn't need to be that high and the vcore may likely be able to drop just the same.

I'm guessing there must have been very poor heat transfer from core0 to the heat spreader while core1 had adequate contact, that might explain the large discrepency with core0 tapping out at a much lower clock than core1.

[MMIQHW]

cracking of the core is no problem for me so far. I use little bumpers, look like the ones amd used for the athlon xp cpus. Works perfect with my nexxos watercooler, the thermaltake big thyphoon and with a modified mounting with the zalman 7000 too. I've used 10 cpus, all without ihs and without problems during the last 6 months. Just be careful an it's no problem to use it naked

[lambusser]

highoctane...

Just in case this helps, here's my info. With help from 'xxmartin' and 'Major' I removed the IHS and it was the best thing I could have done Thank's gentlemen!

I've still want to dirve down my memory some, so I have another 'overnight' session planned with MemTest86+.

When I executed my "final" 2xPrime95 test, I used 'Custom' and 800MB on each instance...due to the 2GB of memory.

Enjoy,
Lam