The Official X2 Overclocking Discussion Thread

[dng29]

can someone verify this for me. my x2 4400 is pretty good. i got it at 11x255 right now for 2.8Ghz. cpu0 or core 1 is great. it will pass prime95 for hours no problem. cpu1 or core 2 sucks as it can't even pass the 1st iteration in prime95. is it correct to say that chances are alot of dual cores have bad core 2? i can play games on cpu0 by changing the affinity and it would be fine. if i change affinity to cpu1 to play games, instant crashes . i also tried aquamark 2k3 and 3dmark 2k5 with affinity set to both cpus, aquamark passes, but 3dmark fails half way thru. i'm running a pair of g.skill pc3200 and my vcore is set at 1.55. man i would love to have two good cores instead of 1 good and 1 bad .

[xgman]

can someone verify this for me. my x2 4400 is pretty good. i got it at 11x255 right now for 2.8Ghz. cpu0 or core 1 is great. it will pass prime95 for hours no problem. cpu1 or core 2 sucks as it can't even pass the 1st iteration in prime95. is it correct to say that chances are alot of dual cores have bad core 2? i can play games on cpu0 by changing the affinity and it would be fine. if i change affinity to cpu1 to play games, instant crashes . i also tried aquamark 2k3 and 3dmark 2k5 with affinity set to both cpus, aquamark passes, but 3dmark fails half way thru. i'm running a pair of g.skill pc3200 and my vcore is set at 1.55. man i would love to have two good cores instead of 1 good and 1 bad .

Most X2's are like this. My 2nd was closer. If you can, rma and try another. it may be worth it. If your reseller won't, AMD still might. Give it a try.

[xgman]

hey. got my x2 3800 from monarch yesterday and i've been running @ 10x2800 2x superpi 32mb stable. i took this ssss (semi suicide screen shot) a couple minutes ago @ 10x290. the 12v is being reported wrong on the ite smart guardian, fluctuating wildly between 11.71 and 11.91. bios, mbm and my dmm read it at between 12.18 and 12.22v. when i get a better fan and/or get the chip burned in so i can use a lower vcore and/or get some watercooling i will go for 3gig. not too sure about popping off the ihs, but dont want to go any higher than 1.55v until i get better cooling. also, my 3dmark2001 score has dropped from 33k to 29k from when i had sandy @ 2800mhz, all else being exactly the same. is this normal, losing 3d performance i mean? my sandra and everest mem scores are the same, my sandra cpu score is about exactly 2x my sandy score, etc, etc. just 3dmark n/g.
ada3800 daa5cd
ccbwe 0529 spmw (e6)


very truly yours,
politenessman

You have a nice 3800 there. Your clocks are near the top of the average 4400 x2's. maybe yields are improving. Actually my Sandra mem scores are less with X2 than with FX55 at roughly the same cpu speed.

[dng29]

Most X2's are like this. My 2nd was closer. If you can, rma and try another. it may be worth it. If your reseller won't, AMD still might. Give it a try.

actually this is my replacement cpu because the first one had the same issue. i don't think it's anything i did or the settings i have in the bios. i'm running pretty relax timings and low htt of 255. it sucks seeing everyone else have good and stable cores while i have 1 good and 1 bad. i hate to rma this one to amd since it'll take forever for them to send the replacement. i'll keep tweaking it and see what will happen. also is the 704 bios good for g.skill pc3200? i might try that bios. i'll also try to pump up the vcore and see what kind of results i'll have. any other suggestions would be greatly appreciate. thanks in advance.

[uniacid]

can anyone explain how I can run two instances of prime95 or another good program to burn in these two cpu's

[eclypse]

can anyone explain how I can run two instances of prime95 or another good program to burn in these two cpu's

Here you go.. Prime x' 2

[uniacid]

thx so much.

[rise4310]

just got mine today

ccbwe 0529spmw

damn these run hot. although that "max temp" program readsthe temp good to 69c

i'll be happy when zalman releases that new cooler.

[alpha0ne]

just got mine today

ccbwe 0529spmw

damn these run hot. although that "max temp" program readsthe temp good to 69c

i'll be happy when zalman releases that new cooler.

Hmmmm most I have seen only show upto 63c, like mine

[dng29]

i got a little better result with my 4400 x2 yesterday. i went into and change some memory timings and pump the vcore up to 1.6. i was able to ran superpi 1M and 2M fine with just the second core(affinity - cpu1). I tried 4M and it failed after the 10th test or something, sucks. next i tried prime95 on just this core. after 5th iteration, pc just hang. so it's either the core is bad or the memory controller is bad. correct me if i'm wrong, but does x2 share memory controller or each core has its own memory controller? it's interesting that mine is an E6 revision, but it's not a stellar overclocker though. i might have to lower it down to 2.7Ghz and try prime95 to see if i can burn in.

[uniacid]

I have a question for anyone with an X2, after overclocking in bios and loading up in windows, I'm using clockgen inside of windows to overclock higher but when I run sisoft sandra it seems to not update the cpu speed and the results (as I go higher results should be higher) although cpu-z displays the updated speed and fsb, I'm wondering if clockgen isn't overclocking correctly within windows, is anyone else experiencing this?

[don_vercetti]

when i use clockgen, it crashes. So i'm sticking to using BIOS, period.

[Flanker]

Why The Hell Isn't This Stickied?!

[chainhoe]

3800+ @ 2700mhz, 1.520v

[All_Star]

I love my X2

X2 3800+
0529 SPMW
Batch 0179

Running 10 x 260 @ 1,5 vcore
Load temp: 40 / 43 degrees.





http://valid.x86-secret.com/show_oc?id=37394

[Quest For Speed]

Clockgen crashes on me if I start under 250 HTT. If I set bios to 255 I have no problem using it.
So far after a extensive burn-in my X2 3800 is now stable 1.44v@2760. The voltage needed to run stable rises quickly when I push it further. The best I can do is 2920@1.67v

[All_Star]

Clockgen crashes on me if I start under 250 HTT. If I set bios to 255 I have no problem using it.
So far after a extensive burn-in my X2 3800 is now stable 1.44v@2760. The voltage needed to run stable rises quickly when I push it further. The best I can do is 2920@1.67v

What burn-in method did you use?

[alpha0ne]

I cannot change the multi in Clockgen ???

[All_Star]

New result:



------
X2 3800+ Manchester
CBBWE 0529 SPMW
Batch 0179
-----

It's not the maximum speed of the CPU, but my OCZ VX won't go any faster then 265 mhz 2-2-2-5 yet, it needs some burn in. More results coming soon.

[eclypse]

EDIT: forget it... i've figured it out.

[Quest For Speed]

What burn-in method did you use?

I ran the cpu with high voltage (1.65) using Cpu Burn-In for 24 hrs, then found the lowest voltage that I could run stable at 2.75, added 20 mhz and ran Cpu Burn -In again with error reporting disabled for 72 hours. I repeated the process again by raising the mhz. It only added 20 mhz to my best OC but I shaved some voltage when running at my 24/7 settings.

[alpha0ne]

Can you run CPU-Burn In with two instances running at the same time, confirmed by task manager showing 100% CPU usage on cores 0 & 1 ???, if not you really need to burn in BOTH cores at once and if needed, isolate the weaker core for some further errr 'treatment'

[Quest For Speed]

When I do my burn in I use two instances Cpu Burn-In. One core tops out at 2880 and the other 2920. I am still working on bringing up the weak core just a bit more.

[Ugly n Grey]

69c? that's too hot for that chip... keep it under 60c in my opinion....wow

[SmackDizzle]

41c is as high as I've seen mine get.. I was a bit surprised!