I aint sure whats going on during IBT...first round testing error but when it comes to second round, success. Any clue

dat means not stable...........

try wit OCCT, choose to only stress the processor.

It doesnt seem to get right at this point...yesterday shows error on "Maximum" in IBT and OCCT Large Data Set keeps prompting error on core #3 in less than 10sec.

Today, i start IBT and OCCT...both r fine and indicating no error until OCCT reported error on core#1 and thts it no more errors after a few stress rounds.
Can anyone figure it out

it shud b smooth sailing 24/7.
but stress testing failure can also cause by other h/w such as mobo & ram.
try & put ur processor on another system & try stress testing again.

processor is design to run @ full load 24/7.
so it shud passes those stress test. 5 loops of IBT is consider as ngth, it shud b smooth sailing.

if indeed the processor got problem, u can argue to intel bout dis.
say something like u ran folding 24/7 & processor often causing crashes.

Run OCCT small data set for cpu stress test...
OCCT large data set failed = incorrect voltage or timings for Ram & vNB...
Set Cpu Margin Enhancement to Compatible or Performance.. (if any)

Since Core 3 is giving premature failures on IBT, I would suggest to send it for a RMA claim

To kill your processor, continue using IBT and have it looped non stop for the next 2 days, that would literally kill your processor.

The bolded.. that seem not right. R u sure?
Did u use the old power supply that involved in the lightning tragedy? I suggest that u try other PSU. It could be your PSU.

This post has been edited by campsol2k: Aug 4 2009, 03:46 PM

bro bryan, although i know tht proc are literally the most non effected hardware in terms of nature disasater unless of course some high generated voltage did overburn it. But if u say kill it off, hw would i be able to explain to them and is this gonna give alot of hustle work ? sorry,,just to new abt killing off a proc using IBT

---

Added on August 4, 2009, 4:27 pmbro tech, if lets say i sent the proc to Intel and inform them like what u told me. Will they take the proc for their own testing to confirm or just RMA it and wait for a new one ?

---

Added on August 4, 2009, 4:30 pmPsu is fine, did a testing on other rig. Its just tht i forgot which is which since i didnt OC for quite sometime after the RMA mobo came in and failed to OC. Any info, do ask me. Thx for ur concern

This post has been edited by his: Aug 4 2009, 04:30 PM

Hmm... btw killing your proc using IBT stressing isnt a great idea... you could kill your RAM and motherboard in the process...

Ok... put everything in stock and stress your PSU using OCCT for 15 minutes... report back to me...
What's your stock CPU VID?

This post has been edited by campsol2k: Aug 4 2009, 04:41 PM

its all in stock settings, change none at all...current CPU VID is at 1.25v.

Psu stressing is a success without failure/error

This post has been edited by his: Aug 4 2009, 05:19 PM

Ok. Seems that the PSU is fine. So the question is which is the culprit? its either motherboard, RAM or processor?

right now, lower the multiplier to 8 and run an IBT stress test at the last stable-unstable FSB with your previous other stable settings(mem clock/mem timings/voltages). I want to make sure its either the proc or the motherboard.

What's your PL setting when overclocked at 360FSB?

This post has been edited by campsol2k: Aug 4 2009, 05:48 PM

u 1st hav to answers some question & fill up some forms.
call dem & dey will email to u.

after dat intel will send delivery guy to ur location to pick the processor up.
now the processor is wit dem, obviously dey will test it.

u just hav to wait now.
intel warranty is usually quite fast, nvr > 2 weeks.

Hey guys

it seems i manage to OC back but with different kind of settings.

FSB : 360
Multi : 8
Fsb Strap : 400
Dram Freq : Auto
Dram Timing : 5-5-5-15

Settings for current from Vcore onwards are all AUTO.

Hope it helps giving u all a bit clues

This post has been edited by his: Aug 4 2009, 11:00 PM

hmm...

If u don't mind do some more experiments...
Try to increase the FSB become 3.0GHz (370X8) and run the stability test... Keep increasing the FSB until u achieve unstable FSB. U may test it with different multi.. let say 7.

If u can't push your processor above 3.0GHz, that's mean something is wrong with the processor.... I'd say it's time to RMA the proc...
Just for your reference, my Q9550 C1(default VID is 1.15V) doesnt need any core voltage increment and stable when clock at 3.0GHz. I believe your Q9550 E0 can do abit better... i mean when your processor is in perfect condition.

This post has been edited by campsol2k: Aug 5 2009, 12:49 AM

Thx for the reply...i will try it but if i up the voltage on dram (which i did tried earlier) & also increasing a slight increment on ram freq, it just runs 100% fan full speed with no post

---

Added on August 5, 2009, 11:06 pmHey guys,

I need to ask for ur experiences. I currently did some OC testing and below r the results

FSB : 400
Multiplexer : 8.5
fsb strap : 400
Dram freq : auto
Vcore : 1.31
PLL : 1.54
Vtt : 1.24
Nb volt : 1.22
Dram voltage : Auto

Result : No post and fan keeps running 100%

Second testing with slight changes on above settings

Fsb : 400
Multiplexer : 8.5
Fsb Strap : 400
Dram freq : auto
Vcore : auto
pll : auto
vtt : auto
nb volt : auto
Dram volt : auto

Result : Able to run and insert windows without a problem

Whats causing the problem?

This post has been edited by his: Aug 5 2009, 11:06 PM

as for my settings for q9550 @ 3.4ghz..(just reference ah. dun follow exactly cuz diff mobo diff settings for stability)
400mhz x 8.5 = 3.40ghz
cpu vcore 1.25
den the other Voltages, i din set auto.. i key in the defaults values instead..

den mch i bump up 0.6v

as for vdimm(vram).. 2.04v.. timing 5-5-5-15..

i think q9550 dun need alot changes to make it to 3.4ghz.. im pretty sure at stock voltage 1.25v.. it can run. i saw ppl using even lower vcore n still stable..
i think important is ur mobo mch and vdimm.. check ur sticker label for ram wats the voltages.. normally if auto, the bios set the ram voltage pretty low..
i think in bios ram voltage they usually set to 1.8v for auto.. its way below wat is recommended by the ram manufacturer.. like i say earlier.. check ram sticker for default voltage to run..

This post has been edited by jcyl2: Aug 6 2009, 12:24 PM

Hey guys

New update from m havoc wreaking PC
Below r the testings

FSB : 400
Multiplexer : 8.5, 8, 7
FSB Strap : 400
Dram Freq : Auto
Dram timing : 5-5-5-15
Vcore : 1.306
gtt : 0.370 / 0.410
pll : 1.52
vtt : 1.24
dram volt : 2.00
nb : 1.22

It runs perfectly but if i upped the gtt to 0.400/0.410, it wont run anymore

Second setting :
FSB : 410
Multiplexer : 8.5, 8, 7
FSB Strap : 400
Dram Freq : Auto
Dram timing : 5-5-5-15
Vcore : 1.326
gtt : 0.370 / 0.410
pll : 1.52
vtt : 1.24
dram volt : 2.00
nb : 1.22

It doesnt run and even when i upped the gtt to 0.400/0.410, it still wont post

It will only post when i stay below 410fsb and gtt at 0.370/0.410

This post has been edited by his: Aug 7 2009, 01:48 PM

ur OCing quad core .
y not reduce d multiplier until like 4-5 ?
and put d vCore to stock vID and push FSB bit by bit .
im saying like 333x7 = 2331
400x6 = 2400

u get wat i mean ?
keep d speed at almost identical as stock and find d stable spot for it .

raising speed u can add multiplier and vCore only so wont be hard .

bro why don't you up your vtt and nb voltage slightly and slowly to see if it does the trick. coz from what I can remember, those 2 looks kinda low to me. I suggest up with the nb first. use the bold'ed setting, but only play with the nb (first) and vtt

This post has been edited by metsatsu: Aug 8 2009, 11:02 PM