World's most pathetic E6600?

[Zinthar]

I'm running a week 44G E6600 (I've heard some awful results with 44G, but also some promising ones) on a Bad Axe 2.

Went ahead and disabled all the features that inhibit OC'ing (such as VT, C1, watchdog, etc etc) and gave it a go.

I'm topping out right now when using 1.55V at around 350FSB, even when using the 1333 strap. My memory is Crucial Ballistix DDR2-800, and it's underclocked at the moment since I'm running it 1:1. Bad Axe 2's are known to be able to easily hit well above 400 Mhz FSB, so I think the CPU is the culprit here.

To note: I'm running Vista 32-bit and all of my failures when running Orthos instantly result in a blue screen of death. No error reported by Orthos ever -- just the BSOD.

I can post up to about 3.4 Ghz or so, but nothing past that. Am I doing something wrong? I manually switch to the 1333 strap when I try anything above 350 FSB, I've made minor rasies to NB & SB voltages, and disabled everything I could think of. Am I missing anything or is this just an awful 44G?

Running 3.25 Ghz now @ 1.5625V. With my tuniq tower I'm getting idle temps of 35 degrees with max load in orthos beta of about 49.

 

[secretanchitman]

damn...that blows. how about running that chip in a different mobo?

 

[caboob]

I feel your pain.
I have a week 47 E6600 that does about the same as well. Tried it on an Asus P5WDG2 WS Pro (975X) and the Asus Commando and get about 3.05GHz.

 

[lopri]

What's your BIOS version #? I remember a discussion some time ago regarding BadAxe 2's BIOS. It's possible that your CPU is bad, but do try out different BIOS if you haven't already. It's not necessarily the case that a newer BIOS is better than older ones.

 

[MarcVenice]

what about trying it on XP ?

 

[jhurst]

I had a 47G that wouldn't go above 3.0ghz (333 x 9), on Asus P5B-Deluxe. The things are hit and miss. I "upgraded" to an 40A but still only does about 3.2 (400 x 8) @ 1.4v, and to get to 3.4 I have to get to 1.55v, which I am not really comfortable running it at.

 

[Shimmishim]

2333 is probably the best bios or the BX2 and the newer ones are fine. i personally liked 2507 the best.

 

[f4phantom2500]

imo u should sell the chip to someone who wants an e6600 but doesnt overclock (assuming you can determine that the cpu is the culprit, or if you are convinced), cut your losses, and buy a new one that will overclock well...or be happy with it.

 

[aclim]

check my sig, im not too ecstatic with my 44g either, but im happy with what I got considering its a new batch. I kinda expected to have this problem

 

[Zap]

Overclocking = YMMV

Careful, or you guys will have fanboys come in here and accuse you of doing something wrong, or of using the wrong parts (HSF, mobo, RAM, PSU).

 

[f4phantom2500]

Originally posted by: Zap
Overclocking = YMMV

Careful, or you guys will have fanboys come in here and accuse you of doing something wrong, or of using the wrong parts (HSF, mobo, RAM, PSU).

true, but everyone who really understands overclocking should know that every processor is different. Think about it, assuming the average oc is 3.4ghz (for the sake of argument), the ratio of chips that hit 3.6ghz to those that hit only 3.2ghz is 1:1, assuming a normal bell curve distribution (which, i think, can be assumed in this case, judging by the results i've read online).

 

[Atty]

How do I tell my week? :\

 

[The Sauce]

I had an F that did roughly the same. I ditched it.

 

[lopri]

I personally think that the batches and weeks can't really be any indication for overclocking with Intel chips. With Opterons, we could kind of see the patterns, but remember that Intel makes 10 times more chips than AMD. With so many chips popping out of the production line, it's really a luck of draw.

 

[petercintn]

I have a 45G I've been playing with about a month, have one of Petra's w/c kits on it with the Apogee GTX and dual rad only cooling the cpu. I can get 3.45g Orthos 10hr stable but am having to push 1.58v in bios and speedfan says it's only seeing 1.53v. If I try to push this up to 3.5 or greater I get the BSOD.

Over at Tom's Hardware they have a chart and a good article on c2d's temps. But it says that my t-junction should be 15c less than the two core temps. I'm finding that the t-junction (or CPU reading in speedfan) is about 10c higher than the two cores. During Orthos stressing my temps run anywhere from 45c to 54c with 48c or so seeming to be average. Can anyone clue me in as to why my temps are bassackwards?

To Snatchface-The Unbeliever series is one of the most incredible reads I ever enjoyed. The sunbane was absolute genius. Stephen R. Donaldson also has a Sci-fi series, 'Into the Gap' or some such that is also excellent.


Rig Specs:
Week 45 E6600 @ 3447mhz
4g of Ocz 1066 @ 1073mhz
EVGA 680i A1
EVGA 8800GTX
2 Maxtor Diamond Max 21 Sata 320g in RAID 0
1 Maxtor IDE 160g
1 ADS USB Box w/WD 250g
1 Sony CD/DVD Burner
Swiftech Apogee GTX
Swiftech Dual Rad
Laing DDC-2 Pump w/Petra'sTech DDCT-01s Delrin Top

 

[The-Noid]

Originally posted by: Zinthar
I'm running a week 44G E6600 (I've heard some awful results with 44G, but also some promising ones) on a Bad Axe 2.

Went ahead and disabled all the features that inhibit OC'ing (such as VT, C1, watchdog, etc etc) and gave it a go.

I'm topping out right now when using 1.55V at around 350FSB, even when using the 1333 strap. My memory is Crucial Ballistix DDR2-800, and it's underclocked at the moment since I'm running it 1:1. Bad Axe 2's are known to be able to easily hit well above 400 Mhz FSB, so I think the CPU is the culprit here.

To note: I'm running Vista 32-bit and all of my failures when running Orthos instantly result in a blue screen of death. No error reported by Orthos ever -- just the BSOD.

I can post up to about 3.4 Ghz or so, but nothing past that. Am I doing something wrong? I manually switch to the 1333 strap when I try anything above 350 FSB, I've made minor rasies to NB & SB voltages, and disabled everything I could think of. Am I missing anything or is this just an awful 44G?

Running 3.25 Ghz now @ 1.5625V. With my tuniq tower I'm getting idle temps of 35 degrees with max load in orthos beta of about 49.

BADAXE reaches high FSB when you up the NB volts. It not uncommon for people to need 1.70v+ to reach fsbs over 370 or so depending on how much cache the chip has. You will definitely need to use the 1333 strap. 1.5625v on the Badaxe gives you about 1.46v under load. Your chip may be bad but it is not as bad as you are saying. Remember also your memory needs 2.2v to run at its rated timings. What voltage do you have on the memory?

Easy way for you to diagnose the problem of if it is the chip of the board. Try to post with 8x400 on the cpu. It is still 3.2 ghz so you will only need the same amount of voltage but you will have orthos errors much faster.

Also some bios on the BadAXE2 are just terrible, which one are you running?

Every person who has had a problem and not able to even get 3.2 ghz stable has always found that with a few tweaks everything worked fine...

 

[Zinthar]

Originally posted by: Yoxxy
BADAXE reaches high FSB when you up the NB volts. It not uncommon for people to need 1.70v+ to reach fsbs over 370 or so depending on how much cache the chip has. You will definitely need to use the 1333 strap. 1.5625v on the Badaxe gives you about 1.46v under load. Your chip may be bad but it is not as bad as you are saying. Remember also your memory needs 2.2v to run at its rated timings. What voltage do you have on the memory?

Easy way for you to diagnose the problem of if it is the chip of the board. Try to post with 8x400 on the cpu. It is still 3.2 ghz so you will only need the same amount of voltage but you will have orthos errors much faster.

Also some bios on the BadAXE2 are just terrible, which one are you running?

Every person who has had a problem and not able to even get 3.2 ghz stable has always found that with a few tweaks everything worked fine...

I hope you're right, but I'm skeptical because so far I've tried everything I know to test where my limiting factor is.

1) I've been running my memory 1:1 (setting mem frequency to double the reference frequency, ie 266-533 & 333-667 for 1066 & 1333 straps, respectively). I'm running 4-4-4-12 @ 2.2V, as per the memory's specs.

I've also tried running on a 5:4 divider (266-667) with 4-4-4-12 @ 2.2V and encountered no issues when I had the FSB at 350 Mhz (I was giving the CPU VCore 1.55V to reach this), and then running memory stress tests in Orthos for hours. I also ran another set of Crucial Ballistix in the machine and this did not affect my overclocking.

2) In order to test different BX2 BIOS's I first got my machine to be stable with the BIOS I began with (2692, the most recent one). My settings:

FSB: 350 Mhz
VCore: 1.55V
FSB/NB: 1.30V
ICH: 1.575V
Ref. Frequency: 266
Mem. Frequency: 533
4-4-4-12 @ 2.2V
Orthos Stable for 14+ hours before manually stopped

I then tried BIOS's 2333 & 2507, first booting into Vista with the above settings (all successfully), and then attempting to to drop VCore while keeping FSB the same. I dropped to 1.5125V and got a BSOD while running Orthos within an hour. After a few other settings changes, I determined that I couldn't find a noticeable improvement with either BIOS, and so I returned to 2692.

3) Unfortunately, I don't seem to be able to alter my multiplier with this board. The BIOS doesn't let me highlight that option. If there's a way to do this, I'd love to go ahead and try a lower multiplier with 400+ FSB to test the BX2.

4) So far with everything I've tried, the only thing helping seems to be raising VCore. This is only my 3rd overclock, but I've done a lot of research and I think I've done everything correctly.

Currently I'm running 356 FSB for 3.2Ghz with 1.5625V Vcore, 1.30V NB, 1.575V ICH, & running the RAM in a 5:4 divider 4-4-4-12 2.2V. I'm Orthos stable 12+ hours at these settings and I'm satisfied with this... so long as the 1.5625V to the CPU isn't going to damage it or degrade my overclock within the next 2 years or so.

With that voltage on my Tuniq Tower it idles in the mid-30's and maxes out in the mid-50's. So even overclocked it runs cooler than my E6400 B batch ran at its stock clock/Vcore with a Zalman 9500 (which idled in the low-40's, but had a curved IHS). But can the chip handle the voltages I'm feeding it, or is that relatively unimportant compared to how well it's being cooled?

 

[The-Noid]

FSB: 350 Mhz
VCore: 1.55V
FSB/NB: 1.30V
ICH: 1.575V

NB and ICH are backwards I think.

NB is a 130nm chip for 975x.

It takes around 1.65v to reach 370+ fsb on i975x.

 

[Zinthar]

Originally posted by: Yoxxy
FSB: 350 Mhz
VCore: 1.55V
FSB/NB: 1.30V
ICH: 1.575V

NB and ICH are backwards I think.

NB is a 130nm chip for 975x.

It takes around 1.65v to reach 370+ fsb on i975x.

Yeah, I may have them backwards, I'm actually listing them based off of my memory at the moment.

That's interesting though, as I haven't gone past 1.6v on the fsb yet. I'll put it at (or a little above) 1.65v with 1.575 to the vCore and try 370 again.

 

[The-Noid]

1.65v for NB or MCH. BX2 runs best around 1.45-1.50v vFSB.

 

[Zinthar]

Originally posted by: Yoxxy
1.65v for NB or MCH. BX2 runs best around 1.45-1.50v vFSB.

Thank you.

I'll use those settings and report back.

 

[Xvys]

Maybe try upping the juice. If I ever get a BSOD or freeze-up, it's always the ram. Too fast, too tight or not enough juice.

 

[Zinthar]

Originally posted by: Xvys
Maybe try upping the juice. If I ever get a BSOD or freeze-up, it's always the ram. Too fast, too tight or not enough juice.

That's what I had thought as well from previous experience, but that doesn't seem to be the case here. I was running underclocked at full voltage -- Crucial Ballistix D9's, rated for 4-4-4-12 @ DDR2-800 @ 2.2v. I've been running them around 700 Mhz... they should be fine. Used a different 2 sticks as well and got the same results.

I ran them 1:1 in both the motherboard's 1066 & 1333 straps.

I tested yesterday giving much more juice to the NB (1.45v). It did nothing for my overclock -- can POST at 370 FSB, but Vista is unstable there almost immediately.

I went with the following as my final settings:
Bad Axe 2 BIOS 2692
FSB: 356 Mhz (x9 for 3.2 Ghz)
vCore: 1.550v
NB: 1.325v
MCH: 1.60v
Ref. Frequency: 266 Mhz
RAM Frequency: 667 Mhz (for a 5:4 divider)
RAM: 890 Mhz, 4-4-4-12 @ 2.2V

With these settings I'm running 12 hours+ Orthos stable. 3.4 Ghz would have been nice, but this 44G chip can't make it there. Assuming the 1.55v Vcore doesn't degrade my overclock in the next year or two, this should carry me through until something appealing with 8 cores appears.

Any thoughts on whether 1.55v is too much and will damage the chip. Most of the time the chip is idling with core temps in the upper 30's with this voltage. Does the voltage really matter if I can keep the heat down on my components? And also, if anyone has any other thoughts on things I could try that might be causing poor performance, I'm welcome to suggestions.

 

[RamIt]

Always wondered why people think a chip is a dud if it overclocks higher than the highest chip offered by the manufacturer.

 

[TankGuys]

I'm not surprised, really... unfortunately, it seems like there's a lot of bad chips past week 40 or so.