Can't get e8400 past 3.06 Ghz.. why?

gousst

Distinguished
Feb 9, 2009
18
0
18,520
So after replacing an e5200@3.4 with an e8400 stepping c0 thinking I could get it at least to the same speed at stock voltages, I can't get past 340x9 (3.06) without failing a prime95 test after an hour. It wouldn't even boot at 378x9 (3.4), whereas I could boot into windows with the e5200@3.6 with stock voltage and cooler (same motherboard).
Since I have a crappy old asrock mobo (n73v-s) with the 610i chipset and no vcore adjustments, I thought that it simply didn't like anything above the rated 1333 fsb, but i just tried the stock speed as 375x8 (3 Ghz) = 1500 fsb and it didn't get any errors running the large FFT prime95 test after 30 mins (still might not be stable, but I was surprised it booted). I have the ram downclocked to get the1:1 ratio so I don't think thats the issue here.

I suspect I just got a crappy chip that needs more voltage, but what do you guys think the issue is? should I revert back to the e5200@3.4 🙁?

Edit: Oh I totally forgot, temps with the e8400 seem too high with everything at stock; 63 ºC cpu and both cores at 80 ºC each running prime95 for an hour. At idle they are the same as the e5200, cpu @ 33 ºC and each core at 43 ºC.. I'm so confused!
 
Solution
Yeah, you're bridge on the mobo may be a limiting factor but wow, great job considering! I'm not sure there's much else you can do. You're already jimmy rigging it as it is... ha! Nice post! Temps look great btw!
You basically answered your own question. Yes, we don't all win the intel chip lottery and sometimes we do in fact get a loser chip. Sounds like you got a dud in terms of oc'ing. Sorry to say, sad but true. I suspect unfortunately that you are correct that you got a crappy chip. Sorry!
 
Well I guess that's it, I got extremely lucky 5 years ago with the e5200 (36% OC) and this e8400 just does a 2% overclock.. I have a hard time believing it (luckily it was almost free) but thanks for the reassurance.

I've looked and there is a way to increase voltage via a "pad mod": http://www.overclockers.com/forums/showthread.php?t=586596 but I can't find info concerning the e8x00 series, is the contact arrangement the same across all the 45nm wolfdale processors? I got the temps down 10 degrees by reapplying thermal compound but I'm willing to risk burning it anyway
 
Interesting. If you aren't afraid of burning it then by all means...
Good job on getting another 10c down with the reseat. Yeah, it's no fun getting a dud. 36% OC on the other chip alternatively is quite nice! 2% is unusually weak but it is what it is. Keep us posted on what you get out of her!
 
After going through the e8x00 datasheet I successfully increased the default vid of 1.225V to 1.325V by bridging two pads with aluminum foil and glue instead of the conductive ink people suggest, and it was no easy, but easier than I first thought I gotta say!

2zjidjp.jpg


I don't know why it shows less voltage than it's supposed to, but I can only get to 3.33Ghz stable before i get freezes (no BSODs this time). Do i need more vcore or is it my cheapo board that can't keep up?
No matter what, if I move the fsb to anything beyond 371 my system will eventually freeze even when it passes prime95 tests after several hours. Before upping the vcore, i would get instant prime95 fails and BSODs instead so I'm not sure what to do now. Is it safe to increase the pcie frequency to anything above 100Mhz?
 
Yeah, you're bridge on the mobo may be a limiting factor but wow, great job considering! I'm not sure there's much else you can do. You're already jimmy rigging it as it is... ha! Nice post! Temps look great btw!
 
Solution
Haha thanks, I've installed an intake and outtake fan (molex) and temps are fine under load now. I just tried multiplier and bus combinations and it looks like the board is what's failing and no matter how high I pump the chipset/VTT/RAM voltages, it just won't give up more fsb Mhz.
Well at least it reached the e8600's stock speed :pt1cable:

34ngr5i.jpg