E8400 overclocking help 4.0ghz!

X

Xenozx

2[H]4U
Joined
Jun 16, 2004
Messages
2,127
OK, I have the specs in my sig. I am running the newest EVGA bios available for my board. I want to overclock my CPU to the 4.0ghz range, and dont understand why I am having issues. I think I may be missing something, and figured I would look here for some help. I think my Motherboard may be my limiting factor, but Ive looked up other reviews on it, and for the most part, people seem to be able to overclock on them fine. Anyway

I cant seem to get past 3.85 ghz or 428mhz. (1720qhz) I can get to 3.8 easy, and can do it at 1.35v. to get to 3.85 and or 3.95 it required 1.40v. My temps are fine up to 1.4675v. 40idle/60 load. However, I dont seem to be able to get stable above 3.85ghz. I can post at 4ghz, I can post at 3.9 and 3.95 and everywhere inbetween, but I get random freezes, and weird curruptions like garbled text on the bios screen, and weird colors often including the color pink. No ammount of voltage seems to fix this. My other voltages are as follows:

FSB 1.4
SPP 1.4
RAM 2.1
VCORE 1.450

I turned off all advanced CPU feature but halt xB or what ever that virus processor level thing is. I tried running 1:1, and experience no increase in stability in comparision to being completely unlinked.

maybe I missed something here, and I am just looking for some input. Any suggestions that might help me reach that stable 4.0ghz barrier? HELP! :)
 
4GB OCZ Reaper X DDR2 1070 mhz 5-5-5-15 2.1v - 4 sticks or 2 sticks? if 4, try to use 2 sticks. btw, can you list the full bios specs?
 
bail_w said:
4GB OCZ Reaper X DDR2 1070 mhz 5-5-5-15 2.1v - 4 sticks or 2 sticks? if 4, try to use 2 sticks. btw, can you list the full bios specs?
Click to expand...

2 sticks. I heard 4 can cause overclock issues, so I spent a little extra for 2 2GB sticks.

CPU freq 3856mhz
FSB - Memory Clock Mode Unlinked
FSB (QDR) 1712
MEM (DDR) 1072
SLI Ready memory DISABLED

Multiplier 9x
PCI-E x16 1 auto 100mhz
PCI-E x16 2 auto 100 mhz
SPP <--> MCP auto 200mhz
NFORCE SPP -------> 5x
NFORCE SPP <----- 5x
Spread Spectrums DISABLED

Memory timing Settings
tCL 5
tRCD 5
tRP 5
tRAS 15
Command Per Clock (CMD) 2T

tRRD 3
tRC 22
tWR 5
tWTR 9
tREF 7.8ns

CPU Core 1.4500v
CPU FSB 1.4v
Memory 2.1v
nForce SPP 1.4v

Limit CPUID MaxVal DISABLED
CPU THERMAL CONTROL DISABLED
C1E Enhanced Halt State DISABLED
Execute Disable Bit ENABLED
Virtualization technology DISABLED
 
I heard rumors that the xfx bios is more stable then the EVGA, is it safe to flash the xfx bios on the EVGA 680i LT boards?
 
i wouldn't try it, but the new 45nm E8xxx series chips stability can be affected by CPU & NB GTL references voltages settings a lot, so i guess you dont have that in any of the bios? btw, you bios setting seems to be fine.
 
Can you adjust the NB voltage? I upped it 2 notches(don't have the readings in front of me) and it made 3.85 very stable.

I'm back down to 3.7 but will return to the quest on the weekend. ;)
 
Ill have to try that. I have it set to 1.4, I can move it up to 1.5. I idle 38ish and load dual primes at about 57 degrees at 1.4v.
 
Be careful when feeding these 45nm chips over 1.36v as that is the theoretical limit of these as stated by Intel. I've heard stories of people burning theirs up at 1.4v and even lower. These things really don't like the volts that the 65nm C2D's do. I think they were earlier batches so if you are feeling lucky, go for it. :p Can't wait to get my E3110 to try my luck with this 45nm stuff.
 
Have you tried it with just 1.3 on the NB voltage. I never had much luck going to 1.4 with mine. It never seemed to help my stability and I could go to 1800 on it at 1.3. I would lower that spp and fsb to 1.3 and lower the multi on that e8400 and try 1800 to see if the motherboard is holding you back or the cpu.
 
ill have to try that. These boards are voltage picky. I know I have tired 1.5 in the past on an overclock that works at 1.4. As soon as it was switched to 1.5 my PC wouldnt post, so I backed down to 1.4 and everything was fine. Maybe trying 1.3 could be helpful.

Yea, even when set at 1.45, it shows up in CPU-Z and Rightmark at 1.41 after Vdroop. Ive seen some reviews take these things strait to 1.5v. I hope thats not true, but it is always a good Idea to play it safe. Since I can do 3.8 at 1.36, maybe Ill just go to there, since the voltages are much safer. Ill try out the 1.3v NB/FSB/SPP first :)
 
No I havent, I have AS5 and Alumina sitting arround, is that a good idea? will it help out NB temps? Is there a way to see NB temps?
 
general rule of thumb for vanilla 680i boards - if you overclock, your CPU FSB voltage should be 1.5v. that is a categorical rule. just do it whenever your messing around with clocks higher than stock.

my experience across 4 different processors yields the same instability at varying FSB's and clocks, all from mild to [H]ard overclocks.

also, try running your ram in synchronous mode. it may be capable of 1070 DDR (535 raw clock) but it could be ripping your stability to shreds. with what you are trying to go for (and what i will be shooting for as well until X48) is 450x9, which translates to 900 on the memory speed.


i also just jumped on the 45nm thing, but am awaiting yet ANOTHER nvidia reference board RMA from BFG.

as soon as i get mine back ill follow suit and let you know what i come across.

prior to the RMA, i was running my q6600 (G0) at 1.38v, 3600MHz (450x8) so im fairly certain that specific board can do 450x9 on the wolfie. we shall see though.
 
@brinox, so far from my experience overclocking the new 45nm chips, they dont require a lot of FSB or VTT voltage, usually around 1.2v for Vtt and 1.3 or 1.4 for cpu fsb voltage to achieve 450 FSB or more.
 
Well I dont think the issue is my board.

IM typing this with my computer set to the following

487FSB 1948(QDR)

487x8 = 3.90 ghz.

Obviously, this shows the board can handle it fine. I actually booted into windows at 500FSB, but backed it down after a lockup.

According to rightmark no matter what I set the voltages to my MCP temp is 64-65 degrees. Is this higher then it should be?

Also, at 487, I tried to boot into windows w/ a CPU voltage of 1.35 and right before windows would load I get a really fast BSOD. the only way to get into windows is to up the voltage to 1.4v. This kinda tells me this CPU is power hungry, and is the limiting factor of my overclock. It seems that if I want into the 4.0ghz territory with this said chip, I would probably need to set my voltage to 1.5, which I am not really willing to do.

3900.jpg


oh, and the 2gb sticks vs 1gb stick overclocking reference. yes you are right, the actual sticks wont overclock as high as 1gb sticks, timings/mhz wise, but this should not effect the overclock on your CPU though right? Only, if your ram was the limiting factor? These 2GB sticks I have overclock better then my patriot extreme 1GB sticks. They made it up to about 1000mhz, where I have Orthos tested these 2GB sticks at 1080mhz 5-5-4-15 @ 2.1v, for over 15 hours.
 
what is Vcore VTT? I keep hearing people mention changing this in the bios to hit 4.0ghz, but I dont have this option. Anyone know what it is? Is it only on intel chipsets or something?

Also, in Core temp, theres something called

VID: 1.1125v

what is that? I hear people who said theres is higher then that, but I dont know what it means or hwo to change it.
 
You must log in or register to reply here.
Back
Top