Overclocking my new e8200..need advices...

[CH@NO]

so here's the problem...

I have a EP35-DS3L from gigabyte, an e8200 with a pair of OCZ Platinum Rev.2 at 800mhz all powered up with a 550 w cooler master extreme power psu.

I've seen some threads that this cpu can reach over the 4 ghz AND with lower voltage increase, I get stucked within the 3.7ghz....the system boot at 3.8 but very unestable. The thing is, surely I'm doing something wrong 'cause I need to raised the Vcore up to 1.45 IN the MIT section of the BIOS, other weird thing is that in the PC HEALTH STATUS the Vcore shows 1.42 (aprox) instead, also CPU-Z and Everest, I'm using Vista right now.

This mobo and CPU are kinda popular and I have a descent PSU, so the RAM....then, any ideas to OC this CPU any higher???

Thanx.

EDIT:Added some screenshots of my BIOS at 3.8ghz

 

[theonetruewill]

The reason you get 1.42v in PC Health is that you have vdroop. This is common and don't particularly worry about it. Get Realtemp (not coretemp- if you want me to explain why send me a pm) to monitor your CPU temps and hit harder on the voltage. I think you should take your RAM out of the equation and drop it's speed right down so you can see the limit of your CPU. Then once you have found that (do not try more than 1.525v) ramp the RAM speed up.

 

[CH@NO]

The reason you get 1.42v in PC Health is that you have vdroop. This is common and don't particularly worry about it. Get Realtemp (not coretemp) to monitor your CPU temps and hit harder on the voltage. You may need to increase FSB voltage (Northbridge).

ok, I'm already aware of the misreading temps of Everest and others and I'm using Realtemp, the temps are pretty close to the BIOS's temp.

I raised the FSB core to +.2 of +.3.....if I raised to the max could I damage the mobo???

Oh, and what are ICH or whatever It say below the FSB voltage control???

 

[theonetruewill]

ok, I'm already aware of the misreading temps of Everest and others and I'm using Realtemp, the temps are pretty close to the BIOS's temp.

I raised the FSB core to +.2 of +.3.....if I raised to the max could I damage the mobo???

Oh, and what are ICH or whatever It say below the FSB voltage control???

MCH voltage is the Northbridge. I fear FSB voltage may be the FSB termination voltage - don't go any higher on that. However, I am not sure on the latter - why the hell can't Gigabyte name the damn things properly.

EDIT: FSB overvoltage is the Southbridge. Don't go anymore than +0.1v, there simply is no need.

Also if you see anything about Command rate in the RAM settings, change it to 2T from 1T.

 

[Jarman]

E8200 chips hit the limit at about 3.8 due to the multiplier. U dont wanna go above 1.4v with them tbh....they are prone to degredation or even failure.

 

[CH@NO]

E8200 chips hit the limit at about 3.8 due to the multiplier. U dont wanna go above 1.4v with them tbh....they are prone to degredation or even failure.

Yeah, Intel says not go above 1.45v.....but how much It's lifespawn is degraded???....from 5 to 2 years???, sometimes It doesn't matter, many of us upgrade our systems within the 2 years, at least I do that.

 

[CH@NO]

MCH voltage is the Northbridge. I fear FSB voltage may be the FSB termination voltage - don't go any higher on that. However, I am not sure on the latter - why the hell can't Gigabyte name the damn things properly.

EDIT: FSB overvoltage is the Southbridge. Don't go anymore than +0.1v, there simply is no need.

Also if you see anything about Command rate in the RAM settings, change it to 2T from 1T.

thanx mate, 'till now the system remains stable at 3.8 with 1.45v, the idle temps are 33-39C core0-1 respectively.

Just i wanna ask you if I put in the MIT section of the BIOS on the vcore 1.5v, CPU-Z shows 1.45v...you say i'ts 'cause the "vdrop" thing, Is it possible that the CPU will recieve the 1.5v or that "vdrop" will always halve the Vcore put it to 1.45v????

hope I'll be clear.

Thanx again

 

[theonetruewill]

- Issue solved for the moment via PM messages -

 

[dadi_oh]

Yeah, Intel says not go above 1.45v.....but how much It's lifespawn is degraded???....from 5 to 2 years???, sometimes It doesn't matter, many of us upgrade our systems within the 2 years, at least I do that.

Well given that the Wolfdales have not been out that long and there are reports of dead and degrading chips already we may be talking months, not years...

Depends on your risk aversion. It seems that the rules have changed with the 45nm parts. It used to be that thermals killed you before voltage but now it is not so certain. It seems that you can kill your Wolfie with volts before temps become an issue.

BTW. I have both an E8200 and an E8400. Since I am uncertain about the voltage risks until there is more data I am keeping my Vcore at 1.365V max (the max in the Intel spec I believe). At that voltage My E8400 runs at 3.92GHz while the E8200 only hits 3.6GHz (Prime95 stable). This seems to be the limit of the CPU, not a FSB limit since my moetherboards are capable of >500MHz FSB.

It is entirely possible (and plausible) that the E8200's are binned as E8200's rather than E8400's specifically because they do not meet the same stability as a E8400. So my expectation is that E8200's will not OC like E8400's regardless of FSB limitation.

My $0.02 worth anyways...

 

[CH@NO]

Well given that the Wolfdales have not been out that long and there are reports of dead and degrading chips already we may be talking months, not years...

Depends on your risk aversion. It seems that the rules have changed with the 45nm parts. It used to be that thermals killed you before voltage but now it is not so certain. It seems that you can kill your Wolfie with volts before temps become an issue.

BTW. I have both an E8200 and an E8400. Since I am uncertain about the voltage risks until there is more data I am keeping my Vcore at 1.365V max (the max in the Intel spec I believe). At that voltage My E8400 runs at 3.92GHz while the E8200 only hits 3.6GHz (Prime95 stable). This seems to be the limit of the CPU, not a FSB limit since my moetherboards are capable of >500MHz FSB.

It is entirely possible (and plausible) that the E8200's are binned as E8200's rather than E8400's specifically because they do not meet the same stability as a E8400. So my expectation is that E8200's will not OC like E8400's regardless of FSB limitation.

My $0.02 worth anyways...

thanx, I'll dig for more info, If the things that you say are true (pretty much sense to me), I'll down de Vcore.

 

[CH@NO]

Well given that the Wolfdales have not been out that long and there are reports of dead and degrading chips already we may be talking months, not years...

Depends on your risk aversion. It seems that the rules have changed with the 45nm parts. It used to be that thermals killed you before voltage but now it is not so certain. It seems that you can kill your Wolfie with volts before temps become an issue.

BTW. I have both an E8200 and an E8400. Since I am uncertain about the voltage risks until there is more data I am keeping my Vcore at 1.365V max (the max in the Intel spec I believe). At that voltage My E8400 runs at 3.92GHz while the E8200 only hits 3.6GHz (Prime95 stable). This seems to be the limit of the CPU, not a FSB limit since my moetherboards are capable of >500MHz FSB.

It is entirely possible (and plausible) that the E8200's are binned as E8200's rather than E8400's specifically because they do not meet the same stability as a E8400. So my expectation is that E8200's will not OC like E8400's regardless of FSB limitation.

My $0.02 worth anyways...

you're correct.....why push the Vcore so high only for 100mhz boost???.....I try 1.37v at 3.7 and the system seems stable, now with that voltaje i'm under safe range, it's just try and see if I could lower the Vcore a little more.

Thanx for the info, It was helpful.

 

[theonetruewill]

you're correct.....why push the Vcore so high only for 100mhz boost???.....I try 1.37v at 3.7 and the system seems stable, now with that voltaje i'm under safe range, it's just try and see if I could lower the Vcore a little more.

Thanx for the info, It was helpful.

I can hit 3.424MHz on 1.50v stable but I use 1.45v 24/7 @ 3.334GHz, because I don't wanna shove more volts on for very little gain. I have approached this thread as a max OC thread so you don't have to run these settings 24/7. Run it at what you feel comfortable in the end- but don't fear whacking a fair bit through it for max OC benching

 

[dadi_oh]

you're correct.....why push the Vcore so high only for 100mhz boost???.....I try 1.37v at 3.7 and the system seems stable, now with that voltaje i'm under safe range, it's just try and see if I could lower the Vcore a little more.

Thanx for the info, It was helpful.

I think we are on the same page. Lot's of volts for 100 or 200MHz seems like a low payback IMHO. Not that I am all that risk averse. I watercool and voltage mod video cards just for fun. In the case of the Wolfdales however, I am sitting on the fence until we have a better sense of how they react to voltages in the 1.4V+ range.