TPU's Core i7 Overclocking and Feedback

[Fitseries3]

 

[Fitseries3]

 

[Solaris17]

damn this board is wicked finiky...lol i havent figured out how to OC at all it refuses...and do all i7's shut down when they restart? doesnt matter if i do it in windows or ctrl-ALT-DEL it shuts down for about 3 sec (as in no power to the rig) and starts back up again...doesnt seem to affect anything...just odd. anyone have some settings to try for me?

 

[Fitseries3]

bad board it sounds like.

 

[Chicken Patty]

damn this board is wicked finiky...lol i havent figured out how to OC at all it refuses...and do all i7's shut down when they restart? doesnt matter if i do it in windows or ctrl-ALT-DEL it shuts down for about 3 sec (as in no power to the rig) and starts back up again...doesnt seem to affect anything...just odd. anyone have some settings to try for me?

it is a normal thing for the EVGA x58 board to shut down on restart. Mine does it completely normal.



Solaris, YGPM.

 

[Solaris17]

it is a normal thing for the EVGA x58 board to shut down on restart. Mine does it completely normal.



Solaris, YGPM.

wew thnx i was scared i was like omg wtf?

 

[Chicken Patty]

wew thnx i was scared i was like omg wtf?

yeah I crapped my pants too! But im used to it now

 

[kid41212003]

There is a guide to overclock EVGA X58 on evga forum. It might help you.

My gigabyte shut down completely for half a second before it power on, that's when I restart the computer.

And when it power up, it will power up for a second, then off for a half (no lights, nothing), and then power on again. I guess it's a power-supply-test, before post.

Because, my old computer would boot up, even though I didn't connect the vga power cable. But with this board, it would make a really noisy beep BEFORE post if I don't connect the power cable for vgas.

 

[Solaris17]

so this thing refuses to OC i think i got almost 3Ghz out of her and that guide...ya doesnt really work CP you have some settings for me? and it wont run my ram either at its rated speed...their 1333mhz sticks and if i oc and try to bring my ram upto speed it wont boot ...funny thing is even if i dont try to oc and i try bringing my ram upto speed it still wont boot im stuck at 1066mhz their 1333mhz sticks 9-9-9-24 and when i enter 9-9-9-24 2T at the correct voltage the system refuses to post?

 

[Binge]

incompatibility w/your ram Sol. Get different sticks.

 

[Solaris17]

incompatibility w/your ram Sol. Get different sticks.

that would totally blow i got these for x-mas and instead of triple channel kits im just going to get 2 more of these. its cheaper./...but that doesnt explain why i can oc the proc even if leave the ram at 1066.

 

[Binge]

you can't get triple channel without triple channel sticks. It will not register. As for your OC it makes perfect sense. The ram is incompatible. If it is unstable as is then making the rest of the system unstable would greatly reduce the chances of successful benching etc.

 

[Solaris17]

you can't get triple channel without triple channel sticks. It will not register. As for your OC it makes perfect sense. The ram is incompatible. If it is unstable as is then making the rest of the system unstable would greatly reduce the chances of successful benching etc.

what im saying is my ram is stable at the 1066 settings it is at now but if i try to OC the rig it will not post

 

[Binge]

And I'm saying "Roger Roger, 10 4 good buddy," I read that loud and clear. 1066 may not be completely stable with those sticks. It could be 99.9999% stable but when you OC the stability exponentially drops even with the same clocks. I know I'm being quick to point the finger at the ram and you don't really want that to be the case, but that seems to be the root of your issue.

 

[Chicken Patty]

Solaris, those are the settings I used bro, perfectly stable at 4.2 GHz. My ram I can set it at 1333 and it boots fine, definitely could be a RAM incompatibility isssue!

 

[Chicken Patty]

Sol, try to leave everything on auto, leave the ram timings on auto too! just set the BCLK to 170 and thats it, vcore auto and everything. see if it boots. That was my air setup and it worked flawlessly for 3.4 GHz

 

[kid41212003]

Ok, after a week of resting myself and my system (got sick of crashing and bluescreen), I'm back on OC.

I think I'm the only one with GIGABYTE board, and air cooler here, incase you missed my posts on this thread:

With HT enable, and TB off at 3.8GHz, and vcore 1.34 in bios, the system passed prime95 for 8 hours+.

With HT disabled, and TB on at ~4.1GHz, and vcore 1.35 in bios, the system passed prime95 for 12 hours+.

These 2 above are the best one I have found.

The surprise part is, the cpu is hotter under 3.8GHz with HT enable than 4.1GHz with HT disable (5-7C different).
At 1.375 vcore with HT enable , the cpu temp would reach 70C, and 80C at 1.4v
And with the same voltage with HT disabled, the cpu temp is lower ~10C

After countless testing, and reading from different forums, my conclusion for my system are:

My system can not handle vcore 1.381vcore+ with HT enabled, too hot. It's possible with 1.4vcore and HT disabled.

Main objective: searching for the highest frequency with vcore under 1.4v

Secondary objectives:
1st. 4.2GHz HT enable under aircooling: Stopped.

2nd. 4.2GHz HT disable under aircooling: In progress.

When I got my cashback on February, I will move to WC, and I will re-do my 1st project.

What I'm doing right now:

I'm testing my system stability under this setting: baseclock 200x20, 4GHz, HT disable, TB off, , 1.35vcore, already passed prime95 for 3 hours.

If the system pass Prime95 for 8 hours+, I will move to the next step with TB on, and that's 4.2GHz

Until Microsoft fix the issue with Threads management, I will not run my system with HT enable, I hope the upcoming Windows 7 will be different.

 

[Binge]

Welcome back to the ring! Don't get KOed

 

[Fitseries3]

i will be reporting my findings soon along with a proper how to on i7 OCing.

i have figured something out though. its alot easier to get it to run after you know what it wants from you

i benched 4.16ghz last night at 1.3625v no problems. in fact... i benched for about 3 hours with no restarts at all.

 

[Chicken Patty]

Ok, after a week of resting myself and my system (got sick of crashing and bluescreen), I'm back on OC.

I think I'm the only one with GIGABYTE board, and air cooler here, incase you missed my posts on this thread:

With HT enable, and TB off at 3.8GHz, and vcore 1.34 in bios, the system passed prime95 for 8 hours+.

With HT disabled, and TB on at ~4.1GHz, and vcore 1.35 in bios, the system passed prime95 for 12 hours+.

These 2 above are the best one I have found.

The surprise part is, the cpu is hotter under 3.8GHz with HT enable than 4.1GHz with HT disable (5-7C different).
At 1.375 vcore with HT enable , the cpu temp would reach 70C, and 80C at 1.4v
And with the same voltage with HT disabled, the cpu temp is lower ~10C

After countless testing, and reading from different forums, my conclusion for my system are:

My system can not handle vcore 1.381vcore+ with HT enabled, too hot. It's possible with 1.4vcore and HT disabled.

Main objective: searching for the highest frequency with vcore under 1.4v

Secondary objectives:
1st. 4.2GHz HT enable under aircooling: Stopped.

2nd. 4.2GHz HT disable under aircooling: In progress.

When I got my cashback on February, I will move to WC, and I will re-do my 1st project.

What I'm doing right now:

I'm testing my system stability under this setting: baseclock 200x20, 4GHz, HT disable, TB off, , 1.35vcore, already passed prime95 for 3 hours.

If the system pass Prime95 for 8 hours+, I will move to the next step with TB on, and that's 4.2GHz

Until Microsoft fix the issue with Threads management, I will not run my system with HT enable, I hope the upcoming Windows 7 will be different.

Thanks a lot for the input Kid, I will give this a shot too! Right now im in the process of purchasing another rad or if I cant find one soon, I will mount my dual fan rad and see how that does against the heat monster i7. Right now 4.2 GHz HT on and 1.411v temps get almost as high as yours. The single fan rad just can't take it.

 

[kid41212003]

I forgot to add, enable EIST on bios help stability.

According to my motherboard, EIST mean, the system will handle the VID for best performance and thermal.

It did not say anything about Speed Step, and I didn't see my CPU lower the speed. Weird?

 

[Solaris17]

all right guys ill give it a try

 

[Binge]

HT on, all power saving features on, skewing and spectrum off, ram 1600 8-8-8-24 1.65v, the unclock trick Fitseries has been talking about, stock NB SB and PLL voltage, and lastly 1.38vCore in bios. WHQL nVidia drivers 180.42 not as good of a cpu score from when I use 181.00, but there are no artifacts. With a PhysX update my cpu score would be back up 2000 points.

I know someone else on these forums should have a good chip.

 

[Chicken Patty]

well, at 4.2 GHz with HT enabled I am perfectly stable, however any higher and it tkaes some more vcore which I cant afford due to temps. So i decided to fiddle a bit with HT off.

I am now 4.3 GHz without increasing voltages over my 4.2 GHz ht enabled setup, stable so far, going to bench in a bit, maybe in the morning, i'm sleepy as heck.


here is what I have for now

 

[LiveOrDie]

What VCore should be enough to run at 920 at 3.2Ghz and what vtt? people tell me the stock VCore is 1.10v is that right?

im running mine on VCore at 1.15v and 200+ vtt? at 3.2Ghz