• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

i7-980x (P6X58D) @ 4.46 GHz - Need feedback to break 4.5!

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

00011011

Registered
Joined
Mar 8, 2010
Location
Roselle, IL
Trying to break 4.5 barrier. Here are my settings below. I could really use some advice from some of you veterans in order to squeeze out the last few MHz that I need...

BCLK: 144
CPU Ratio: 31 (4462 MHz)
DRAM Freq: 2020 MHz
UCLK Freq: 3032 MHz
CPU Volt: 1.39375
QPI/DRAM Volt: 1.40

Load-Line Calibration: Enabled
C1E Support: Disabled
Intel C-State: Disabled
Virtualization: Disabled
Intel HT Tech: Disabled (for now)
SpeedStep: Disabled (for now)

stresstest2.jpg


pcis1c50per.jpg
 
Have you tried a different multi (lower) with a higher bclk? What are your other major voltages set at?

EDIT: Crap, you have pics I cant see b/c I am at work...
 
I have mine at 21x215 (4.5GHz) with 1.4v set in the bios. If I go any lower on the voltage in the bios, I can't pass prime for any respectable length of time. From several other users, this seems to be about the norm regardless of multi and blck settings - I apparently have one of the more sought after batches (3003Bxxx) but I don't think it's going to matter that much.

My QPI is actually set to 1.3v - I really don't think you need yours set quite so high.
 
Have you tried a different multi (lower) with a higher bclk? What are your other major voltages set at?

What other voltages are significant, I've only touched CPU Voltage and QPI/DRAM...

Also If the CPU frequency comes out the same (for example 4.4 GHz), is 200x22 and different than 220x20 as far as the CPU is concerned? Becuase I thought that with an unlocked muliplier there wasn't really a need to try to maximize the BCLK for isolation purposes or whatever...

I have mine at 21x215 (4.5GHz) with 1.4v set in the bios. If I go any lower on the voltage in the bios, I can't pass prime for any respectable length of time. From several other users, this seems to be about the norm regardless of multi and blck settings - I apparently have one of the more sought after batches (3003Bxxx) but I don't think it's going to matter that much.

My QPI is actually set to 1.3v - I really don't think you need yours set quite so high.

I will give those setting a try. Have you passed LinX with those settings? I also have a 3003B batch, there are discussed as being high performers.

Regarding the QPI, when I was using the XMP profile for my 2000 MHz RAM, the BIOS set it to 1.6 automatically. Strangely though when I configure everything manually, I am Prime95 stable and able to pass 20 rounds of LinX at 4.46 GHz with only 1.4 volts QPI...
 
Regardless of multi and blck settings, it took 1.4v for me to maintain 4.5GHz. No, I haven't used LinX, but I have used Prime with small FFts - it passed for four hours, I was happy enough with that.

I've really only touched the CPU and QPI voltages myself - other than setting the vDimm to 1.65.
 
My QPI is actually set to 1.3v - I really don't think you need yours set quite so high.

Does anyone else think that 1.4v for QPI/DRAM is too high for RAM at 2000 MHz?

Another question: I have a top filling Res, and it seems that I have to add a couple ml of coolant every day so that it stays abolutely full and not become airated. Does coolant naturally burn off or evaporate or something in water cooled systems, or do i have a poor seal some where and it may be evaporating? I know there are no "leaks"...

(Also, today I just passed the LinX 20 Run Stable test @ 4.5 GHz :)
 
Last edited:
Back