Does Gigabyte GA-G1975X have the voltage drop problem when overclocking?

[Manoj]

Asus P5WD2-E has the voltage drop problem when overclocking, how about the Gigabyte GA-G1975X?

[vapb400]

yep vr-zone did an article on it
http://www.vr-zone.com/?i=3187

[IluvIntel]

Which mobo does'nt when O/C ?

Answer = probably none !

[Manoj]

Which mobo does'nt when O/C ?

Answer = probably none !


I mean is the problem as bad as the Asus P5WD2-E?

or is the Gigabyte board better?

[Liquid3D]

Correct me if I'm wrong, but as i understand this, as Multi-phase motherboard CPU-power circuits increase in the number of stages (now 8 on mobo's supporting Pentium-D) chips known as "multi-phase buck converter chip and PWM controller" are utilized. For example the Intersil ISL6556B;

"The ISL6316 controls microprocessor core voltage regulation by driving up to four synchronous-rectified buck channels in parallel.
Multiphase buck convertor architecture uses interleaved timing to multiply channel ripple frequency and reduce input and output ripple currents.
Lower ripple results in fewer components, reducing overall component cost, power dissipation and board area.
The device provides very accurate over-current protection, phase-to-phase current balance and loadline accuracy through differential current sensing, as well as very precise voltage regulation through differential voltage sensing.
Differential current sensing (measuring the current across the current sense element) ensures the multiphase convertor eliminates any current measurement inaccuracy due to PCB layout.
The ISL6316 features adjustable precision reference voltage offset to enable precise loadline programming.
The offset pin allows accurate voltage offset settings that are independent of voltage identification (VID) setting.
The device senses current by measuring the voltage across the on-resistance of the lower MOSFETs or direct current resistance (DCR) of the output inductor.
The ISL6316 is paired with Intersil's ISL6612/12A/13/13A/14/14A and ISL6605/08 FET drivers.
Dynamic VID technology allows seamless VID changes.
Buck controllers such as this are able to drive multiple synchronous-rectified buck-converter channels in parallel. The advantages of a multi-phase power delivery architecture lies in the fact that they utilize interleaved timing in order to multiply the ripple frequency, and lower the input and output ripple currents. By lowering ripple currents, the developer can improve power throughput and efficiency, reduce heat dissipation, and utilize a smaller area on the board for implementation"

I believe this should revive the argument CLAIMING the actual motivation behind multi-phase power circuitry was a cost saving measure as smaller low quality capacitors are used. Adding insult to injury the fact the Pentium-D board require the 2x4 power connectors and will most likely reverse the decision behind 12V ATX Power Supply Desing Guideline (http://www.formfactors.org/developer/specs/ATX12V%20PSDG2.01.pdf) limiting 12v rails to 18A. It's been rumored those days of 45A 12V rails may be a cure for the common Vdroop? What curdles my milk, spices my chilly and squeezes lemon juice in my eye is the fact were using 8-phase power circitry and experiencing the most basic of core current supply issues. I wonder if we should take a page from the original Socket-478 Prescott fiasco? The answers are there somewhere, what's frustrating is we have to modify motherboards approachng the 300USD price range to accomodate processors built on a .065-micron process?

[Tony]

The issue is the CPU's are drawing 20A plus on the 12v line that supplies them, many PSU's just can not cope with this.

Intel made an AMP greedy monster out of their dual cores, the 955x is real bad, needing 23 to 26A at high clocks.

that will be the cause of the droop I bet..if had the cpu and boards i will confirm it real quick

[Manoj]

I am using the Asus P5wD2-E motherboard for my Pentium D 955XE. I can only overclock stable up to 4.51Ghz for benchmarks and gaming.

Any higher and it will not be stable. I can get into windows at 4.71Ghz but it is not stable.

I have noticed that the Voltage drops alot and also the CPU temp gets warmer at higher CPU clocks.

I was wondering if I should change to the Gigabyte 975x motherboard but according to you guys it also has the same Voltage drop problem.

[IluvIntel]

Correct me if I'm wrong, but as i understand this, as Multi-phase motherboard CPU-power circuits increase in the number of stages (now 8 on mobo's supporting Pentium-D) chips known as "multi-phase buck converter chip and PWM controller" are utilized.

The Asus P5WD2-E only has 4-phase cpu power circuitry, just like the Gigabyte GA-G1975X. As far as I know the Workstation version of 975X Asus mobo has this 8-phase circuitry.


what's frustrating is we have to modify motherboards approachng the 300USD price range to accomodate processors built on a .065-micron process?

You had to do that on 90nm processors as well, I think the mobo vendors have designed the O/C boards to be stable to a certain point and beyond that, they just don't want to spend the money and put higher quaility components. Really, when you think about it, it is pretty cheap to solder in a VR to stabilize the droop - but where do you solder it ? varies from mobo to mobo. :)

[JoeBar]

The Asus P5WD2-E only has 4-phase cpu power circuitry, just like the Gigabyte GA-G1975X. As far as I know the Workstation version of 975X Asus mobo has this 8-phase circuitry.

That's the case. Only P5WDG2-WS has 8 phase power.