EVGA

VDroop - Explanation

Author
Sum1uNo83
CLASSIFIED Member
  • Total Posts : 2295
  • Reward points : 0
  • Joined: 4/15/2009
  • Location: Florida
  • Status: offline
  • Ribbons : 12
Thursday, March 25, 2010 3:50 PM (permalink)
I see thins topic discussed over and over but with no end result or conclusion. I understand enabled over volts the CPU and disabled undervolts it. also enabled protects the CPU from spikes. the question is weather or not vdroop is a good thing or a bad thing? How does it work exactly?
 
Example:
my cpu is overclocked at 4ghz / no turbo / with HT at 1.2v in bios. in windows at idle it shows 1.2 - ok but under load it shows 1.18v. does this mean if i set 1.18v in bios it will still be stable under load?
 
is this how the true voltage under load is found with vdroop on?
 
the reason im leaning more toward vdroop enabled is that it just seems to be better for temps and the cpu. please corect me if im wrong and advise any input on this matter that seems to not have an answer.
 
Thanks guys

                                       
                                                           
#1

3 Replies Related Threads

    chememan
    SSC Member
    • Total Posts : 816
    • Reward points : 0
    • Joined: 5/21/2009
    • Status: offline
    • Ribbons : 2
    Re:VDroop - Explanation Thursday, March 25, 2010 5:23 PM (permalink)
    On the previous generation of boards (790 series) the Vdroop simply reduced the amount of droop under load, but there was still some droop.  For example, My Q9650 needs 1.25v minimum to remain stable under load.  When the volts were allowed to droop, I had to keep the voltage at 1.3v+ to keep them above the 1.25v minimum under load.  Without Vdroop I was able to reduce the voltage to 1.275 and it never drops to 1.250 under load.

    If you set your voltage to 1.18v in the BIOS, it will likely droop to 1.16 under load.  If your chip needs 1.17v to remain stable it will flag an error or BSOD.  If it requires less, you might be ok and you were overvolting it before.

    With the X58 board they seem to have increased the compensation of the Vdroop control and some folks actually see a voltage increase under load because the board is over-compensating for the load.  I checked my particular board and the voltage does not move more than .01 from loaded to unloaded without Vdroop, but varies .05+ when allowed to droop.   

    There was a claim that a CPU would need less voltage to remain stable with Vdroop but I was not able to confirm that.  I needed the same minimum voltage (on the Q9650 and on the W3540) to remain stable regardless of the setting.  If there is a difference, it was too small to detect with my testing.



    4930K @ 4.6 X79Dark + 4x4 G-Skill 1900-C9 + 780Ti Classified + 840 Pro SSD + AX1200 + HAF 932 
     
    W3540 @ 4.4g  on H2O + EVGA E760 Classified + 3x2GB G-Skill Perfect Storm 2133 + 
    2x EVGA GTX480SC in SLI on H2O  + Intel X25-M 160 G2 + WD Velociraptor 300 + Corsair AX1200 + HAF 932
     
    3570k 4.5 on H2O + Z77FTW + GTX480SC + G.SkILL 1800 + Intel X25M + Corsair HX850 + HAF 932

    3770k 4.6 on H20 + Z77FTW + GTX480SC + G.SkILL 1800 + Intel 520 + CM1000+ HAF 932

    HEATWARE  

    #2
    d.burnette
    CLASSIFIED Member
    • Total Posts : 3508
    • Reward points : 0
    • Joined: 3/8/2007
    • Status: offline
    • Ribbons : 17
    Re:VDroop - Explanation Thursday, March 25, 2010 5:23 PM (permalink)
    Your pretty close.
    Having vdroop enabled, means your voltage will drop under load of course. Once you are stable you can view the vcore under load like LinX and see what the voltage is that it drop down to. Ideally for overclocking you want to find the sweet spot, which is just above where it failed previously, so you know you are not giving it more vcore than really is needed to remain stable.

    Ok so you have vdroop enabled at 1.2 vcore in the bios, and it drops down to 1.18 under heavy load and remains stable.
    If you disable vdroop, then what you set in the bios will be the min voltage, ( with the exception of say speedstep/cxe being enabled) and as load is added to the system, your vcore will actually increase.  Mine typically increases in the neighborhood of  around .02-.03v.
    So technically, you could probably get away with setting an even lower vcore in your bios, to where it hits around 1.2v whilst under a heavy load like running LinX or Prime.

    What this gets is lower vcore at idle, hence lower cpu temps when the system is not under load. Some fear that a voltage spike under heavy load may hurt the cpu, and is why vdroop is there in the first place. Certainly it was designed as a safeguard, however I have been on these forums since 2007 and have yet to see anyone at least on these forums have a problem with this.

    It really boils down to a matter of personal preference, some like not having vdroop, others prefer having vdroop. Back in the 680i and 780i days, many overclockers would use the pencil trick to shade between the appropriate transistors to achieve the same - with the X58, we now have it as a bios option.

    I myself prefer to disable vdroop and have the lower vcore setting at the start.

    Don 
     
     
    EVGA Z390 Dark MB | i9 9900k CPU @ 5.2 GHz all cores | EVGA RTX 3090 FTW3 Ultra | 32 GB G Skill Trident Z 3200 MHz CL14 DDR4 Ram | Corsair H150i Pro Cooler | EVGA T2 Titanium 1000w Power Supply | Samsung 970 Pro 1TB m.2 Nvme | Samsung 970 Evo 1TB m.2 Nvme | Samsung 860 Evo 1TB SATA SSD | EVGA DG 87 Case |
     
     
     
     
     
     
     
     
     
     
    #3
    chememan
    SSC Member
    • Total Posts : 816
    • Reward points : 0
    • Joined: 5/21/2009
    • Status: offline
    • Ribbons : 2
    Re:VDroop - Explanation Thursday, March 25, 2010 5:48 PM (permalink)
    The problem that some have against running "without Vdroop" is that they have the opportunity to pass 50 passes of LinX and then BSOD under no load because the Vcore was lower at idle than under load.  I have not noticed that happening to me, although that could technically happen.  I would assume that a small voltage bump would keep the minimum above the chip's minimum and resolve the problem at less voltage than if set "with Vdroop".  Another fear are the alleged voltage spikes, but I do not have the equipment necessary to measure those.

    4930K @ 4.6 X79Dark + 4x4 G-Skill 1900-C9 + 780Ti Classified + 840 Pro SSD + AX1200 + HAF 932 
     
    W3540 @ 4.4g  on H2O + EVGA E760 Classified + 3x2GB G-Skill Perfect Storm 2133 + 
    2x EVGA GTX480SC in SLI on H2O  + Intel X25-M 160 G2 + WD Velociraptor 300 + Corsair AX1200 + HAF 932
     
    3570k 4.5 on H2O + Z77FTW + GTX480SC + G.SkILL 1800 + Intel X25M + Corsair HX850 + HAF 932

    3770k 4.6 on H20 + Z77FTW + GTX480SC + G.SkILL 1800 + Intel 520 + CM1000+ HAF 932

    HEATWARE  

    #4
    Jump to: