Voltage not reduced when idle

[Speakerator]

Hi folks,

 

Pre-text: I recently upgraded my old 2500 to a 2600k (too cheap to build a new system at current ram prices :D). Since my P8Z68V Pro board as been idling along the last six years, I went ahead and overclocked the chip to 4.6GHz (103 Core x 45) @ 1.33V. It's stable so far, folding along for some time, running at 65 - 70C° with a Dark Rock 3, which seems fine for me.

 

However, here's my issue: I was expecting the core voltage to drop below 1V when idling (as also depicted by the 0.9V -something VID at the lower powerstate. However, CPU-Z reports a Core Voltage of 1.33V, even when Core Temp reports that the Multi dropped to 16 / 1.6 GHz.  I remember that there is a BIOS setting that deactivated / reactivates that feature, but I can't put my finger on it. Can somebody point me into the right direction? The chip is idling at 35C°, but even though I probably will not be using it for another 6 years, I guess there is no need to put unnecessary stress on the chip.

 

Best,

Chris

[xg32]

35 minutes ago, Speakerator said:

Hi folks,

 

Pre-text: I recently upgraded my old 2500 to a 2600k (too cheap to build a new system at current ram prices :D). Since my P8Z68V Pro board as been idling along the last six years, I went ahead and overclocked the chip to 4.6GHz (103 Core x 45) @ 1.33V. It's stable so far, folding along for some time, running at 65 - 70C° with a Dark Rock 3, which seems fine for me.

 

However, here's my issue: I was expecting the core voltage to drop below 1V when idling (as also depicted by the 0.9V -something VID at the lower powerstate. However, CPU-Z reports a Core Voltage of 1.33V, even when Core Temp reports that the Multi dropped to 16 / 1.6 GHz.  I remember that there is a BIOS setting that deactivated / reactivates that feature, but I can't put my finger on it. Can somebody point me into the right direction? The chip is idling at 35C°, but even though I probably will not be using it for another 6 years, I guess there is no need to put unnecessary stress on the chip.

 

Best,

Chris

either intel speedstep or turbo boost, and then you need to use an offset voltage to get back to the same 1.33v you are using right now, not manual.

[Speakerator]

On 31.12.2017 at 1:55 PM, xg32 said:

either intel speedstep or turbo boost, and then you need to use an offset voltage to get back to the same 1.33v you are using right now, not manual.

oh, wow, es, that did the trick. Guess this will need some more fiddling, since I currently only get it to boot with offset mode auto, which results in 1.36V. Negative offset of .03V is a BSOD during boot. Oh well, guess that's the fun of it. Anyway, thanks for the input! =)

[xg32]

3 hours ago, Speakerator said:

oh, wow, es, that did the trick. Guess this will need some more fiddling, since I currently only get it to boot with offset mode auto, which results in 1.36V. Negative offset of .03V is a BSOD during boot. Oh well, guess that's the fun of it. Anyway, thanks for the input! =)

If you get it to boot ur close, the 2600k will run as long as u get 0.9v idle, and then u can just match the clock speed with the load voltage it gives u.

 

I think my faster 2600k gave 1.28v load 4.4 with 0.9v idle and a -.07 (ya i know) offset. The slower one gave the same voltages (i'm guessing it depended on the mobo) but only went 4.0.

 

The faster cpu only need .87v idle while the slower one needed .9v.