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

3770k voltage isn't throttling anymore

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

soulwatch5

Member
Joined
Jun 16, 2010
Location
Burbank, CA
So finally got my 3770k hooked up and running and decided to start ocing. Below is the setup i'm using atm.

3770k
asrock z77 extreme6
G.Skill 1600 ram
Venomous X heatsink

So my goal was simple, 4.5ghz with as low voltage as the cpu will use and still be stable. I first set the multiplier to 45 and kept everything on auto just to see if it would work and still be stable. It does but the voltage would jump from 1.15~1.2volts on idle to 1.38 under-load and temps would hit mid to high 80s C and one core up to 92C. So i then switched voltages to manual at 1.30v with load line calibration to level 2 and turned off some features of motherboard i wasn't using. (rear 1394, esata ports, and Intel VT). So with just that much lower voltage setting (reads 1.28voltage even though its set to 1.3) temperature dropped significantly. Basically it dropped temps to low 70s and that one core close to 80c. I was was then planning on trying to lower voltage more but notice now that when the cpu throttle's down to 1.6ghz the voltage doesn't throttle down with it. It stays at 1.28. Not sure if its really that big of a deal just wondering why now did it now stop lowering voltage when it lowers clock speed?

Also read somewhere you probably could lower the voltage on the cpu llc to help lower temps just not sure how much is safe to lower.

Thanks in advance for any help u guys might provide.
 
I personally have a 3570k, but I have seen many achieve 4.5Ghz on a 3770K with lower than 1.3v. You could use this spreadsheet as a ROUGH guideline for what voltage to use.

http://www.overclock.net/t/1247869/...voltages-temps-bios-templates-inc-spreadsheet

As you see, at 4.5Ghz, there are not many people at 1.3v. Now, for the cpu voltage not lowering, I'm sure your aware that LLC as a major affect on this. While I think you should still see a drop when the CPU downclocks, the difference will not be as great when at LLC level 5 (assuming your extreme 6 is similar to my extreme 4). Also, are you adjusting the turbo voltage offset or the core voltage offset, changing the core voltage offset results in a non-fluctuating voltage. The turbo voltage offset is generally the best to use if you want to run at less voltage 24/7.

However, <1.3v 24/7 is nothing to worry about if you keep your temps low. 80C is pretty high, so hopefully a smaller voltage will be stable. Also, in general, due to current spikes, LLC should be used as little as possible (on Asrock meaning level 4-5), unless needed for stability (which is usually only at 4.9+ overclocks) The current spike almost always equate to higher temperatures overall, and I have recorded about a 4C difference between a stable voltage at LLC level 1 and LLC level 5. Hope this helps!
 
As you see, at 4.5Ghz, there are not many people at 1.3v. Now, for the cpu voltage not lowering, I'm sure your aware that LLC as a major affect on this. While I think you should still see a drop when the CPU downclocks, the difference will not be as great when at LLC level 5 (assuming your extreme 6 is similar to my extreme 4). Also, are you adjusting the turbo voltage offset or the core voltage offset, changing the core voltage offset results in a non-fluctuating voltage. The turbo voltage offset is generally the best to use if you want to run at less voltage 24/7.

Nice post. :welcome: to OCF. :)

The latter part of this quoted comment is more relevant to his question than the former.

LLC will get you less fluctuation of voltage under load, as the voltage normally fluctuates as load does.

However what he is asking about is actually the difference between a static voltage setting and a dynamic voltage setting. When he had everything at auto, he was using a dynamic voltage setting. At lower load levels, the CPU would enter a different C state, which means it would drop the the multiplier and the voltage to power save. Once he manually set the voltage, the CPU is still dropping the multiplier, but it is now using a static voltage.

I'm not familiar enough with the BIOS on that ASRock board to tell him exactly what settings to change in order to allow dynamic voltage at different load states, but that is the core of what his questions ask, and what you are addressing in the later part I quoted. It sounds like you might be on the right track talking about turbo voltage offset, but I'm not familiar enough with the bios to say for sure.
 
Thanks for the info guys. Been kind of busy and haven't been able to mess with everything just yet but I'll do so when I get home tonight.

Also something I notice when setting the voltage it gave me two option to set it by offset or a fixed value. I think this might be the culprit. If I set it to offset it give me values of ±.10 volts but I have no idea what I would set it to for the target voltage but when set to fixed it give me values from 1.10volts up to 1.7. I set it to fixed so I could actually see what I was setting it too.

From what I remember off the top of my head I changed the voltage from specificly the cpu not the turbo cpu volts. I'll check that when I can.

Again thanks for the help.
 
Offset voltage will add volts to VID. So if your chip has a VID of 1.1, an offset voltage setting of +.2 should result in a static 1.3V approximately. It should function the same as setting a fixed value.

That is in my experience, on non-Asrock boards.
 
Back