Overclock.net › Forums › Intel › Intel CPUs › Kaby Lake Overclocking Guide [With Statistics]
New Posts  All Forums:Forum Nav:

Kaby Lake Overclocking Guide [With Statistics] - Page 216

post #2151 of 2759
@Darkwizzie A few days ago we were arguing that not only high overclocks should be submitted and included in the chart. So here's mine before my Phanteks PH-TC14PE cooler arrives in the next days.

It's not a high overclock, it's more to show the limits of the Arctic Freezer i32 cooler (and similar) when overclocking. It's a decent cooler, similar to Cooler Master Hyper 212 Evo in both price and performance. Silent and would be perfect match for an i3, i5 or slightly overclocked i7.

Did only a 2 hours test because I did not have more time and it's just to show what to expect with a cheaper cooler while staying below 80 °C. Temperatures averaged 76 °C with only a couple spikes to 85 °C, nothing to worry about.

So here's my submission, purely for statistics.

Username: dirtyred
CPU Model: i7-7700K
Base Clock: 100
Core Multiplier: 47
Core Frequency: 4700
Cache Frequency: 4200
Vcore in UEFI: 1.24 (Adaptive: Turbo 1.22 V, Offset 0.02 V)
Vcore: 1.216 V (mostly stable, occasional spikes to 1.232 V, never below)
FCLK: 1000
Cooling Solution: Arctic Freezer i32, not delidded
Stability Test: RealBench 2 hours (I know it's on the edge of stability, crashes if Vcore is set 0.02 - 0.03 V lower)

Batch Number: L640G413 Malaysia
Ram Speed: 3000 16-16-16-28-2T (tightest while stable on 1.2 V)
Ram Voltage: 1.2 V (to keep CPU temps as low as possible)
VCCIO: 1 V
VCCSA: 1.05 V
Motherboard: Z270 Asus Prime Z270-K (BIOS 0610)
LLC Setting: LLC Level 5
Misc Comments: Just a quick benchmark to test how far can you overclock with this cooler while staying below 80 °C. AVX offset 2 (45, could force 46 but temps are a bit too high for my taste).

    
CPUMotherboardGraphicsRAM
i7 7700 K @ 4.7 GHz ASUS Prime Z270-K Gigabyte GeForce GTX 1060 WINDFORCE OC 6G  G.Skill Ripjaws V 16 GB 
Hard DriveCoolingOSMonitor
Samsung 940 EVO 256 GB Phanteks PH-TC14PE Windows 10 Pro Dell U2312HM 
PowerCaseMouse
Seasonic M12II-620 EVO Fractal Design Define C Razer Taipan 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
i7 7700 K @ 4.7 GHz ASUS Prime Z270-K Gigabyte GeForce GTX 1060 WINDFORCE OC 6G  G.Skill Ripjaws V 16 GB 
Hard DriveCoolingOSMonitor
Samsung 940 EVO 256 GB Phanteks PH-TC14PE Windows 10 Pro Dell U2312HM 
PowerCaseMouse
Seasonic M12II-620 EVO Fractal Design Define C Razer Taipan 
  hide details  
Reply
post #2152 of 2759
Quote:
Originally Posted by cxmachi View Post

Yes, but I'm still stuck wondering what exactly is going on with my voltages? Isn't setting a voltage for adaptive voltage supposed to ignore the VID and set your stated voltage in the UEFI?

Again, my end goal here is to have 1.25V Vcore when I'm at full load, regardless if it's AVX or non-AVX. Currently, with my UEFI settings, I'm getting 1.28V during AVX loads and 1.25V during non-AVX loads like gaming. Bumping down LLC from 5 to 4, I can get 1.264V on full load on AVX, and 1.232V on non-AVX/gaming (which I'm not confident will be stable since 1.25V is what I tested stability on)

Is there no way to do this when using adaptive voltage? Is this a common issue/normal?
Adaptive works off VID with a scaling that you set with maximum Vcore.

Example DVID, Auto LLC, Prime95 Max 1.332v, RealBench Max 1.284v, Web browsing Max 1.320v idle ~ 0.200v-0.800v, idle power saving fetures off 1.296v
post #2153 of 2759
Quote:
Originally Posted by wingman99 View Post

Adaptive works off VID with a scaling that you set with maximum Vcore.

Example DVID, Auto LLC, Prime95 Max 1.332v, RealBench Max 1.284v, Web browsing Max 1.320v idle ~ 0.200v-0.800v, idle power saving fetures off 1.296v

So maximum Vcore in UEFi isn't really maximum then? It just scales off of that value? Since I'm clearly going over that value in AVX loads.
post #2154 of 2759
Quote:
Originally Posted by dmo580 View Post

And I'm telling you I'm using Adaptive voltage with my max set at 1.35V. My benchmark was Prime 95 28.10, so definitely there's heavy AVX loads going on.

Here's a screenshot of my Vcore (1.344) versus VID (1.359). I suppose one could argue it's "close enough" within the 0.016 resolution that we have? However if I look in my notes I have recorded load Vcore from 1.312 all the way up to 1.344 as I tried to obtain stability at 4.9 GHz. So the fact that my load Vcore in Prime 95 can go that low suggests my chip was able to go below VID (assuming it's been 1.359 all the time) at turbo.



Perhaps @cxmachi perhaps you can measure your VID and Vcore both under load in BF1 and Realbench.

I just did this but I forgot to do graphs. I also used BF4 in this case so as to match yours.

Basically, the results are as follows:

30 minutes of BF4: 1.248V Vcore, 1.255V VID. Both averages.
1 run of x264: 1.28V Vcore, 1.299V VID. Both averages also.

It's close like yours but my Vcore is able to go below VID here. I don't really know what to make of this. Like I said, I'm getting conflicting answers from all of my questions on here.
post #2155 of 2759
Quote:
Originally Posted by cxmachi View Post

I just did this but I forgot to do graphs. I also used BF4 in this case so as to match yours.

Basically, the results are as follows:

30 minutes of BF4: 1.248V Vcore, 1.255V VID. Both averages.
1 run of x264: 1.28V Vcore, 1.299V VID. Both averages also.

It's close like yours but my Vcore is able to go below VID here. I don't really know what to make of this. Like I said, I'm getting conflicting answers from all of my questions on here.

True but I wouldn't make too much of Vcore going under VID given that we only have 0.016V resolution on our boards. Not to mention the measured Vcore isn't always correct. But it's interesting to note taht you basically are seeing the VID dictate your voltage, which is why X264 is forcing a higher VID and therefore your VCore is higher.

I'm going to try some more testing but I remember intially trying like 1.32vcore. It's not stable enough for Prime 95 but I think I can get a few minutes out of it. I'll record the Vcore and VID again to see if there's a difference.
post #2156 of 2759
Quote:
Originally Posted by dmo580 View Post

True but I wouldn't make too much of Vcore going under VID given that we only have 0.016V resolution on our boards. Not to mention the measured Vcore isn't always correct. But it's interesting to note taht you basically are seeing the VID dictate your voltage, which is why X264 is forcing a higher VID and therefore your VCore is higher.

I'm going to try some more testing but I remember intially trying like 1.32vcore. It's not stable enough for Prime 95 but I think I can get a few minutes out of it. I'll record the Vcore and VID again to see if there's a difference.

Yeah, that is the weird part since almost everything I see online basically says that at adaptive voltage mode, and at the turbo ratio, it's supposed to ignore VID and looks at the voltage you set in the UEFI. Which isn't what I'm seeing here. Thanks btw, you can also try the x264 stress test linked in the OP, I think you should be able to run 1 loop of that atleast.
post #2157 of 2759
Quote:
Originally Posted by cxmachi View Post

So maximum Vcore in UEFi isn't really maximum then? It just scales off of that value? Since I'm clearly going over that value in AVX loads.
Depends how the motherboard manufacture sets up the Adaptive VID voltage ramping calibration to the actual Vcore. All you need to know is VID scales with Vcore when using Adaptive.


My Gigabyte maximum DVID/Adaptive Vcore is what I set exactly in Bios. My rig CPU Vcore 1.260 + Dynamic DVID +0.075 = 1.335v

Gigabyte DVID, Auto LLC, Prime95 v28.10 Max 1.332v, RealBench Max 1.284v, Web browsing Max 1.320v idle ~ 0.200v-0.800v, idle power saving features off 1.296v

Quote:
Originally Posted by cxmachi View Post

Yeah, that is the weird part since almost everything I see online basically says that at adaptive voltage mode, and at the turbo ratio, it's supposed to ignore VID and looks at the voltage you set in the UEFI. Which isn't what I'm seeing here. Thanks btw, you can also try the x264 stress test linked in the OP, I think you should be able to run 1 loop of that atleast.

With ASUS they have 3 options to use 1# fixed Vcore, 2# Adaptive uses VID and your voltage setting or AUTO stock, 3# then there is offset no VID use, it ramps off of the CPU frequency.

Serial Voltage Identification "SVID": A few generations back, Intel introduced serial voltage identification (SVID) which is a protocol the CPU uses to communicate with the voltage regulator. The power control unit inside the CPU uses SVID to communicate with the PWM controller that controls the voltage regulator. This allows the CPU to pick its optimum voltage depending on current conditions (temperature, frequency, load, etc.). You can actually use a combination of SVID and LLC to get an optimal VCore instead of manually setting it. If you start your system without making any changes, your VID (which some refer to as the stock voltage) might be 1.25v, but if you lower your CPU multiplier and restart, you will find your VID has dropped automatically. The reverse happens if you increase your clock and do not set any VCore. Intel's latest CPUs are able to pick their own voltage, and this comes into play if you want to utilize "offset" / "adaptive" voltage. The good news is that if you come from Haswell, you should look forward to a CPU that has the same or better durability.

Read more: http://www.tweaktown.com/guides/7481/tweaktowns-ultimate-intel-skylake-overclocking-guide/index5.html
post #2158 of 2759
hey guys, new around here, just wanted to see if theres anything I can do to improve my OC

Username: Zornyan
CPU Model: 7700k
Base Clock: Bclk.100
Core Multiplier: 50
Core Frequency:5ghz
Cache Frequency: stock
Vcore in UEFI: 1.375v .
Vcore: 1.360.
FCLK: (cant see this on hwmonitor?_
Cooling Solution: not delidded, kraken x62 with two corsair ml140 fans
Stability Test: realbench 1 hour  

Ram speed: 3200mhz (need to check timings)
Motherboard: asus maximus z270 code
LLC Setting: 5
Misc Comments: AVX offset -2




this is what my system is reporting running realbench, voltage at 1.375 in bios but seems to sit a 1.36 whilst stress testing, LLC is already at level 5 so no idea whats causing this? anything under 1.375 in bios seems to be unstable, crashing after approx. 30-45 mins realbench.

I have had one crash at the end of a 1hour realbench, but I'm sure this is to do with my voltage not sitting at the correct level? temps hover in the mid 70s most of the time so I don't think its that.

also, can anyone explain what the LLC/RING voltage that hwmonitor is reporting at 1.4v is? and why my VID is reporting lower than my vcore?

think with a few tweaks I should be able to get it more stable, anything like XTU it passes fine for hours, so If I can figure out why the voltage is dropping already in realbench that should keep me stable.
post #2159 of 2759
What's the most effective way to determine the chips overclock potential?
post #2160 of 2759
Quote:
Originally Posted by Xperimental View Post

What's the most effective way to determine the chips overclock potential?
Try running cinebench r15 to get an idea of how good a chip is.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Intel CPUs
Overclock.net › Forums › Intel › Intel CPUs › Kaby Lake Overclocking Guide [With Statistics]