Overclock.net banner

Skylake Overclocking Guide [With Statistics]

2M views 11K replies 851 participants last post by  djsolidsnake86 
#1 ·
Skylake Overclocking Guide

:devil-smi
:devil:
:devil-smi


Haswell Overclocking Guide [With Statistics]
Skylake Overclocking Guide [With Statistics]

Welcome to the Skylake Overclocking Thread. Because a lot of people found my Haswell Overclocking Guide to be a useful source of information, I have decided to try my hand at a Skylake guide as well. For those new to Overlock.net, know that you can open an image in a new tab to view it in its actual size.

I am not responsible if your CPU blows up, your motherboard blows up, your computer blows up, your house blows up, or if your life blows up due to my guide.
:buttkick:


My recommendations on what parameters are "safe" are my opinions. I do not have many chips I can subject to high voltages and loads to figure out what is truly safe or unsafe.
What is Skylake?

Skylake is the codename of the microarchitecture launched at August 2015 to succeed Broadwell. The unlocked desktop parts were the first to come out. They are the i5-6600k and the i7-6700k.

How are the 6600k and the 6700k different?

It's the same pattern as in previous generations.

6600k:

Base Clock: 3.5GHz

Max Turbo: 3.9GHz

L3 Cache: 6MB

Hyperthreading: No

Box Cost: $243

6700k:

Base Clock: 4.0GHz

Max Turbo: 4.2GHz

L3 Cache: 8MB

Hyperthreading: Yes

Box Cost: $350

In most cases the L3 cache difference is immaterial. If you want hyperthreading, the choice is obvious but bear in mind it makes the CPU hotter in testing. Will the i7 part overclock better? That remains to be seen. However, if my Haswell Overclock Chart is any indication, the difference will likely be small. If that 100MHz matters that much to you, you should've bought a binned chip in the first place.

What's new with Skylake?
  • No stock heatsink.
  • z170 chipset.
  • Socket 1151. Past coolers compatible with Sandy Bridge, Haswell, etc still work.
  • Neither the CPU nor the socket are backwards compatible.
  • Removal of Fully Integrated Voltage Regulator, or FIVR introduced at Haswell due to issues at low power applications.
  • Lack of FIVR may mean lower temperatures (although that's just one factor among many) and could mean better response under extremely low temperatures.
  • According to Sin from OCN and Raja from Asus, changes in architecture requires less robust power-handling components. Because of this, boards tend to have a lower power phase count.
  • DDR4 is standard. DDR3 and DDR3L support very rare, and no motherboard supports both DDR3 and DDR4. Still dual channel memory controller.
  • It is believed that the Integrated Memory Controller in Skylake is very strong, allowing very high speeds without compromises.
  • A smaller die, thinner PCB, and smaller process than Haswell/Devil's Canyon.
  • Generally acceptable core voltage has increased.
  • Expect overclocks to be similar to that of DC. (In this guide I will refer Haswell as Haswell and Haswell Refresh as DC.)
  • Base clock can be changed in a very fine-grain way. Instead of "straps" jumping for 100 to 125 MHz etc, now you can increase it 1 MHz at a time. It's also possible to deviate from the base clock more without instability. Changing the base clock here has no effect on the DMI. However, it still affects ram and PCIE. In total, this makes BLCK changes a more useful thing to try.
  • In a similar spirit, ram overclocking is less granular. What used to be 200/266 MHz steps are now 100/133 MHz steps.
  • Power usage is similar enough to DC/Haswell. Temperatures similar to DC.
  • Graphics updated to HD 530.

What about the "x20 PCIE lanes" stuff I've been hearing about?

  • The Direct Memory Interface has been upgraded to version 3. This bumps up the spec to 8.0 gigatransfers per second.
  • There are "20 lanes of PCIE 3.0". 16 are direct lanes to the CPU. If you want more you need to go to the enthusiast platform.
  • The remaining 4 lanes are handled by the DMI. In the past you had 2 GB/s cap, now you have 4 GB/s cap. There's still overhead, so in reality you'll get maybe a maximum of 3.5 GB/s. You can now run a single Intel 750 Series ssd without bottlenecks. DMI PCIE lanes have a higher latency than direct CPU lanes but the difference should be minor. You can now RAID PCIE ssds through Intel Rapid Storage Technology, and it can be a mix of NVMe and ACHI drives, although that is not recommended. Procedure remains tricky.
  • The DMI upgrade also means you will be seeing a lot of new USB 3.1 ports, with a chance of seeing USB Type-C ports. Of course, if all of those are running at full blast, expect your PCIE SSD speeds to suffer a bit.
  • Usually in motherboards, the top PCIE lane is x16 PCIE 3.0 lane, second is the same, and the third full PCIE lane is for the SSD. Since Skylake CPUs only have 16 CPU PCIE lanes, putting a second GPU in means both GPUs run at x8.

How does Skylake stack up to past generations?
Refer to this lovely chart from PCPer.





(Credits: PCPerspective at www.PCPer.com)

Haswell overclocked to ~4.5GHz and DC maybe ~4.6-4.7Ghz. It's far too early to say what the average Skylake overclock will be. Clock for clock Skylake has an extra large improvement in x264, of about 28% over Snady Bridge. The chart might be using an old version of x264 though. Other sources report 35% gain. In Cinebench however, that lead shrunk to 11% over Haswell and 25% over Sandy Bridge. Broadwell isn't often discussed due to low TDP, probably low overclocking, and the emergence of Skylake.

Quick Word about Temperatures and Delidding:

Intel said that Skylake has a similar thermal solution under the Integrated Heat Spreader to that of Devil's Canyon. The temperatures are also similar to DC when I consider margin of error. However, it still benefits from delidding. Expect 10-13C improvement in temperatures. Lower temperatures can decrease the amount of voltages required to stabilize. By delidding yourself, you are risking permanent destruction of your CPU and resale value of the CPU will decrease even if the delidding went perfectly. You can pay $50 to SiliconLottery to have them delid it for you. This service is insured so that if anything does go wrong, your chip will be replaced.

There are also two types of delidding. You can either choose to take off the IHS, put on paste, and mount the cooler directly onto the die. This is called 'bare-die' and requires a more delicate touch. If the cooler is mounted too tightly you can warp the die and you're toast. If you elect to put the IHS back on, this is the safer method. Bare die isn't a large improvement over putting the IHS back on, so I recommend putting the IHS back on.

Is there any insurance for my CPU?
Yes. The Intel Protection Plan still exists. For $25-$30 you can stop worrying about overclocking leading to death of your CPU. This is on top of the warranty you get with a CPU purchase. Visit here for details. In general people found that Intel has been pretty lenient in accepting replacements.
Here are some things to think about before the instructions are read:

  • Coolermaster 212 Hyper Evo -> Noctua D14 -> x61 Kraken -> Custom Loop
  • Do you want to delid?
  • Ripjaws 4 were out during the x99 times, which is pre-Skylake. Both Ripjaws 4 and Ripjaws 5 are DDR4 modules, but it is recommended to use the latter. There have been some anecdotal reports of Ripjaws 4 requiring more voltage to be stable on the z170 platform.
  • Unlike Haswell, LLC affects core voltage like the pre-Haswell times, and adaptive voltage mode is no longer dangerous under heavy synthetic loads like Prime95.
  • On a similar note, input voltage is no longer a setting. Same goes for cache voltage.
  • C-states decrease the voltage and in turn power usage during idle. This has a marginal effect on SSD performance.
  • The recommended utility for looking at your stats is HWInfo available here.
  • Terminology Check:

    Uncore = Cache Ratio = Ring Bus (Not technically 100% true, but when people say these things that's what they mean.)

    BLCK = Base Clock

    100 (Base Clock) x 45 (Multiplier) = 4500MHz or 4.5GHz


  • Again, changing the base clock affects PCIE and ram speeds. You will have to readjust the ram setting accordingly if you change the base clock.

  • Skylake is still a new launch. Keep track of BIOS updates. Things may be improved as time passes.
  • If you Bsod, you can look at some details from the crash log. BluescreenView will pull up the information for you. If you want, you can download it here.
  • VID is the voltage the processor requests. Generally it is not a useful reading in HWinfo. Vcore when read in real-time in a tool like HWinfo is a measurement of the voltage actually given. When you put in 1.3v into core voltage in the BIOS, maybe only 1.25v is given to the cores under load. This discrepancy is called Vdroop. To counteract that you can simply raise the voltage you entered or you can use Load Line Calibration or LLC. This setting impacts the real-time Vcore reading and increases it. Voltage delivered can have very quick drops, so quickly that specialized gear is required to detect it. LLC helps counteract that.
  • Can't find Vcore reading in HWInfo? Don't know where F-clock (Fclk) is on HWinfo?
  • Whenever possible, change as few settings at a time. If you crash with core, cache, ram, and GPU overclocks enabled, it's hard to tell which overclock(s) caused the crash.
  • Increase multiplier by 1 each time.
  • The voltage required to stabilize the next multiplier increases each time.
  • Write down the settings you've tried for better organization.
Core Overclocking

Skylake comes with far more freedom to tweak than Haswell, but this also means more complexity when you are in pursuit of the absolute best settings possible.

0. Update your UEFI.
1. Manually set your cache ratio and ram to stock. Don't even use XMP profiles. Hell, turn your GPU's overclock off.
2. All Skylake CPUs so far can hit 4.4GHz. Try 4.4GHz at ~1.35v. It should work and be stable. If not, apply 1.4v. Stable? Good.
3. Just go up a multiplier. Increase voltage if you crash during stress testing with our x264 test. Remember, recommended maximum voltage is 1.45v. Never run a stress test and leave without monitoring the temperatures for the first 2 minutes.
4. Eventually you will find the highest overclock you can hit without breaking 1.45v and this overclock will pass x264 test overnight. Another option is to try Prime95 v27.9, but v28.9 is overkill and masochistic. To read more about different stress tests and to access quick download links to them (including out modified x264 test), check the "Stress Testing" spoiler.

5. Decide if you want to tweak the base clock. The next section is all about that. If you don't want to tweak the base clock, continue reading about Fclk if you want a possible 1% boost to your FPS in graphics-related stuff like gaming and skip to ram overclocking and cache overclocking. The sections with a blue title are related to base clocks, but the Fclk section applies to those who want to tweak Fclk without touching the base clock as well.

Blck/Fclk tweaking time!

Blck, or "base clock", affects multiple things including: Fclk, core frequency, cache frequency, and ram frequency. Let's look at each of these things.

Fclk (Optionally with Base Clock Changes):

Fclk is a setting that has to do with the way the GPU contacts the CPU. The default setting was supposed to be 1000MHz, but due to some complications, z170 boards end up having 800MHz as the default setting for the earlier versions of their UEFIs. The ability to adjust the Fclk was added with later BIOS updates. Fclk is a GPU-oriented setting, so CPU benchmarks won't notice a difference. The difference between various Fclk settings is relatively small. It varies depending on the configuration (especially from GPU to GPU), but for a 980ti the difference is within the margin of error. Open the spoiler here to look at Anandtech's results:



Notice how different the gains are depending on the GPU used.

800MHz was the original default, with 1000MHz supposedly being the default, but it's possible to overclock this further. From what I can gather, the performance gains of overclocking the Fclk is really quite small, even on graphical benchmarks like Unigine Valley. If the difference in performance past 1GHz Fclk doesn't even show under some graphical benchmarks, the difference is really quite small. However, if you have a GTX 770, this section is for you.

The exact settings will vary because motherboard vendors love assigning different names to the same thing. The instructions below are meant for Asus z170 Hero boards but it should be similar enough to whatever you have to make sense.

There should be a setting to adjust the Fclk directly in your BIOS, allowing you to set the Fclk to 400MHz, 800MHz, or 1000MHz.

For example, in the Asus Hero z170 UEFI, under "Tweaker's Paradise", there is an option called "FCLK Frequency for Early Power On". With all motherboards there is typically an option for 1000MHz. What this setting actually does is set the Fclk multiplier to 10, and if the base clock is 100, 100 x 10 = 1000MHz Fclk. In HWinfo, under "System Agent Clock" (refer to Overclock Preparation spoiler), it should now read 1000MHz.

So, if you choose to only overclock the Fclk through the dedicated Fclk setting, make sure your base clock makes sense. If you have it set to 1000MHz and you forget about it and you go back to changing the bclk to overclock your core clock, you won't understand why you won't POST at 170 bclk. The answer is that your Fclk has been overclocked to an insane value.

So let me state it again: Your final Fclk frequency is affected by both your Bclk and the setting you've chosen in the dedicated Fclk setting.

Fclk setting at 1000MHz (Fclk multiplier = 10)

Bclk set to 100MHz

-----------------------------

10 x 100 = 1000

Fclk is 1000MHz

Fclk setting at 800MHz (Fclk multiplier = 8)

Bclk set to 110MHz

-----------------------------

8 x 110 = 880

Fclk is at 880MHz

If you didn't pick a Fclk setting and you left it at auto in the dedicated Fclk menu, then your motherboard will try to adjust the Fclk multiplier so that you will not crash.



Fine tuning Core Clock with Base Clock:

Base clock x Multiplier = frequency in MHz

Recall that multipliers can only be whole numbers. If we only tweak the multiplier, we can only do 4.5, 4.6, 4.7GHz etc. What if I can do 4.5GHz but I cannot do 4.6GHz? Maybe I can stabilize 4.55GHz. To get 4.55GHz we have to change the base clock. The base clock can contain decimals (like 100.1MHz, etc).

So let's say we want to do just that: Apply a 4.55GHz core clock. We know our 4.5GHz overclock is stable and this is our fallback. Let's pick a number that's kind of close to 100 that multiplied by something gives us 4550. 130 base clock with 35 core multiplier will do just that.

Old OC:
100 x 45 = 4.5GHz Core clock
100 x 40 = 4.0GHz Cache clock
2133MHz Memory clock

Base clock set to 130:
130 x 45 = 5.85GHz Core clock
130 x 40 = 5.20GHz Cache clock
2773MHz Memory clock

Base clock set to 130 and multipliers adjusted accordingly:
130 x 35 = 4.550GHz Core clock
130 x 31 = 3.900GHz Cache clock
2080MHz Memory clock

If the above passes, we now have a stable 4.56GHz. Maybe we could aim for 4.57GHz now.

In the example, changing the base clock to 130 causes the core, cache, and memory clocks to be far higher than stock. (Why stock? Because if you were following directions, all of your overclocks should be set back to stock!) Since the cache and ram are now overclocked, we don't know if the new cache and ram overclocks are stable, and if the new settings crash we don't know if it's the core clock, cache clock, or memory clock causing the problem. We must adjust the multipliers to set it back to stock. 31 as the multiplier for cache gives us something close to what it was originally, so we will not crash. Setting the ram back down to close to 2133MHz gets us stable ram once again.

The higher you go from 100 base clock, the harder it is to stabilize. Generally the stability at 170 bclk and up will vary depending on the motherboard. You will sometimes fail to boot if the bclk is too high. There's usually no good reason to set the base clock above 170 though. With smart math, it should be possible to get very close to any frequency without exceeding 150 bclk. Don't forget that bclk can have decimals.

Base Clock Balancing Act:

If you want to adjust the Fclk and ram settings on top of fine tuning core and cache clocks with base clock changes, you have to ensure that:

  • Bclk that is not too far from 100 to cause instability.
  • Bclk when combined with a core multiplier, gives you the absolute highest core frequency that is stable. (Highest priority, unless using GTX 770.)
  • Bclk when combined with a Fclk multiplier (4, 8, or 10) has to result in an overclocked, but stable Fclk.
  • Bclk when combined with a right memory divider, gives you a ram frequency that is overclocked somewhat near its maximum.

If you are going to adjust the base clock but don't care about Fclk overclocking, please make sure your Fclk isn't overclocked when your base clock is changing. You can check this in HWinfo as "System Agent" clockspeed.

Ram Overclocking
Once both the core and cache ratio are set to stable and overclocked values you don't want to touch anymore, go ahead and overclock your ram. You now have more fine-grain jumps in frequencies to choose from. Don't forget that timings matter as well, and the "tighter" or the smaller the numbers are, the better. According to Asus, System Agent and VCCIO voltages can help stabilize a ram overclock, although more isn't always better. The ram itself could use some extra voltage. The default is 1.3v, let's bring that to 1.35v which is a safe amount.

Here are rough guidelines for figuring out how your ram is doing for those too lazy to benchmark:

Latency:
Ram can have lower latency or higher frequency. Generally for gaming purposes, lower latency is considered to be more important (unfortunately for you, DDR4 is generally worse in this regard than DDR3). To calculate latency, do 2000 x (Cas/Frequency). Lower is better.

Frequency:
On the other hand, a higher frequency is generally considered to be useful for video editing workloads which do sequential reads. These types of work favors higher frequency. To calculate how long these reads take we do 1000/Frequency. Lower is better.

Anandtech's Rough Ram Performance Formula:
Frequency/Cas = Performance Index
Whichever has a higher performance index is generally faster. If two sets are very close, the higher frequency kit wins.

Cache Overclocking

Cache overclocking is the easiest thing to overclock but has the least impact on performance (comparisons listed in the later sections). By now, everything should already be overclocked and stable except for the cache. Simply set the cache to the same frequency as your core and stress again to check if it's stable. If it's not stable, lower the multiplier by one and repeat. Unlike Haswell, cache can generally get to the same frequency as the core. There isn't even a cache voltage to worry about.

Final Step:
Go back and see if your overclocks still function perfectly with less voltage. How low can you go? This is just fine tuning of your voltages.

Safe Voltages (TENTATIVE):
Vcore: 1.45v/1.4v
VCCIO: 1.25v/1.2v
System Agent (SA): 1.3v/1.25v
Vdimm: 1.4v/1.35v
No Cache voltage or Input Voltage with Skylake

The first value shows voltages a pretty ballsy person can use. The voltage after the forward slash shows voltages for regular users who don't want to live on the edge. Please do not exceed these parameters unless you are a serious power-user that knows what he/she is doing and is willing to accept the risks.

Quick Word About "24/7 Stability and Safety"
24/7 stability is useful for people using their CPU 24/7. Playing video games a couple of hours a day or week is not the same as hammering your CPU at 100% load for hundreds of hours in a row. If you're really that concerned about CPU longevity you shouldn't be using Prime95 to stress. "100% load 24/7 safety" is a meaningless and vague goal people strive for.




Unlike Haswell, Prime v28 and Linpack are no longer much hotter than other tests. They are still the hottest tests around, but it's not quite as ridiculous anymore. For example, Prime v27.9 is similar in temperature to v28.7, As the settings chart notes, I detected temperature fluctuations in Linpack, IBT, and XTU stress even though the load on the CPU still read 100%.

Just like in Haswell, note how XTU stress is cooler than XTU bench, and AIDA64 varies in temperature wildly based on the settings checked. Without a way to loop the test, applications like XTU bench and Cinebench are not viable stress tests. As expected, custom x264 at 16 threads is hotter than the 4 thread setting, and using more memory for Linpack causes a hotter test. With the Haswell temperature chart I had 8gb of ram to use, and for this chart here I had 16gb.

My temperatures are lower than what most people will observe because I am not running hyperthreading and I am also delidded. My case as good airflow. Hyperthreading will cause a chip to be hotter and harder to pass stress tests.

Below is a hierarchy of stress tests, listed in order from hardest to pass to easiest to pass. For time-to-crash data, please visit the 'Miscellaneous Testing' spoiler. Anything that is counted as easy to pass or even easier are not recommended and will not be enough to be entered into the main overclocking settings chart. More details in the charting form spoiler.

Marathon-Man:

OCCT S

Linpack (Max) (From Intel's website, not from OCCT or any other place or XTU.)

P95 28.7 S

Tough:

P95 27.9

IBT (Max)

Medium:

x264 16T

ROG Realbench

Easy:

Stockfish (Chess, BMI2 version)

XTU

Aida64 (Full Suite)

Walk in the Park:

Cinebench

Firestrike

Booting into Windows

x264 is the recommended and the default go-to stress test for this thread. If you feel the need to use a hotter test that is your right, but know that your overclock may be hampered by that choice. You could forego delidding in many cases simply by switching to x264. The downside to this method is that the overclocking process will take longer because we are replacing a very stressful program and a short test duration with a less stressful one and a longer duration.

I highly recommend trying our x264 encoding test if you are looking for a test that can stress while still being pretty cool. For a peace of mind I recommend running x264 looped all night as you sleep once, and if it passes, it's stable. Run it, sleep, wake to see the test still running, pass, smile.

Angelotti and JackCY have tweaked the x264 Bench utility and turned it into a stress testing tool. You no longer need to download other programs to get it to work; just download, unzip and run. Simply put, our version of x264 test is better in every way to the original x264 benchmark. There is no reason to use the original utility. There is a readme inside to tell you what options to pick but I will also summarize it here: By default, try the 16 thread setting (yes, even if your CPU is an i5) with normal priority.

'Prime95 is not Certified for Haswell/Skylake/Insert Nonsense Here'

This was a myth that was perpetuated by some Youtubers. For Skylake it seems this has died down. But know that Prime95 will not eat your CPU and spit out the remains.

Prime95:
When you are closer to stability, Prime95 may stop with an error. This is a rounding error, meaning the crash was minor enough so that your computer itself does not crash. There is some data to suggest that Prime95 gives out rounding errors very frequently, even in overclocks considered functionally stable. With Skylake, unlike Haswell, version 28.7 is not significantly hotter than version 27.9. Still, from testing thus far, v28.7 has been shown to crash unstable overclocks faster than 27.9, so consider it a harder test.

There isn't conclusive evidence so far about which setting in Prime is the most stressful and prone to crashing unstable overclocks. It is known that smaller FFT sizes tend to cause higher temperatures. 8 is the smallest size (in K, but that's a technicality). It is unknown if using more memory causes unstable overclocks to crash faster. Here is a picture showing how to set your own FFT size:


Linpack/IBT
Linpack is a newer version of IBT. Please note that with larger ram usage settings, it will take a while for the memory to be used up and the temperature to increase. This may take 2 minutes or so.

Below among the list of stress test download links there is a link for "Linpack Package". I have taken Linpack and added Linx GUI to it. You can now use Linpack as Intel originally intended or run the GUI to easily change the test settings.

Stress Test Download Links

Custom x264 with Loop Functionality and Other Improvements v2.06

Aida64 v5.30.35

IBT v2.54

Linpack Package v1

OCCT v4.4.1

Prime95 v27.9

Prime95 v28.9

ROG Realbench v2.4

XTU v6.0.2.2

y-Cruncher 0.6.8

Latest Version of HWinfo (Monitors temps, voltages, etc.)

Memtest v6.2.0 (For testing ram overclocks.)

'I must pass all stress tests!'
So if I made a program that crashes you at stock clocks, you would feel compelled to underclock your CPU, even if that application in no way represents real-world usage? Passing "all stress tests" really means passing "all stress tests that people happen to have made". If nobody decided to make ultra-mega-Prime95, you would think your overclock is stable. That seems like a random, haphazard way to figuring out if your overclock is stable or not. Computers are built for using, and whether you crash at Prime95, what really matters is whether you crash often enough while using it normally. Forcing yourself to pass a stress test "just in case you use it to its limits" makes no sense either. No point in going down "what ifs" which have no signs of ever happening. And if it does, work it out when it does.

Run 2 different types of stressing programs, and then use your computer normally. If you crash, then it's not stable. What's stable for you might not be stable enough for me. Some people need 100% reliability because of their jobs.

Let's not get into a semantic debate about the word 'stability'. If you define stability as 'never crashes on anything, ever', then I don't care about your notion of stability. That criteria makes no sense either because the only way to be sure you are stable forever is to test your CPU forever. The world doesn't end if your CPU crashes on you. Run a stress test overnight, then go play video games to test things out. If you ever end up crashing in the heat of the moment, lower the multiplier by one and you should be perfectly stable.

Miscellaneous Settings:
It is recommended by Asus to disable 'spread spectrum' and 'CPU SVID Support' in the UEFI when overclocking. Please note that disabling SVID support disables adaptive voltage. In order to get Cstates to work correctly with an overclock, you need to have adaptive voltage mode on.
I have redone benchmarks to test the performance difference between a high and low cache frequency.



As you can see, a decrease of 100MHz in core clock has a larger impact on performance than a 1,000MHz decrease in cache frequency. Therefore, my position on cache frequency remains unchanged: It is a secondary setting that you should only overclock and worry about once everything else is done.

If you'd like to see Haswell cache frequency testing, open this spoiler.

Credits to Maxforces for the second part of the benchmarks. From my personal benchmarks, I found the drop of 0.7Ghz for the cache to be an equal performance hit of 0.05Ghz decrease in core clock and this difference shows in a very CPU reliant benchmark like chess.





And here are the most recent tests for cache frequency that I have done:



The 4.2 vs 3.4 is the cache setting. The core multiplier for this test was x45.

Testing methodoloy in this test is much more well documented by me.

Chess: Houdini 3, 9mb hash, starting position, 5 minutes.

BF3 Multiplayer: 64 player server in a closed map (Canals). Regular gameplay for entire round.

BF3 Campaign: Second misson, following scripted NPC movement.

Enemy Territory: 30 vs 30, Fueldump.

Runescape: GE, World 3. Capturing FPS while stationary. Max detail, non HTML5. x4 AA Bloom enabled. (It seems to use CPU to do AA)

Oblivion 1: Walk out in the wild, through Oblivion gate, to town gate.

Oblivion 2: NPC combat in Imperial City. Several guards/NPC vs Umbra. Spawn 50 player copies and begin combat once Umbra dies.

These were done on tests, as you can see, that vary from CPU benchmarks to CPU reliant games.

Maxforces Says:
Test setup


Results


















but if you play 3dmark you will gain some points


Some of my cores are hotter than others. Is this normal?
If the variance from hottest to coldest core is 10C or under, I would call that normal. If it's greater, consider doing a re-paste.

Ram XMP profile doesn't work.
Make sure the motherboard bios has been updated to the latest version. If that doesn't work, try adding a bit more ram voltage, SA voltage, and VCCIO.

Monitoring software shows incorrect data.
Make sure you are using HWinfo and make sure it is the latest beta version.

Prime95 stopped and says there's an error.
Most likely it is a rounding error. This means you've failed the test, but in a more minor way such that the computer doesn't crash. There is some data to suggest that Prime95 gives out rounding errors very frequently, even in overclocks considered functionally stable.

My temperatures are through the roof!
  • Stop using IBT or Linpack or Prime95. Use our custom x264 test or use something similar like ROG Realbench,
  • If your temperatures from core to core vary over 10C, consider a re-paste of your thermal paste.
  • What ambient temperatures are we talking about here? Are you sitting in an oven?
  • Hyperthreading makes your CPU hotter.
  • CPU delidding service is $50 in the United States from SiliconLottery.
  • I recommend using D14 or better in terms of cooling.
  • Ensure the cooling solution is mounted properly.

My CPU is downclocking under load!

Check your motherboard's power settings. Set them to max.

The Vcore is far higher than what I've set in the BIOS under load!

Your LLC is probably being overly aggressive. If possible, please set it to a lower amount manually.
Here is an IPC test. (Please note that I showed PCPerspective's IPC charge in the general Skylake spoiler and they have a page about IPC gains as well.

The gains in chess were smaller than I anticipated and the gains in gaming were larger than I expected.



Oblivion and Skyrim were tested in a heavily CPU bottlenecked area. For music encoding I turned 2 flac tracks, 1 wav track, and 1 podcast in mp3 to OGG files through Foobar at 70kbps bitrate. Chess was tested with Arena 3.5 GUI and the latest Stockfish beta at the time with Sedat Canbaz's benchmarking position. Also worth noting is that overclocking the ram from DDR4 2133 to DDR4 3000 gave me a 6% FPS increase in Oblivion (CPU bottlenecked game, apparently also ram as well).



Above is a power consumption test. To state the obvious, this is a measurement of the power draw of the entire system as measured by Kill-a-Watt. ROG Realbench and Battlefield 4 use up a lot of the GPU, making the test pointless. Ram stressing is also a part of the power consumption when looking at tests like Prime or Linpack.

The PSU used was an EVGA Supernova P2 1000w. It is platinum rated, but being a 1000w rated power supply, it is not very efficient at such low loads.



Here is a small chart showing the power consumption differences between 1.4v and 1.2v on the Vcore, and idle power draw differences. Some of you may have seen a chart from Overclocking.Guide showing no power consumption differences between different power saving options. That test is relatively pointless because none of the settings changed idle voltage. Running C-states and such and setting the processor to use 1.4v on load is no difference power-draw-from-the-wall-wise compared to simply setting 1.4v manual voltage and putting that under load. While C8 state saved only 35w when compared to an idle 1.4v setting, many people do not turn off their computer or put it on standby when they are not using the computer. 35w-hour saved for 8 hours a day over a year can add up.

But don't turn on C-states out of concern for the longevity of your processor. The processor is obviously idle when C-states would kick in if it's on, so there's really not much difference.

Time-to-Crash Data:

The settings used for this test are modified ones from temperature testing. The ram was clocked back down to 2133. Linpack takes a minute to load up to maximum load, so for the results I have removed a minute from each result.

Notation is as follows:

3:00, 3/5 meaning 5 runs were done, 3 passed an hour, and 2 failed. Of the 3 that failed, the average time to crash was 3 minutes.

4.7/4.7GHz:

1.36v:

Linpack Max - 2:47, 0/5

OCCT S - 3:38, 0/6

OCCT L - 3:52, 0/3

P95 v28.7 8k - 7:00, 0/35

P95 v28.7 4069k - 1:47, 0/35

P95 v27.8 8k - 10:33, 2/5

IBT Max - 47:00, 1/2

1.3v:

x264 16T passes overnight.

1.29v:

x264 16T passes overnight 50% of the time.

4.6/4.6 GHz:

1.34v:

OCCT S - 4:34, 0/10

P95 v28.7 8k - 13:41, 2/10

Linpack Maximum - 25:04, 1/4

1.335v:

OCCT S - 2:35, 0/6

OCCT L - 3:14, 0/5

Click here to view the Skylake Overclocking Chart in a new tab!

https://docs.google.com/spreadsheets/d/1wQwYMGsSnMpKxrEesNVUSoP7hGykFWw1ygJsxwx64e8/edit?usp=sharing;output=html&widget=true

Sample Size142
Average OC4.68Median OC4.70
Average Vcore1.38Median Vcore1.38



*The Vcore values in the graph above and the average Vcore statistic are derived from the Vcore under load column of the overclock settings chart. Non 6600k or 6700k entries are ignored for the purposes of finding average & median OC/Vcore.

6600k6700k
Average Clock4.644.69
Average Voltage1.381.39

This gives you a rough idea of where you stand.
Top 1%5
Top 4%4.9
Top 22%4.8
46th Percentile4.7
Bottom 32%4.6
Bottom 9%4.5
Bottom 2.5%4.4
In order to be charted you need to fill out this form:

Username:
CPU Model:
Base Clock:
Core Multiplier:
Core Frequency:
Cache Frequency:
Vcore in UEFI: This is the CPU core voltage value you input into BIOS/UEFI.
Vcore: This is the *average* CPU Vcore reading from Hwinfo or HWMonitor under load. "Load" depends on what you're stressing.
FCLK: Reminder: In HWinfo, it is "System Agent Clock".
Cooling Solution: If you are delidded, note it here. Please explicitly state if you are doing bare die or not.
Stability Test: Please list the version of Prime95 and what FFT preset/size it is if you are using Prime95. Please list the number of threads used if using custom x264 test. In other words, please provide as many details as you can. Acceptable stress tests will be listed at the bottom.

Batch Number: What country? Please list the entire batch number if you can. You can find it on the box.
Ram Speed: State the frequency and timings (3200 16-16-16-35, etc etc)
Ram Voltage: If you've tweaked VCCIO and System Agent, please note it here.
Motherboard:
LLC Setting: If you didn't change default, say AUTO
Misc Comments: Anything else you'd like to add on top of what you've listed?

To be charted in the main chart, you must fulfill one of the following requirements:
Prime v28.7 1 hour
OCCT 4.4.1 1 hour
Linpack from the Linpack Package download run at max settings (not recommended) 2 hours
Prime v27.9 3 hours
IBT 3 hours
x264 16T 5 hours
Realbench 5 hours

Aida64 and XTU do not count no matter the length of the test.

Picture Verification:
To have this optional column ticked off, you must submit a picture showing that the stress test has been completed as you claimed. You also must have HWinfo open, showing both the frequency and Vcore. (Many people forget to make sure the Vcore reading is showing.)

Please try to be honest about what stability your CPU has. If the CPU Bsods later please come back and make a followup post. I'm spending a lot of my own time to maintain this chart and write this guide to help others and if you can submit your results and keep them validated and current, it will help me and other people a lot.
To Do List:
Test power usage as voltage/frequency changes

Separate temperature chart for different voltage and freq

3/30/2016
Link to Prime95 has been updated to v28.9 from v28.7.

3/21/2016
Typo fixed ('must meet the follow')

11/2/2015
IO and SA safe voltage varies switched around.

10/27/2015
Added spoiler that displays the percentile stuff and other misc statistics from the bottom of the settings chart.

10/19/2015
Removed part about Aida64 FPU only in the stress testing hierarchy.
Modified charting form to be easier to read and more clear.
Minor cleanup in the chart.

10/6/2015
Fixed a funny typo about "if possible" -> "impossible"

10/05/2015
Minor cleanup in the chart.
Added tip about too high vcore caused by LLC.

10/2/2015
Minor cleanup in the chart.
Chart's other chart has been moved up for more visibility.
Overclocking section has been overhauled yet again. Fclk stuff might be too confusing.
Specified what exactly the Vcore statistic for average vcore and the vcore in the chart is from.
In the chart, Asrock has been fixed. It's now "ASRock".
Removed the 'it's here' line in pre-overclocking spoiler for cosmetic reasons.
Added part about downclocking chips in troubleshooting section.
Noted FPS gain in Oblivion with ram overclocking.
Updated section about max voltages.
Added note about Aida, XTU, etc being too easy to pass and not recommended in stress testing spoiler.
Updated temperature settings picture to change ghz -> GHz, and remove the time to crash column.
Added Time-to-Crash data in the misc spoiler, and added a piece that ranks tests by difficulty.

9/21/2015
Chart's text is now aligned to the middle of the box.
Various cleanup in the chart.
Chart now has percentile system working.
6600k vs 6700k color coded in chart, and average for each model shown.
Trendline in chart.
Chart now loses many of the columns on the right hand side.
Chart's graph has been replaced with Excel's... gotta find a way to quickly paste it over

9/15/2015
Removed thread status notice and Skylake availability notice.
Updated Fclk overclocking information.
Updated guide accordingly to new x264 minimum requirement for charting in all relevant placed.
Fixed a typo about Prime95 28.7 where I called it 28.5.
Removed "much more minor" when talking about rounding errors.
First serious attempt at editing the guide to reflect Fclk, and major overhauls of the rest of that spoiler.
Charting form modified. Now has requirements for stressing.
x264 version has been updated.

9/14/2015
Hwinfo picture has been changed to more clearly show what is what, and include Fclk setting shown.
Charting form and chart changed to account for Fclk.
Minor fixes in the chart.
Added info about SVID and adaptive voltage, and C8 and adaptive.

9/11/2015
Added tip about opening image in a new tab to view it in full size.
Added a picture showing how to set custom FFT sizes in Prime.
In pre-overclocking spoiler, added info about input voltage/cache voltage being removed for Skylake.
In that same spoiler, fixed "Nuctua" typo.
In that same spoiler, added info on Ripjaws 4 vs Ripjaws 5.
Recommendation to check temperatures for a test before leaving has been extended to 3 minutes due to behavior of IBT/Linpack.
Added some stuff about cache overclocking which I forgot to all this time.
Renamed 'quick overclocking' to 'core overclocking'.
Fixed typo... 'on changed' vs 'unchanged' in the cache frequency doesn't matter spoiler.
Chart's "VID" column changed to "Vcore in Bios".

9/10/2015
Updated the temperature chart.
Updated information about the temperature chart so the description makes sense.
Added power consumption chart of various stress tests.
Added C state power usage comparison.
Added bits about Fclk overclocking.
Small changes to disclaimer spoiler contents.
Ugh, LLC description changes in second spoiler.
WIP tag has come off, 2 minutes ahead of schedule! wewt!

9/9/2015
Added part about ram overclocks crashing CPU tests, along with 2 bsod code info.
Removed part about Linpack/IBT being bad tests, added info on delayed stressing for those two.
Typo fixed about "x254 can do the same".
Added part about base clock allowing for decimal places.

9/8/2015
Charting form stress test form has been edited a little for more precise information gathering...

9/6/2015
Added chart's graph to guide.

9/4/2015
Fixed adaptive voltage information.
Fixed LLC information.
Fixed Prime95 and Linpack/IBT information.
Fixed charting data (VID, amended stress testing, fixed batch number, not on IHS anymore). Altered some info after the charting form.
Removed Vring under "safe voltages".
Amended VID vs Vcore section. More work is needed.
Added troubleshooting spoiler.
Cleaned up the chart.
Put Haswell cache testing in a spoiler within a spoiler. No more clutter, but that extra information is there if somebody wants it.
Added Skylake cache testing chart.
Updated tirade about passing all stress test so the examples fit Skylake.
Changed some spacing in stress testing section.
Added IPC testing chart with the new spoiler.
Some changes to base clock overclocking section. Amended information about base clock change, affects ram, did not clarify in second spoiler.
More stress test data change... Right under the chart.
Added y-Cruncher to the download links.
Replaced Linpack with Linpack package, added information on that.
Added "update UEFI" as step 0 of overclocking, probably for the better...
Changed 'click here to view chart in a new tab' message to orange instead of blue...
Fixed all instances where I called GHz "ghz", need some consistency in the guide. (Same for MHz).
Changed second spoiler name to show the section has data on base clock differences compared to Haswell.
Added picture of HWinfo and where the vital readings are so people know where Vcore is.

9/3/2015
Obtained Skylake chip. Removed the WIP disclaimer in the thread but kept the title tag. More information pending.

8/28/2015
Prime95 link updated to v28.7.

8/24/2015
Removed cache voltage in chart form.
Added section about vid vs Vcore.
Moved top paragraph of Overclocking spoilers to pre-overclock spoiler.
Moved around some columns in the chart.
Fixed small spacing inconsistencies in Overclock spoiler.
Added snazzy message up top showing how many days until 6700k stock.
Charting form now mentions VCCIO and SA, plus a section for 'misc comments'.
Pointed out that more VCCIO and SA voltages isn't always better in ram overclocking section.
Removed suggestion to try max amount of those voltages.
Removed PCH column in chart for now.
Made sure all the spoilers are spaced equally to each other. Cosmetic improvement.
Lowered starting multiplier from 45 to 44... to prevent issues I guess...

8/23/2015
Overhauled the section on base clock changes. The original explanations were crap.
Changes to the headers in the overclock guide section to make them stand out more. Easier to read and looks better.
Changed the charting form to reflect more Skylake-specific settings. Changed styling to make it easier to read.
Removed all mention of "ring bus" or "uncore". The standard term is "cache", and so it will referred to as such.
Altered IPC comparison chart, added VID vs Freq graph.
Fixed typo at intro.

8/22/2015
Changed the chart to reflect more Skylake-specific settings like BLCK.

8/21/2015
Modified stuff about x264 bench. Updated our custom x264 to v2.05.
Re-added link to P95 v27.9 I accidentally deleted.
Stress tests now all show version number.
Stress Test spoiler title has been changed.
Made the chart larger.
Added this changelog.
Added the "THIS IS WIP" disclaimer in bold.
Modified text in "Obligatory Disclaimer"... Don't want to seem arrogant.
Did I never add info about Linpack directly in either Skylake or Haswell thread?!
Added multiple new download links to other stress tests.
Added into on BLCK differences in Skylake
Finally added info in the 'Ring Bus dun matta' section.
New spoiler for charting form, along with a section outside of spoiler showing average OCs/etc a la Haswell guide.
Added ending to the guide asking for suggestions, etc etc.
Separating out the DMI stuff from the rest of 'what's new' because it's more important.
Fixed error saying that Skylake is 25% faster than Haswell in x264
Typo? How "is" 6600k and 6700k different. Should be "are"?

8/20/2015
Crazy thread bug fixed, allowing me to edit the thread once again.

8/19/2015
Thread is created.

Thank you for checking out my guide!

Feel free to ask questions or provide suggestions!

Please read the guide before asking questions though!

Please do not PM me unless you think I've missed your post!
I love using exclamation marks!
:cheers:
 
See less See more
32
#9 ·
Quote:
Originally Posted by JackCY View Post

Feel free to copy, link, the x264, x265, Prime stuff from my sig.

Where are da OC results?
biggrin.gif
Ehhh.

My Haswell guide did well so I think I'm set up with too high of an expectation. So far it's not going as well as I would've liked... for example the forum glitch that forced me to push out a half-finished thread to begin with, with a large disclaimer saying 'if you post in here, your post is toast!'. I was also thinking about how the Skylake club is already out with a chart of its own. Although granted, when I did the Haswell guide the official thread for that was long open and Sin had his guide as well. I was one of the earlier people to break out the idea of lowering the uncore and ignoring "1:1" ratio with the core, so that helped a lot. There's not much new this time I fear. This combination of Skylake not being available to me but available to many other people and the a Skylake thread already out with a chart will make it harder for my chart to settle in. But I still think my chart is more complete and it should only get better with time. It's hurting me a lot psychologically to see the availability be 8/31 from SiliconLottery, + more time to bin. I guess I should be more focused on doing well at the task at hand. What can I do with my guide that other people have not?

But I digress (extensively).
:D


Right. So I am assuming that your test is a modified version of Angelotti's?

EDIT:

I see it definitely is. I'm going to start testing it right now and some other things. I believe I also saw Raja say that Handbrake is one of the hardest tests for Skylake, and I'll have that ready...
 
#10 ·
Im really dissapointed of my new 6700k.

I cant get it stable @4,6ghz with a vcore set to 1.4v... tried different levels of llc and so on and so on.
My old 3770k from 2012 was i running @ 4.7ghz @ same voltage as this cpu.

The temps are not my problem. When im running @ 1.4v im only hitting MAX 79c in one hour Prime95 testing so i could go further if i wanted to.

How much can i increase the vcore before the cpu could take damage? Is it the temperature wich set the limit or is it any maximum vcore?
 
#12 ·
Buy an unbinned one. I don't know why someone would pay the crazy extra for a binned chip.
Here Skylake is in shops everywhere, if you don't like it you can return it within 2 weeks no questions asked. Hell you could bin a few yourself for the cost of shipping or walking to a shop. Of course the shops wouldn't like you much probably but it's doable.

The x264 test? I downloaded the test from your mega link and rewrote the batch files to what I like so they behave more sensibly and have more options. Did that a year ago, it's easier to run them over and over, setup a shortcut with parameters and that's it. Logs don't get lost on a crash, etc. Just add/replace the new batch files. You can update the x264 binary with the latest released from web. That's how I do it. You still need the original video file etc. from your pack. I've only shared the batch files in the post as text so I don't have to keep a track of some file sharing going dead like they often like to.
Added x265 for fun not long ago. And you can combine the two in one folder, I have it that way. Just place the batch files side by side in the same folder. Add HandbrakeCLI to the binaries and that should be it.

Code:

Code:
SET Version=2.05
REM Updated by JackCY 07/05/2015
REM Based on Darkwizzie's and Angelotti's script.
So yeah based on that, the computation itself should be the same, it's just the stuff around such as user input handling and log output are modified and to me more user friendly. Supports infinitely running loop, etc. should be noted in the changes
Quote:
Updated things as far as I can remember:

x264 version is printed by the exe itself and when someone updates their x264 exe it will show the correct version
revised configuration and CLI
POPD instead of "cd.."
names are defined as variables, easy to change at one place
all options listed for priority except realtime that should rather not be used
results are generated on the fly, continuously and are not lost when test crashes due to bad OC
results show loop start time and one can find how long it ran before a crash, equal to command line output
crashed logs are deleted on a new start and not appended to
parametrized options, easy to create predefined configs with shortcuts and batches using the optional parameters
added help for parameters -h, --help
infinity now works instead of a number of loops as well
fixed loop counter when running infinite test
v2.04 cosmetic polish of code nothing more, default values used if you hit enter when asked for config values

Meaning one can also input a very high number of loops or infinity, keep it running and then stop it as desired and not lose the results.
Of course unfinished results are temporarily stored in the test folder.
Prime95 supports profiles too, haven't seen that shared anywhere so I summed it up and posted it.

eXposee: DC does on average 4.6GHz around 1.3V. It's always a lottery. You go as high as you want, of course higher voltage can break down the CPU, it's not the overall temperature of the package that will kill it.

I don't know where you've seen Intel recommendation. But even stock voltages on Skylake seem to be all over the place or read wrong in the software. My bet is that Intel is pushing the voltages to the limit to get better clocks on stock as to match up the DC performance, because otherwise nearly no one would buy newer but slower CPUs. The performance boost is what, again only around 5% over DC at stock clocks. Lets do that math shall we? DC 4.4GHz is 100%, SL is 105%, so to match the DC performance you need SL to run... wait for it... 4.4*100/105=... moment of silence... 4.2GHz. There you go, exactly what Intel is clocking them at stock. Sure SL is nice, we will see how well the changes do, how mobo manufactures do with the voltage control back in their hands and how well the lower node clocks. Also SL consumes more power than DC despite being on a smaller node, funny ain't it? Well they are pushing all they can out of it IMHO so they can match the DC.

I would have to check the reviews again but that's the way it seemed to me. SL more power hungry and as expected only 5% faster. Sometimes DC faster than SL. BW was a step back compared to DC due to lack of clocks and the gain otherwise couldn't match DC. Now with SL the boost in IPC is big enough to compensate for the clock loss. But that's about it. Make a DC and SL system, do a blind test and the person wouldn't know a difference to tell them apart based on application performance. Upgrading from Sandy is probably worth it, from Ivy might not be. From HW/DC no way.

I do like the boost SL i5 has gotten though. Would I tell a difference compared to my DC 4.5GHz i5... no way.
 
#13 ·
Quote:
Originally Posted by eXposee View Post

Im really dissapointed of my new 6700k.

I cant get it stable @4,6ghz with a vcore set to 1.4v... tried different levels of llc and so on and so on.
My old 3770k from 2012 was i running @ 4.7ghz @ same voltage as this cpu.

The temps are not my problem. When im running @ 1.4v im only hitting MAX 79c in one hour Prime95 testing so i could go further if i wanted to.

How much can i increase the vcore before the cpu could take damage? Is it the temperature wich set the limit or is it any maximum vcore?
Hi there, same is true for me.
I cant get 4.6 GHz stable without exceeding 1.4V in Bios settings.
Nontheless i have 4560 Mhz with 190 BLK stable at 1.4 Volts.
Beats my Sandybride @ 4.6 GHz by every aspect.

See specs here:
http://www.sysprofile.de/id190143
 
#14 ·
Quote:
Originally Posted by eXposee View Post

Im really dissapointed of my new 6700k.

I cant get it stable @4,6ghz with a vcore set to 1.4v... tried different levels of llc and so on and so on.
My old 3770k from 2012 was i running @ 4.7ghz @ same voltage as this cpu.

The temps are not my problem. When im running @ 1.4v im only hitting MAX 79c in one hour Prime95 testing so i could go further if i wanted to.

How much can i increase the vcore before the cpu could take damage? Is it the temperature wich set the limit or is it any maximum vcore?
Intel does not recommend to go over 1.45v... is that because of the temperatures or any other reason?
Hey!

Firstly, can you show me where Intel says 1.45v is the max? Asus' guide says the same thing, I suspect Asus just took whatever Intel said.

Here's some stuff I found on the internet:



This site says that going above 85C makes their CPU unstable no matter the voltage. I'm no so sure about that though.

Following the guy's chart, it's going to take over 1.46v to stabilize.

Digitaltrends says:

Quote:
Asus and Intel documents recommended a maximum of 1.45V
Anandtech:

Quote:
Out of the four samples, one engineering sample had 4.7 GHz at 1.4V two engineering samples achieved 4.6 GHz at ~1.4V and the one retail sample had 4.5 GHz at 1.275 volts before declocking when it was running at 4.6 GHz / 1.4 volts.

I have had two different manufacturers (MSI and ASUS) both confirm that internally they are seeing the majority of their samples hit around the 4.6 GHz mark, and it seems to be very consistent. A couple of my fellow reviewers have also been in contact with what they have, with more reports around the 4.6-4.7 GHz mark

So it seems that an average overclock, albeit with a good cooler in nice conditions, is around the 4.6 GHz mark with a great overclock more towards to 4.8 GHz. Note that this is still early in the product lifecycle and BIOSes can still improve.
SiliconLottery:

Quote:
From the few I've played with, we're looking at a 4.5GHz-4.8GHz range of overclocks.
Asus guide:

https://drive.google.com/file/d/0Bz2VRRbLPrZnMXBnOXRWeVlHcHM/view?pli=1

Quote:
Good samples can achieve 4.7GHz with around 1.40 Vcore fully stable. The highest voltage we

recommend using is 1.42V with triple-rad water-cooling if running stress tests with AVX2 routines

(lower if ambient temps are high). Those not concerned with stress testing may wish to use up to

1.45V for maximum CPU frequency.
I've looked at Haswell spec sheet from Intel. Thing's like a hundred pages, and all I got was max VID, not telling me much of anything. There isn't even a Skylake spec sheet for the public.

I don't have the chip in my hands yet, but these are the things I'd say:

-LLC, if it's like Haswell, only increases input voltage. Raja says that setting LLC on very high puts voltage above what the user sets (which is normal) so that:

Quote:
This is so the momentary transient voltage does not dip too far below the user set voltage.
-If you raise voltage by a large amount and see zero improvement in stability, then it's likely input voltage that needs increasing.

-Make sure Uncore (if it's on Skylake) isn't overclocked (manually or automatically) to a very high level. Best to have uncore set low so we don't have to worry about it crashing an otherwise stable overlock

-What Prime95? If v28, stop it and use 27.9. Or, even consider x264 test. The latter will give you more headroom, both in terms of voltage and temps.

-If 1.4 is seen at stock, intuitively I'd think 1.45 is safe, especially for people who only play games. Safety depends a lot on how you use the CPU in the first place. Gaming a few hours is not the same as constant 100% load on the CPU over days which I've done.

-Base clock can be adjusted in a very fine-grain way unlike in the past, maybe that will help.

I hope I helped in some way, and I'll update as I find more info.
 
#15 ·
Quote:
Originally Posted by Darkwizzie View Post

Ehhh.

My Haswell guide did well so I think I'm set up with too high of an expectation. So far it's not going as well as I would've liked... for example the forum glitch that forced me to push out a half-finished thread to begin with, with a large disclaimer saying 'if you post in here, your post is toast!'. I was also thinking about how the Skylake club is already out with a chart of its own. Although granted, when I did the Haswell guide the official thread for that was long open and Sin had his guide as well. I was one of the earlier people to break out the idea of lowering the uncore and ignoring "1:1" ratio with the core, so that helped a lot. There's not much new this time I fear. This combination of Skylake not being available to me but available to many other people and the a Skylake thread already out with a chart will make it harder for my chart to settle in. But I still think my chart is more complete and it should only get better with time. It's hurting me a lot psychologically to see the availability be 8/31 from SiliconLottery, + more time to bin. I guess I should be more focused on doing well at the task at hand. What can I do with my guide that other people have not?

But I digress (extensively).
biggrin.gif


Right. So I am assuming that your test is a modified version of Angelotti's?

EDIT:
I see it definitely is. I'm going to start testing it right now and some other things. I believe I also saw Raja say that Handbrake is one of the hardest tests for Skylake, and I'll have that ready...
Well right now there's no Skylake overclocking guide out there at all, apart from some generic increase voltage until stable guides. So that would be something to make your thread unique.
 
#16 ·
Quote:
Originally Posted by Z0eff View Post

Well right now there's no Skylake overclocking guide out there at all, apart from some generic increase voltage until stable guides. So that would be something to make your thread unique.
The ASUS guide contains more insider information, I hope I can incorporate that, add in: My stress temperature chart updated, a pimped out and easy to use x264 stressing solution, and an extensive overclock charting system like I did with Haswell, hopefully improved in some way...
:D


I just hope I do a good job.
:D


Trying to answer the first question in this thread and I'm all a sudden swamped with things to do and test before my chip even goes into stock.
 
#17 ·
Thank your input!

Here i found the recommendation.
"That was only possible with the CPU Core Voltage at a whopping 1.43, which is pretty high (Asus and Intel documents recommended a maximum of 1.45V)."

http://www.digitaltrends.com/computing/overclocking-intel-core-i7-6700k/

Im @ work right now but is there an option for Input voltage also?
Quote:
Originally Posted by Darkwizzie View Post

Hey!

Firstly, can you show me where Intel says 1.45v is the max? Asus' guide says the same thing, I suspect Asus just took whatever Intel said.
 
#19 ·
Quote:
Originally Posted by eXposee View Post

Thank your input!

Im @ work right now but is there an option for Input voltage also?
You're welcome.

I'd like to highlight the fact that I don't have a Skylake chip yet. I'm assuming that these things are the same from Haswell, which may be a false assumption (since Skylake took out the FIVR). In Haswell, input voltage is VCCIN is Vrin. I just read this from the ASUS guide:

Quote:
Unlike Haswell, the Skylake architecture does not contain an integrated voltage regulator

for Vcore and other "major" voltage rails (some minor rails are derived internally). The DC

regulation for each of these rails is supplied by the motherboard rather than having a lone

1.8VDC (VCCIN) input rail. This change helps reduce on-die heat generation - which has

benefits for overclocking. As an added bonus for benchmarking fanatics, this also makes the

processors more tolerant of low temperatures when being cooled with liquid nitrogen.
So it says that the removal of FIVR means there probably isn't an input voltage setting? Would you mind testing what voltages increase when LLC is set to minumum vs max under load?

Quote:
Here i found the recommendation.
"That was only possible with the CPU Core Voltage at a whopping 1.43, which is pretty high (Asus and Intel documents recommended a maximum of 1.45V)."

http://www.digitaltrends.com/computing/overclocking-intel-core-i7-6700k/
I've read that from that very web page. But it doesn't tell me where the data is from (as in, link to source). I see where the ASUS guide says 1.45v, but it'd be cool to see the Intel one. I don't even see an official Skylake spec sheet.
 
#20 ·
I've been playing games at 4.8GHz with my 6700K using an override voltage of 1.47, which looks more like 1.488 to 1.5 in CPUz. If 1.45 really is the max that is recommended then I wonder how long my chip will last until degradation becomes apparent...
 
#21 ·
Quote:
Originally Posted by JackCY View Post

Buy an unbinned one. I don't know why someone would pay the crazy extra for a binned chip.
Here Skylake is in shops everywhere, if you don't like it you can return it within 2 weeks no questions asked. Hell you could bin a few yourself for the cost of shipping or walking to a shop. Of course the shops wouldn't like you much probably but it's doable.
I am CPU bottlenecked for many things though. I bought a 980ti from 7970ghz when I wasn't even GPU bottlenecked on Skyrim. In order to maintain logical consistency, I should be buying a binned chip.

Hell, starting up Skyrim is CPU bottlenecked. So is starting up CCleaner with CCenhancer. I'm also getting intermittent stutters in Skyrim now I've applied 4k textures everywhere, and that's probably either CPU or IO bottlenecked.

Granted, even in a CPU benchmark, the extra 200mhz won't really make a difference.

Quote:
The x264 test? I downloaded the test from your mega link and rewrote the batch files to what I like so they behave more sensibly and have more options. Did that a year ago, it's easier to run them over and over, setup a shortcut with parameters and that's it. Logs don't get lost on a crash, etc. Just add/replace the new batch files. You can update the x264 binary with the latest released from web. That's how I do it. You still need the original video file etc. from your pack. I've only shared the batch files in the post as text so I don't have to keep a track of some file sharing going dead like they often like to.
Added x265 for fun not long ago. And you can combine the two in one folder, I have it that way. Just place the batch files side by side in the same folder. Add HandbrakeCLI to the binaries and that should be it.

Code:

Code:
SET Version=2.05
REM Updated by JackCY 07/05/2015
REM Based on Darkwizzie's and Angelotti's script.
So yeah based on that, the computation itself should be the same, it's just the stuff around such as user input handling and log output are modified and to me more user friendly. Supports infinitely running loop, etc. should be noted in the changes
Prime95 supports profiles too, haven't seen that shared anywhere so I summed it up and posted it.


Your link to the x264 binaries brings up two possible downloads. I'm assuming the top one is 10b and the bottom one is 8b and is the one I'm supposed to use? I've tested both and I've found the bottom one to cause 2C higher temps.

I'm working on the update right now. Thanks for your work, jackyCY.

On an unrelated note, I'm considering doing stess test stress testing. As in, I get an overclock that is not fully stable. Then I run the stress test 5 runs. Each run is a run until it crashes. That way, I can put an estimate on how well a test finds instabilities.

Quote:
Originally Posted by Z0eff View Post

I've been playing games at 4.8GHz with my 6700K using an override voltage of 1.47, which looks more like 1.488 to 1.5 in CPUz. If 1.45 really is the max that is recommended then I wonder how long my chip will last until degradation becomes apparent...
Too bad I can't really test this. Degrading chips on purpose aside (which I can afford), the test would take a long time. And if I say "degraded at 1.5v after half a year of 24/7 Chess work", how does that even translate to gaming loads where a person is gaming just for a few hours at a time?
:thinking:
 
#22 ·
]
Quote:
Originally Posted by Darkwizzie View Post

So it says that the removal of FIVR means there probably isn't an input voltage setting? Would you mind testing what voltages increase when LLC is set to minumum vs max under load?
Yes i will post back when i get any time to test. I think maybe today!
smile.gif
 
#23 ·
Quote:
Originally Posted by Z0eff View Post

Well right now there's no Skylake overclocking guide out there at all, apart from some generic increase voltage until stable guides. So that would be something to make your thread unique.
*cough* since like 90s, OCing is all the same. Adjust voltages and clocks. That's it. No more HW changes possible, and they were doing the same thing too XD I don't know why people need a guide for increasing and decreasing a few parameters
biggrin.gif

Quote:
Originally Posted by Z0eff View Post

I've been playing games at 4.8GHz with my 6700K using an override voltage of 1.47, which looks more like 1.488 to 1.5 in CPUz. If 1.45 really is the max that is recommended then I wonder how long my chip will last until degradation becomes apparent...
I wonder what do you cool the CPU with?
With a full load from x264 that thing must be cooking and running linpack a suicide without a delid and water cooling or sub zero.

Or are the Skylake chips that might actually have a worse TIM again than DC somehow magically running lower temps while consuming more power?
I don't get it.
1.4V+ that's nuts.

DC is cooking above 1.3V and even the difference between 1.23V and 1.28V is 5C, pretty much equal to the change of 0.05V on my air cooler and regular load, games, encoding.
Quote:
Originally Posted by Darkwizzie View Post

I'm working on the update right now. Thanks for your work, jackyCY.

On an unrelated note, I'm considering doing stess test stress testing. As in, I get an overclock that is not fully stable. Then I run the stress test 5 runs. Each run is a run until it crashes. That way, I can put an estimate on how well a test finds instabilities.
You want the 8bit version, at least that's what you guys used for your original test. Sure you could swap out for 10b probably, the parameters should be the same but there is no gain in that I think for stress testing.
Handbrake x264 maybe also x265 works well for crashing unstable OC. At least that's what got mine after a year, didn't recheck with the x264 test if it would also fail. But you're on that now. So have fun.
 
#24 ·
Quote:
Originally Posted by JackCY View Post

You want the 8bit version, at least that's what you guys used for your original test. Sure you could swap out for 10b probably, the parameters should be the same but there is no gain in that I think for stress testing.
Handbrake x264 maybe also x265 works well for crashing unstable OC. At least that's what got mine after a year, didn't recheck with the x264 test if it would also fail. But you're on that now. So have fun.
4 threads got me 65C (yeah, D14 is on low RPM atm).

16 threads of Angelotti's x264: 67.6C

16 threads of JackCY's exe + updated files: 67.98C

Looks like there's no regression. Tested chess for giggles and got 63-64C. I'll update the link in the thread shortly.

Quote:
*cough* since like 90s, OCing is all the same. Adjust voltages and clocks. That's it. No more HW changes possible, and they were doing the same thing too XD I don't know why people need a guide for increasing and decreasing a few parameters
biggrin.gif
Hey, with Haswell some folks were trying to run 46/46 core/uncore and that was enough to cause instability. Extra vrin was rarely needed except for 1.4v and up, but it's still there. Then there's the peripheral stuff, like our x264 test, stress test temperature chart, intro showing the specs and changes with new generation (for example, some confusion about what the "x20 PCIE lanes" mean), etc. Somebody must have found it useful.

Quote:
I wonder what do you cool the CPU with?
With a full load from x264 that thing must be cooking and running linpack a suicide without a delid and water cooling or sub zero.
I recall hitting ~90s degrees but below 100C when doing 1.52v on chess I think it was. That's with Haswell and a D14. DC being cooler, Skylake being DC temps. Of course, Prime95 v28 or god forbid Linpack would be insanity. 1.52v at chess is cooler than 1.25v on Linpack.
:devil:


EDIT:

Correction, 4 threads got me 65C. Running more threads does make the stress test hotter. (Even though my CPU doesn't have HT.)
 
#25 ·
Yeah most of the OC guides I've seen are mostly a copy of specs, review charts, etc. bundled with more focus on what the parameters mean for those who don't know or can't figure it out.
Personally what I look for in the OC guides are charts of OC results, and OC statistics. What scaling there can be, what are the average clocks for the chips at what voltages. How temperature and wattage raises with volts, etc. comparisons. Stuff I can't or don't wanna be bothered to do just out of curiosity because it's time consuming.

The rest turns into chaos and debate over who's right.
I'm happily running 1.23V VID with 1.65V Vccin but some would argue that it's too low and you should keep it 1.8V+ and so on. I say to each their own, tweak it to what works for you
smile.gif


Tweak one at a time, most important first, and that's about it. Same thing for anything I got to OC in the last 20 years.
 
#26 ·
CPUs seem to be very different from CPU to CPU. Mine is stable at 4600 MHZ and 1.26 Volts. I have overwritten voltage manually in BIOS.

Here is a pic running Prime95 for 2 hours (I6700k, Asus 170 Deluxe, Corsair H110i GTX). Temps normally around 60 Celsius max peak at 70.

http://abload.de/img/alltagsocstefxyz58s5m.jpg



Ram runs at 2933 MHZ.

Also putting mine on more than 1.4 Volts increases Temps even with the water cooling to 90 Celsius peaks. So looks like chips vary a lot. I am very happy with mine.

Stefan
 
Top