Overclock.net - An Overclocking Community - Reply to Topic

Thread: Kaby Lake Overclocking Guide [With Statistics] Reply to Thread
Title:
Message:

Register Now

In order to be able to post messages on the Overclock.net - An Overclocking Community forums, you must first register.
Please enter your desired user name, your email address and other required details in the form below.
User Name:
If you do not want to register, fill this field only and the name will be used as user name for your post.
Password
Please enter a password for your user account. Note that passwords are case-sensitive.
Password:
Confirm Password:
Email Address
Please enter a valid email address for yourself.
Email Address:

Log-in


  Additional Options
Miscellaneous Options

  Topic Review (Newest First)
09-20-2019 10:21 AM
sir_dakz03
Quote: Originally Posted by BoredErica View Post
Submissions closed like 2 years ago. ._.
Yeah, I figured. But it's a good guide though, and I just recently delidded my 2-year old procie. Was just eliciting a response.
09-18-2019 07:55 PM
BoredErica Submissions closed like 2 years ago. ._.
09-16-2019 08:24 PM
sir_dakz03 Hi All,

Here's my entry:
CPU Model: i7-7700k
Base Clock: 100
Core Multiplier: 50
Core Frequency: 5.0
Cache Frequency: 4.5
Vcore in UEFI: 1.35
Vcore: 1.378
Fclk: 1
Cooling Solution: DEEPCOOL Gamer Storm Captain 240EX Delidded
Stability Test: x264 16T infinity normal 9hrs

Batch Number: Malay L639G012
Ram Speed: 2666 15 17 17 35 2T
Ram Voltage: 1.35
VCCIO: N/A
VCCSA: N/A
Motherboard Z270 MSI Gaming M3
LLC Setting: N/A
Misc: No AVX Offset; Also tried Prime95 25.11 for 3hrs but no screenshot.
07-24-2019 11:13 PM
BoredErica I wanted to do some basic maintenance on the formatting of my guide here, but with the OCN website transition and such, the OP has become a giant spaghetti monster to edit.
04-11-2019 06:08 PM
NIK1 One to Four hours..Thanks a bunch...
04-11-2019 06:07 PM
n1kobg Dont use only one Benchmark Or Stress test.
04-11-2019 02:01 PM
wingman99
Quote: Originally Posted by NIK1 View Post
I am testing and stressing my Asus Z270 IX Apex with a I7 7700k and using Rog RealBench to stress test if stable..How long should I let RB run for to test for stability..I just did a 30 min test and it stopped with errors after 5 min.I upped the core volts a bit and it was successful after the 30 min run..Should it run longer than 30 min for testing stability..Any suggestions..
I would test 1-4 hours to make sure it is stable.
04-11-2019 01:34 PM
NIK1 I am testing and stressing my Asus Z270 IX Apex with a I7 7700k and using Rog RealBench to stress test if stable..How long should I let RB run for to test for stability..I just did a 30 min test and it stopped with errors after 5 min.I upped the core volts a bit and it was successful after the 30 min run..Should it run longer than 30 min for testing stability..Any suggestions..
01-23-2019 03:12 PM
Ironclad17
Quote: Originally Posted by jambazz View Post
I betcha it is the AVX offset parameter that is set on your system.
If the x264 stress test does AVX workloads, then the CPU will lower its multipliers as not to generate too much heat. This offset can be set to 0 in the BIOS. Then normal and AVX workloads all run at the same multiplier.

This is why many people stress test with Prime95 26.6 small fft, which does not use AVX workloads, instead of the newer Prime95 versions.
I may be dumb, but I'm not that dumb. If that was the case I would have had an AVX offset of like 8! I did figure it out finally. The bios on this board seem to force the cpu back down to base clock multiplier (42) whenever it hits any instability if turbo boost is enabled, and any kind of multiplier overclock requires that setting to be enabled. I got around this with 125x40 or 122x41 core clock. This of course was unstable I'm only now learning. Turns out you really do get what you pay for with motherboards. That or I have a really awful bin.

Actually it's starting to look like my fclock was also causing instability.
01-20-2019 02:57 AM
dtfkev
Quote: Originally Posted by Falkentyne View Post
Going to have to run a battery of tests. Whatever you do, do NOT run small FFT AVX prime95. Only use AVX as I instruct here below..

First, download prime95 29.4 https://www.mersenne.org/download/
Run it then close it then open the file LOCAL.TXT and add these lines at the top:

CPUSupportsAVX=0
CPUSupportsFMA3=0

Grab HWinfo64 (another nice program) while you're at it and have it running the same time as prime95). https://www.hwinfo.com/download/
Now you can do some more reasonable tests.

First do a CPU core reliability test by doing a small FFT stress test. This will put a strong load on your CPU without it getting out of control. Don't let it run 12 hours. I personally find that you don't need to let it run more than 30 minutes, as it can test the entire 8k-16k range by then. But if you really want to be sure, you can run it an hour. Have the program HWINFO64 running and look at the bottom for WHEA or CPU L0 errors also.

If that passes, now do a Vcore reliability test by running a custom test, checkbox "run FFT's in place" and enter 1344K-1344K for the range.
This stresses vcore stability more than heat stress. You absolutely MUST pass this test. 1 hour is enough for this. Half hour if you're in a hurry.

Then when you're done, you can stress the computer's IMC and main memory by doing a custom test, uncheck FFT's in place, and enter range 512K-4096K and set the memory size to 75% of your system memory. Example: if you have 32 GB of RAM, the total is 32768, so set it to 24576. For 16 GB, it's 75% of 16384. Ezpz.

This test will take much longer to do. Might have to leave it running overnight as this tests a lot of ranges now.

Finally, you should test the AVX stability test. This is similar to the 1344K test.
Open the file local.txt and change the 0 to a 1 for AVX.
Do not change FMA3.

Then run the same 1344K (don't forget to checkbox in place fft's) and it will run with AVX. This may run a bit hotter than small FFT without AVX (oddly enough it runs cooler on a 9700K and 9900K), so let it run for up to an hour.

Always monitor HWinfo64 for WHEA errors or CPU L0 errors if a thread doesn't crash.

This will take you awhile. Post your results when finished.
Wow this is wonderful haha

Should I run each one of these tests after each change in my OC?

Or should I run only the individual test associated with the change?
(ie small fft for core ratio changes, fft 1344 in place for voltage changes, custom fft for imc and memory changes?)

I just passed all of those tests with [email protected] with everything else in bios stock. Should I manually enter my memory ratings and skip right to the custom 512k-4096k test or start at the beginning?
This thread has more than 10 replies. Click here to review the whole thread.

Posting Rules  
You may post new threads
You may post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off