Prepar3D Forums

Hi, please need some advise and recommendation what to do.
CPU i7 Intel 6700K Z170 Skyline 4.0 to 4.2 Ghz turbo, 32 GB Ram at 3000 MHz,
Nvidia GTX 1080ti.
Is it normal for P3D to CRASH constantly, while overclocking my cpu to 4.6 Ghz with 1.37 Volt?
Please need your Advise and help what I can do.
Benchmark is stable, no Temp issues.
Thank you
Giorgos

I have an i7-7700K, stock speed 4.2 GHz, overclocked to 4.66 GHz and running v. 4.5 + hotfix with no issues at all.

Jorgen

Stable overclock speeds can be application specific. A clock + pwr combination that's stable when running one application or benchmark may not be stable when applying the exact same profile to run another application or benchmark. I guess that's the most nuanced caveat of playing the silicon lottery.
--------------------------------------------------------------------------------------------------------------------------------------------
Memory timings might come into play for trouble-shooting reasons why an expected-to-be-stable overclock crashes. Oh GOD, what a hassle!
---------------------------------------------------------------------------------------------------------------------------------------------
Z170 are about ~5yrs old now?
Before going through any other hassle, I would order new paste and remove + reapply thermal paste to confirm your CPU is getting a uniform thermal distribution, then try running again. Unless you re-applied within the last 3yrs
---------------------------------------------------------------------------------------------------------------------------------------------

I have the same i7 6700k and it will not go beyond 4.4 with P3D v4.
Any higher and P3D crashes. A that is after running the "stress test" and passing.

4.4 is quite good enough.

And.

My motherboard, on which I have the afore-mentioned i7-7700K, is a Z170-A.

The issues reported with the 6700K MAY - I repeat MAY - be issues with the BIOS on the motherboard, the particular RAM in the system, the graphics adapter (including its driver) or all of the above in combination with a Windows version etc etc... there are literally thousands, if not millions of possibilities here.

What I would suggest as a course of action is:

1. Update the BIOS to the latest level

2. Check the installed RAM against the Qualified Vendor List - replace if necessary

3. Update Windows to the latest level - if this is under Windows 7, you need seriously to consider going to Windows 10

4. Update your graphics driver to the latest level

5. Clean out the interior of your system, taking proper precautions - yes, the system may be overheating

All of the above not necessarily in the order given.

Jorgen

My addendum to the JorgenSA post immediately above: Add another simple check, "Confirm XMP memory timing is enabled in the system BIOS"
-------------------------------------------------------------------------------------------------------------------------
As a caveat to the clean out the system as routine maintenance, I will add that thermometers and software can't be trusted to spot problems all the time or before they become significant. System components can't have thermometers everywhere, and so your software will only record a temperature after it has reached a thermometer from a source that's likely even hotter. Furthermore, the information you get is also dependent on IF <--[big "IF"]the software used by the hardware manufacturer to display or even affect a particular PCB's condition is adequate to the task. For 99.9% of users it will never matter. For the 0.1% only a fraction of them will notice that a hardware manufacturer's monitoring software might be displaying either the wrong information, or the right information in the wrong spot. So there's not much market-economics to drive investment into optimizing system-monitoring software adopted for use by a board's manufacturer.
-------------------------------------------------------------------------------------------------------------------------
The test's methodology can give you some information, but it's not complete information, and it comes with specific limitations. You can't always trust the temps being reported as accurate or relevant for troubleshooting your specific observed phenomena. The thermometers should generally be viewed as a potential safety-net to help prevent catastrophic overheating, but is best viewed as a tool to give you general guidance of "better" or "worse" thermals regarding basic environmental factors in which the hardware is running [eg: ambient room conditions, airflow, thermal pastes, copper vs tin]

Daniel,

I put that remark about cleaning out the system in there, because in my experience many users have their system units sitting on the floor - which is a BAD™ idea! It will get clogged up....

I should probably also have put in there to get a footstool or something like that to raise the system unit a least a foot off the floor... (for us metric guys, a foot is 30,5 cm)

Jorgen