i5 7600K OC temps

Your CPU can go up to 80 C max load. Your temps are good.

Remember max load should mean after a little while (half hour) should show how high temperatures will go. I at least find on my AIO cooler that temps are amazing at first but creep up, nothing bad here tho.

On air they usually spike right up anyway.

and if the 7600k is like my 7700k

they definitely run hotter than my previous skylake cpu did when set to the same speed,voltage etc

intel using rubbish heatspreader tim again i guess

stress testing by running prime95 etc to torture your cpu isnt exactly compulsory

though some people seem to think it is

if it doesnt crash doing what you actually do on your pc thats stable as far as i am concerned

The temps you get on air after 10 minutes usually don't vary between those of an hour, so yeah I'd say that's looking good.

yes could slowly lower it to see

profiles usually use more voltage than required

can use offset or adaptive

though my last board was asrock and same thing profiles used a fixed voltage

all its probably done in the profile is upped the multiplier and voltage dont think it would tweak anything else for 4.6ghz

so you could just do that manually then pick offset or adaptive depending which or if both are in your bios

Prime95 / Synthetic Stress testing programs shuda gone out the Windows w/ Haswell. You have two choices:

a) Run the older version w/o AVX and other modern instruction sets which proves that your OC is stable ... but not when AVX is present.
b) Use the new one where the AVX voltage bump is capable of permanently damaging your CPU.

The other things worth noting are:

a) you can be 24 hour prim 95 / AIDA 64 stable and fail under application based multi-tasking benchmarks.
b) did you build the PC to run applications and games or get ya name posted on competitive overclocking web sites ?

I now save a lot of OC testing time simply by using RoG Real Bench which uses AVX but in a manner whereby it's used just like "in real life" using reall applications and not in a manner that presents a danger to the CPU or which artificially limits your OC.

The profiles basically pre-selected data that works on 99% of CPUs. Therefore you particular CPU has a 99% chance of being able to run higher and cooler.... more so if you do not artificially limit the OC by asking it to sustain synthetic loads which it will never see again.

Here's some data from overclock.net

Average OC 5.08
Median OC 5.10
Average Vcore 1.36
Median Vcore 1.38

1. Many sites require x hours of P95 or other synthetic testing before you can validate your OC to be posted on their leader boards. But again, what is the point of proving that your OC is stable with a synthetic test program that you will never use again ? What happens when you try actual programs ? What happens when you multi-task several programs ? When building my current rig, I was P95 stable for 24 hours but when I tried to run and application like H.264, the system crashed. So of what value was that 24 hour test ?

2. Your P95 test proved that as long as you don't run any modern programs with AVX or any other modern instruction sets ... yeay ! But was the goal in building the PC to be able to run P95 ... or was it to run games and programs ? If it was the later, then why not use a stress test that combines several of the most demanding ***real applications*** ?

3. 10 minutes of P95 proves nothing. To be included on overclock.net's list of OC's ... you must fulfill one of the following requirements:

Synthetic tests
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

Application based test
x264 16T 8 hours
Realbench 8 hours

As i said previously, I have seen OCs pass all of those criteria and then fail in the application based tests.

3. I think you should forget the "pre-sets" and move on to determine an OC for your particular CPU. One way to show how much unnecessary headroom you have in your OC is to lower the core voltage. After using a pre-set I am usually amazed at how much lower the CPU can run.

4. Also ... you should proceed sequentially with the following to determine limits:

a) Find maximum core multiplier leaving cache at default and RAM at default JEDEC setting
b) Find maximum core multiplier leaving cache at default and RAM at XMP setting
c) Find maximum cache multiplier at max core and RAM at default JEDEC setting
c) Find maximum cache multiplier at max core and RAM at XMP setting

5. Remember that your OC is limited by the quality of PSU to deliver clean, stable voltage as well as the MoBos ability to do the same.

6. I would suggest downloading RoG real bench and HWiNFO 64 to provide stress testing and monitoring. Check your rail voltages and see how much they vary from nominal during the tests (< 1% ideal)

7. For guidance on the process, you will find some good instructions here:

http://www.overclock.net/t/1621347/kaby-lake-overclocking-guide-with-statistics