5800x overclock info

[MaXxBoulton]

Are you all able to pass 45min asus realbench at the set speeds?

4.7/4.75 looks to be consistent on the 5800x tbf thats impressive on all threads. I may look around for a chip myself

 

[mattup]

I haven't tried real bench but haven't had a crash with any games or benchmarks so far.

I will give it a go and see what happens.

I am really impressed with this chip, its great value and performs very well compared to the 6900K that cost me a small fortune a while ago.

By all accounts I have a golden sample... only taken me around 20 years of the silicon lottery!!

Matt

 

[Jarrod]

Hope this helps:

Temps are a solid 15c lower max! I'm guessing your frequency doesnt stay at 4.7 when stressing? Would explain the difference of 15Gflops in the intelburntest scores

 

[MaXxBoulton]

Temps are a solid 15c lower max! I'm guessing your frequency doesnt stay at 4.7 when stressing? Would explain the difference of 15Gflops in the intelburntest scores

The 15c could be down to many things. Cooler, ihs contact, open case, colder room. Hwinfo shots show the cpu is holding boost.

 

[RayRayFPS]

I do have a great push pull config and AIO is top end.

 

[Jarrod]

I kind of figure its more to do with the actual frequency dipping below 4.7 under stress - like i said would explain the why your score is 15gflops lower - what i did to check this was reset hwinfo stats, run the stress test, screenshot at the 8th test to show the average stats

 

[MaXxBoulton]

like i said would explain the why your score is 15gflops lower

Burn test uses memory calculations to stress the cpu. A lower score in burn test shows nothing about the cpu. And more about ram speed and FCLK

 

[Jarrod]

Burn test uses memory calculations to stress the cpu. A lower score in burn test shows nothing about the cpu. And more about ram speed and FCLK

ooh - strange im only using basic 3600 ram

 

[mattup]

Real bench run for 10 mins. No issues. Not going to run it longer as I don't see the point.

 

[RayRayFPS]

Burn test uses memory calculations to stress the cpu. A lower score in burn test shows nothing about the cpu. And more about ram speed and FCLK

Would make sense as was using a spare memory stick from another pc whilst running test. My 32gb has had to be RMA'd

 

[MartinPrince]

Real bench run for 10 mins. No issues. Not going to run it longer as I don't see the point.

You're using a quite old version of Realbench, you should at least get the latest one.
https://www.techspot.com/downloads/7050-asus-realbench.html

Also 10min Realbench is only for really unstable overclocks, you should really run for 30min minimum for a more thorough test. I've quite often had it pass 10 mins but then give errors after ~25min.

 

[mattup]

You're using a quite old version of Realbench, you should at least get the latest one.
https://www.techspot.com/downloads/7050-asus-realbench.html

Also 10min Realbench is only for really unstable overclocks, you should really run for 30min minimum for a more thorough test. I've quite often had it pass 10 mins but then give errors after ~25min.

Thanks for the link.

I will run it for 30 mins later.

Regards

Matt

 

[mattup]

10 mins or 30 mins all good for this CPU. Happy with that.!!

 

[MaXxBoulton]

ooh - strange im only using basic 3600 ram

If you have 3600mhz and a 1800mhz fabric clock for 1:1 ratio. You test would be a lot better than say 3600mhz with stock 1200mhz fabric clock.

But the cpu load and stress would be the same. The tests eith slower ram runs longer so the full test it still completed by the cpu

 

[Jarrod]

10 mins or 30 mins all good for this CPU. Happy with that.!!

what frequency / core voltage are ya running at bud

 

[mattup]

CPU voltage on auto

Negative curve of -25 on all cores.

3200 CL14 ram at 3600 1.4V

Nothing technical!