My First Ever Overclock (I7-4790K) Can anyone tell me if all this looks ok?

Associate
Joined
21 Apr 2017
Posts
175
Overclocked to 4.7GHZ and Changed the CPU Voltage to 1.25V. Not really familiar with a lot of this jargon but get some of it.

21c6uxe.png
 
Soldato
Joined
19 Apr 2003
Posts
13,513
As above - on the face of it you may have a gem.

Cheers guys just did 2 stress tests and 1 benchmark and didn't get any errors or BSOD looks good to me!

How long were your stress tests?

If they were of decent length i would use and abuse for a few days - as synthetic tests unfortunately don't guarantee stability in everyday use - and then shoot for 5GHz.

But that clock at 1.26v would be pleasing enough volt to clock ratio - but would be interesting to see if you get 5GHz stable.
 
Associate
OP
Joined
21 Apr 2017
Posts
175
As above - on the face of it you may have a gem.



How long were your stress tests?

If they were of decent length i would use and abuse for a few days - as synthetic tests unfortunately don't guarantee stability in everyday use - and then shoot for 5GHz.

But that clock at 1.26v would be pleasing enough volt to clock ratio - but would be interesting to see if you get 5GHz stable.

Around 15 minutes tbh. Very short I will leave it running abit longer this evening. Up to 2 hours :)
 
Soldato
Joined
19 Apr 2003
Posts
13,513
Up to 2 hours :)

Yes, you got a bit more testing to do - that's your minimum really.

I used to run tests overnight - but i got fed up of running lengthy stress tests only for it to BSOD while browsing the net or using sketchup etc... Plus, modern CPUS and motherboards are far more accommodating/forgiving than they used to be.

Stress for considerably longer as that will at least identify quickly an obvious stability issue - but it won't identify cold boot problems, freezing, BSODs that you may experience in everyday use if there is an underlying stability problem. Only general everyday use will pick up these issues. All, takes time i'm afraid - things look hopeful - most would buckle after 5 minutes at those voltages... :)
 
Associate
OP
Joined
21 Apr 2017
Posts
175
Yes, you got a bit more testing to do - that's your minimum really.

I used to run tests overnight - but i got fed up of running lengthy stress tests only for it to BSOD while browsing the net or using sketchup etc... Plus, modern CPUS and motherboards are far more accommodating/forgiving than they used to be.

Stress for considerably longer as that will at least identify quickly an obvious stability issue - but it won't identify cold boot problems, freezing, BSODs that you may experience in everyday use if there is an underlying stability problem. Only general everyday use will pick up these issues. All, takes time i'm afraid - things look hopeful - most would buckle after 5 minutes at those voltages... :)

With regards to most systems would buckle after 5 minutes at those voltages-can this be a bad thing in a way? I will test tonight. What's your to-go software for stress tests?
 
Soldato
Joined
19 Apr 2003
Posts
13,513
With regards to most systems would buckle after 5 minutes at those voltages-can this be a bad thing in a way?

No, not really, it can fail for a lot of reasons - but thermal problems is not going to be one of them for you, even after 2 hrs, at 1.26v.

You could experience, freezes, reboots, BSOD, black screen, thermal throttling, warning "Error on core 4 expected 0.x..." (depending on software your using) etc...

Just reset - start the process again. Obviously, if this happens too many times you could experience Window problems due to frequent crashing - but they can be fixed with 'Last Known Good configuration' or system restore.

Adds to the excitement ;)

What's your to-go software for stress tests?

I'm lazy now - i use prime95 (old school) and CineBench. With Prime i run small FFTs to torture the CPU and then a blend test as this pushes the memory and CPU combined - if i'm going to get errors i tend to get them during this test. But i go for modest clocks at the lowest possible voltage. That's why i like your 4.7 at 1.26v - it's a great clock to voltage ratio - but it's also fun to explore the limits for bragging rights.

Here's a link to stress testing software - clicky
 
Last edited:
Associate
OP
Joined
21 Apr 2017
Posts
175
No, not really, it can fail for a lot of reasons - but thermal problems is not going to be one of them for you, even after 2 hrs, at 1.26v.

You could experience, freezes, reboots, BSOD, black screen, thermal throttling, warning "Error on core 4 expected 0.x..." (depending on software your using) etc...

Just reset - start the process again. Obviously, if this happens too many times you could experience Window problems due to frequent crashing - but they can be fixed with 'Last Known Good configuration' or system restore.

Adds to the excitement ;)



I'm lazy now - i use prime95 (old school) and CineBench. With Prime i run small FFTs to torture the CPU and then a blend test as this pushes the memory and CPU combined - if i'm going to get errors i tend to get them during this test. But i go for modest clocks at the lowest possible voltage. That's why i like your 4.7 at 1.26v - it's a great clock to voltage ratio - but it's also fun to explore the limits for bragging rights.

Here's a link to stress testing software - clicky


Hello again pal, thanks for the links! I'm setting it going tonight at 8 and I shall leave it going for a good 4 hours. Appreciate your help and will post results :D
 
Associate
Joined
19 Dec 2005
Posts
1,429
What are you running cooling wise?

My 4790k is really struggling under the kraken x52 even at stock settings hitting 70deg c, I have remounted / re-applied decent thermal paste and no joy.
 
Soldato
Joined
19 Apr 2003
Posts
13,513
Hello again pal, thanks for the links! I'm setting it going tonight at 8 and I shall leave it going for a good 4 hours. Appreciate your help and will post results :D

No problem - best of luck.

I've just realised you're the guy that bought it for a bargain too - this buy just keeps on giving...
 
Associate
OP
Joined
21 Apr 2017
Posts
175
No problem - best of luck.

I've just realised you're the guy that bought it for a bargain too - this buy just keeps on giving...

That's me pal! By the way stress test went fine running at 4.8GHZ as we speak really don't want to push it any further! I don't want to risk frying anything :p
 
Soldato
Joined
19 Apr 2003
Posts
13,513
The same. Is this wise? Stress tests are ok.
Absolutely fine - great even.

The ideal is to attain the clock you're want with the lowest possible voltage possible - and be stable in all environments (cold boots too).

You have a gem - if it proves stable with everyday usage over the coming weeks.
 
Back
Top Bottom