in my experience an unstable voltage may not crash for quite some time if you aren't stressing it harder than any real world application ever would via p95, but once you do start that up, if it's unstable, it blacks out FAST. heat-wise it's like putting the thing under 140% cpu usage, no other synthetic ive ever used generates this kind of load

This post was edited by DCSS on Apr 23 2017 05:47pm

You know overclock can become unstable over time too
And p95 shouldn't really be used now days for stability

And why is that? It pushes the chip very hard. Is there something that pushes it even harder now?

This post was edited by DCSS on Apr 23 2017 05:50pm

Maybe you should research why aida is recommended for haswell

aida is garbage IMO for stress testing.

How so?

Too easy to pass.

Can run it all day & crash within an hour on other stress tests.

I remember p95 being so intense on Haswell that they released a nerfed version to stop them from overheating
But tell me again how Aida is a more torturous test

This post was edited by DCSS on Apr 23 2017 10:50pm

p95 is considered a poor stability program because its TOO hard on cpu's. Nothing real world will ever come close to touching it.


You can have an overclock thats 110% stable in every game / program / stress test you can think of 24/7... But fail prime in an hour.



Custom x264 loops is where its at, imo.

That's why I don't run prime for an hour. If it survives for a few minutes it'll survive anything else indefinitely, at least in my experience..