Overclocking

can this be considered ''mild'' overclocking ?

>>333633
what cpu do you have. most likely you will not be able to overclock

>>333642
it's an integrated intel i5

>>333633
That's a soft requirement. If the game says you need DX12 and your GPU only handles up to 11, or you're not on Windows 10, then you literally can't play it. But if it says 2 Ghz and you have 1.8, there's nothing stopping the game from running on your computer. And if you can't run it at a playable framerate, overclocking won't make it playable. Unless your BIOS has a built-in OC function, I wouldn't bother.

But if it does, and you decide that you want to try it, AND you have an aftermarket cooler for safety, then do it slow. Run a temp test using something like RealTemp, increase the speed by .1 Ghz on all cores, run the test again, and stop once your temperatures get dangerously high and/or you get a BSOD.

>>333633
>>333770
What he said, that's not a "real" requirement, that's just a recommendation. Squeezing another 0.2 GHZ out of your CPU isn't going to make a performance difference. (And overclocking laptops in general is a bad idea, as they tend to already be running hot.)

Usually, with games, the GPU is where you get hammered, and, as he also said, only when a different version of DirectX than your OS supports is the game going to refuse to run at all. (Possible exception for OpenGL games and some versions of the GMA chips.)

Sadly, unless you have a particularly fancy gaming laptop, you may not have the option of upgrading your GPU, and if it isn't an NVidia or AMD close to what the game wants, overclocking that isn't going to help much either. Just set the game to "shit graphics for potato" mode, and hope for the best, until you can get a new machine.

>>333831
Might be worth making sure your CPU cores aren't parked though. Laptops have a nasty habit of doing that.

https://bitsum.com/parkcontrol/

>>333836
I don't say this lightly, but that looks like a giant load of steaming bollocks.

"parked" is not a term from computer architecture; they made it up.

The crux of their "problem" seems to be:
- processors slow down when there's nothing to do (this is why your PC is not 87 degrees at idle)
- it takes a few milliseconds for them to speed up again
- OMG wasted performance in those milliseconds
- Let's just run the processor at full speed all the time what could possibly go wrong

File: Overclocked_Laptop.png (162KB, 807x546px) Image search: [Google]

162KB, 807x546px

>>333633
Do it. I overclock my laptop and it works.

>>333836
Oh, and any processor with "core i" in the name can only set a single core to its highest speed if all the other cores are at their lowest speed, so by preventing unused cores from being "parked" you're limiting your maximum turbo speed and harming your single-thread performance.

>>333944
The problem is that laptops under the Windows OS go into power saving mode, park CPU cores, and then never unpark them.

>>334133
That problem is better solved by updating your BIOS than by completely disabling Speedstep and Turbo.

And obviously if (like most people) you don't have that problem in the first place, there's no need for snake oil.

>>334138
It's a problem with the Windows power saving features and compatibility, not the BIOS.

>>334191
No it's not. If it were, every computer with an Intel processor and Windows would be affected by it.

I think someone might have noticed that.

>>334202
Someone did, that's why these programs exist.

Various northbridge chipsets and motherboards handle it in different ways, and Windows is far from perfect. Hell, how often do you see a thread about a laptop refusing to charge when its plugged in, or a monitor refusing to turn off or come back on, or a laptop refusing to sleep, come back from sleep, or hard drives refusing to come back after power down, or... Suffice to say, shit like this happens all the time. Far too many hardware solutions, too many legacy products and software layers, various third party power management solutions by HP, Dell, and their ilk, all running on a closed source OS - stuff breaks.

>>334338
>how often do you see a thread about a laptop refusing to charge when its plugged in, or a monitor refusing to turn off or come back on, or a laptop refusing to sleep, come back from sleep......
That's the BIOS, you absolute imbecile.

Google "system management mode", and educate yourself.

>>334351
No, it's not, and none of those problems are - at least not usually. On the few occasions when it is, the CMOS usually isn't going to give you any access to the incompatibility problems involved, you absolute imbecile. Certainly it's rare to see a laptop interface with any options regarding core parking. More often, the BIOS isn't involved at all, it's just third party power management software and Windows not communicating, or occasionally, legacy software not informing Windows that it parked a core. For the rest of those issues it tends to much the same coupled with driver compatibility. There's only a few companies that make BIOS setups, even if every company tweaks them to hell, but there are literally thousands that make drivers and power management solutions, thus they tend to be where the problem lies. (Hell, the fleet here has two dozen HP's that have to be tweaked every time we set them up, due to Windows and battery driver incompatibility, and there are thousands of second and third party solutions to all these problems as a result.)

There's so many disconnected hands in these projects and so many combinations of hardware possible, it's really amazing anything works at all, and nearly everything has one problem or another.

>>333971
that's a gpu overclock my dude. op wants to know if he needs to overclock his cpu

>>334577
>battery driver
>windows crashes
>whilst the computer is left on the blue screen, the battery explodes