Install Steam
login
|
language
简体中文 (Simplified Chinese)
繁體中文 (Traditional Chinese)
日本語 (Japanese)
한국어 (Korean)
ไทย (Thai)
Български (Bulgarian)
Čeština (Czech)
Dansk (Danish)
Deutsch (German)
Español - España (Spanish - Spain)
Español - Latinoamérica (Spanish - Latin America)
Ελληνικά (Greek)
Français (French)
Italiano (Italian)
Bahasa Indonesia (Indonesian)
Magyar (Hungarian)
Nederlands (Dutch)
Norsk (Norwegian)
Polski (Polish)
Português (Portuguese - Portugal)
Português - Brasil (Portuguese - Brazil)
Română (Romanian)
Русский (Russian)
Suomi (Finnish)
Svenska (Swedish)
Türkçe (Turkish)
Tiếng Việt (Vietnamese)
Українська (Ukrainian)
Report a translation problem
Especially when you are not a PC Tech., Software specialist or the likes.
Re-adjust your fan curves, Set Enhanced Sync. to On, cap your FPS to your monitor's VRF.
Re-adjust your filters, save this setting, no UC/OC. Frequency & Voltage should be on Auto.
- visual artifacts.
- system instability.
- driver crashes.
- performance degradtion.
- overheating.
- noise or coil whine.
from which i don't have any.
it runs as usual.
A good system is a stable system.
A stable system is a system running stock parameters.
Just reset it back to defaults and go from there. If needed, reboot the PC after resetting the GPU back to defaults. Then re-test.
Either leave the Voltage alone and OC the Core+Mem; or leave the Core+Mem alone and Undervolt it; don't do both.
It might not be too much to crash the OS or Driver; but just enough instability to cause visual rendering issues; odd colored pixels in your OS and/or Games and such. That's the early warning sign to make some GPU related changes, either increase voltage slightly, or down-clock either Core or Mem; or both.
Seriously doubt it has damage. The purpose of an operating system crashing is to prevent damage in the first place.
Most modern hardware has additional safeguards that you must bypass to do any real damage (like modded BIOS).
But because hardware and software are unpredictable and bugs may exist, you will always be warned that you are operating at your own risk.
PCs are imperfect environments with many variables no one can predict. Because of that it's not a good idea to move outside of specs unless you can replace a part if it fails.
I also agree that overclocking is kind of pointless these days. Most GPUs are already overclocked out the box, and some will boost far higher than others on stock setting depending on driver and conditions they are at.
AMD GPUs will autoboost past their limits as long as its within power and heat specs without you ever having to overclock or risk voiding warranty. If it does not autoboost then its because it's pointless for the task or you have hit some other limit other than clock speed
Also here is an example from a game last year that was buggy that I patched up some issues in, Note that the GPU automatically boosted past its limits anyway because that is the default behavior in the drivers.
Jedi Survivor: warning spoilers...
https://www.youtube.com/watch?v=Rg9q2wgmQC0
Though it is a 6650XT, official boost is only 2635 Mhz on the card, but it autoboosts higher without any manual overclocking as long as it's possible. It actually boosted to 2999MHz at some points off the video, but this 6650XT was limited by power consumption (150W cap), Junction temperature, and VRAM. These three parameters will cut how much you gain from an OC anyway
Much like a CPU, there are parts of a GPU that don't really operate faster just because the clock speed is increased and you will also be hardcapped by the performance of those parts of the GPU.
TLDR: Let the driver do what is best and stay in warranty, even if there is little risk for damage you never know. Don't overclock what you cannot afford to lose.
The junction temperature on this card was at 115C -- past the absolute safe limit of the card and it was still not damaged.
^ Yea this basically.
Whatever you use to OC with just do not set it up to apply OC at OS Startup. That way if it proves unstable you don't want that applied after a PC reboot/startup. Simply save stable settings to a Profile Preset so you can easily switch back to that when needed.
it wont damage itself from overheating
too low core voltage just make it less stable and run cooler
raising core clocks also makes it less stable
other than crash/bsod/reboot or whatever the hardware is fine, just windows may get messed up, for whatever files are being written when it crashes
chkdsk can fix alot, but some programs that need those files may not work properly (ie windows or boot files)