Trying The Experimental Radeon RX 480 Overclocking With AMDGPU OverDrive Isn't Going So Well

[LinuxID10T]

4% is about the highest overclock you will get on the RX 480 without undervolting and increasing the power limit, so it doesn't surprise me you weren't able to get higher.

[dungeon]

If it move it works

[mibo]

note1: This all only applies to the AMD rx480 reference design card with the release firmware. An updated firmware could redistribute the current, so that the PCIe standard is met.

The rx 480 takes an average of 83 up 88 watts which is more than specified for PC I express.

There are Windows games where the rx480 draws more than the standardised 5.5A over the 12V line of the PCIe connector. This connector can handle much more than double the current without any problem. On the mainboard the 12V line is just connected to the 24pin ATX connector to the power supply.
So, there is no real problem except that it is of course stupid to not meet the standard requirements. But who knows what AMD decision makers were thinking...

Overcloking can destabilize the main board which can lead to instability at all.

Yes, oc can lead to instabilities. That has always been true.

[fuzz]

So, there is no real problem except that it is of course stupid to not meet the standard requirements. But who knows what AMD decision makers were thinking...

It's my understanding that the "standard" people are freaking out about only applies to PCIe 1.0, as 2.0 and 3.0 have more standardized capacity.

[Spazturtle]

note1: This all only applies to the AMD rx480 reference design card with the release firmware. An updated firmware could redistribute the current, so that the PCIe standard is met.


There are Windows games where the rx480 draws more than the standardised 5.5A over the 12V line of the PCIe connector. This connector can handle much more than double the current without any problem. On the mainboard the 12V line is just connected to the 24pin ATX connector to the power supply.
So, there is no real problem except that it is of course stupid to not meet the standard requirements. But who knows what AMD decision makers were thinking...


Yes, oc can lead to instabilities. That has always been true.

The current draw issue had already been solved on Windows, latest driver reduces power consumption and improves performance. The issue should be fixed on Linux whenever the firmware is updated. The issues seams to have been caused by bug in the power optimisation that didn't exist in earlier versions.

[SaucyJack]

Sorry this not nonsense. The rx 480 takes an average of 83 up 88 watts which is more than specified for PC I express. Overcloking can destabilize the main board which can lead to instability at all.

Stop spewing bullshit, the PCIe power draw was never an issue. AMD 'fixed' it to shut the idiots up.

[smitty3268]

Not if the Windows and Linux drivers don't check the same things and limits.

It sounds like you are talking about crappy drivers then. I'm not sure the root hardware issues matter if you agree that driver software can fix it. Linux just needs better drivers.

[jiggytom]

Hey, do you have a list of all the files/commands and what they control. I had to find Fan control for the RX480 by myself in the Power folder..it's no longer in the profile setting. Looking to crease the power limit and OC.

Also my OC doens't seem to do anything with the files you sent.

[Taisserroots]

Sorry this not nonsense. The rx 480 takes an average of 83 up 88 watts which is more than specified for PC I express. Overcloking can destabilize the main board which can lead to instability at all.

Which would cause a hard shut down. if the motherboard can't supply the power.