How to switch between integrated gpu and dedicated gpu?

I will switch between my Radeon R9 and the integrated gpu. I have installed the propietary drivers amd catalyst 14.9. Any changes in Bios to make igpu default can’t be saved. Opensuse 13.1 runs still with dedicated gpu.

I believe it will as long as the ATI driver is active. you would need to change xorg.conf to not load it.

With Intel+NVIDIA there is a a program called bumblebee that allows you to control things but I don’t know about Intel+AMD hybrid. Also I assume this is a notebook??? If not why would you want to? In any case the BIOS as little to do with it except for allowing turning off and on the GPU on the CPU. The choice of GPU is controlled from the xorg system based on drivers installed or selection made in the /etc/X11/xorg.conf or /etc/X11/xorrg.conf.d directory


Try to load Catalyst Control Centre and choose which gpu you wish to use on the user session there. On my Aspire I was (it is currently broken) able to do the ‘switch between’ this way, however, I was unable to disable the integrated chip all together (which I attempted to do). In my case it seemed like only the integrated one was available at boot time, blacklisting the intel driver made the system un-bootable. So, I ended up using both, with dedicated gpu on user session.

Acer Aspire E1-572G, Intel integrated chip and AMD Radeon HD 8750M.


If the AMD is like the NVIDA optimus the GPU’s are tied together at the hardware level and the Intel is used to render. But this only applies to notebooks with this Rub Goldberg set up. If a desktop with a real AMD or NVIDA card it is a whole different thing. This is important to know before moving forward

Yes of course, I presumed a laptop.
If he had a desktop, he wouldn’t have an image if not connected (physically) to the right device :slight_smile:

{Assuming a notebook}

From OS:
If Catalyst is installed, try this

# aticonfig --px-igpu

to run X on the integrated videocard.
Two similar commands

# aticonfig --px-dgpu
# aticonfig --pxl

are used to switch to the dedicated card and to see which card is currently active, correspondingly.

In the case of opensource radeon driver, as far I know, vgaswitcheroo can be used.

From BIOS:
What does “can’t be saved” mean? Could you describe this in detail?

I have a similar notebook with Intel HD Graphics 4000 and AMD Radeon HD 8750M GPUs, but I cannot run X with fglrx driver. As I understand, you could do this on your Acer Aspire E1-572G. Could you give some advice at AMD/Intel drivers problem, please (now I have openSUSE 13.1 and the last drivers, but the content of outputs hasn’t changed)? Despite that thread is old, the problem is still actual… Thank you in advance

After every reboot there is still the pci-e card in the bios settings. But this is not of importance. I have a desktop pc. The reason why I will switch between igpu and dedicated gpu is that I will save energy. I will only use dedicated gpu when I am playing games.
This might be interesting?

In your case, perhaps try to mix the solution, if you get it to work, with black-listing the Radeon driver, modprobing the driver-module only when you need it.
If you use the same monitor and just switch connection port, add in anyway a second monitor (clone of the first) in the xorg configuration (section/file).

openSUSE’s method is to use the files in the xorg.conf.d directory for x11 set-up, instead of the xorg.conf file as in the example above.

It might work?

The above post was perhaps a bit far fetched, the more I think about it the more flawed the suggestion seems to potentially be.
I don’t know, for instance, if this (not loading the module) will prevent a default load of the hardware(gpu) itself, power up the fan(s) etc; I don’t think so.

On the other hand, aren’t there usually some power saving modes on most graphic-devices now; so perhaps disabling the integrated gpu instead and switching between different performance modes of the dedicated one, might be an acceptable solution for you?