Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Using integrated graphics rather than dedicated card #7756

Open
dougg0k opened this issue Feb 20, 2024 · 13 comments
Open

Using integrated graphics rather than dedicated card #7756

dougg0k opened this issue Feb 20, 2024 · 13 comments

Comments

@dougg0k
Copy link

dougg0k commented Feb 20, 2024

Hi,

Pretty much the title, why alacritty are using Intel (integrated graphics) instead of nvidia, which I have confirmed to be the one that the system are using.

ver

It uses the gpu on the desktop, since it's only the GPU, but not the laptop.

@dougg0k dougg0k changed the title Using integrated graphics rather than gpu Using integrated graphics rather than dedicated card Feb 21, 2024
@chrisduerr
Copy link
Member

@kchibisov Afaik glutin does nothing to select which GPU should be used, correct? So should just be up to the OS to pick?

If you want to run it on your nvidia card, I'm pretty sure there's tools to select which graphics card you want to run an application on.

@kchibisov
Copy link
Member

We can try to ask for faster config, so it'll try to pick nvidia, but we don't control which GPU to render to, since it's not generally present in OpenGL (with the way we pick configs).

@chrisduerr
Copy link
Member

The thing is we also had people with the opposite issue in the past. So I don't think people would be happy about either.

@kchibisov
Copy link
Member

I'm pretty sure they can force a particular vendor via libglvnd's env variable, I just forgot how to do that. One can also force specific driver for everyone there.

@dougg0k
Copy link
Author

dougg0k commented Feb 23, 2024

Why force rather than just follow what the OS are using as the main card? Or maybe even provide an option if anyone would ever need to force to a specific card due to whatever reason.

@kchibisov
Copy link
Member

kchibisov commented Feb 23, 2024

We don't pass preference or force anything. We explicitly pass that it should be picked automatically.

So your automatically is not what you like and you want to force it into faster GPU, besides the fact that the default is integrated.

@dougg0k
Copy link
Author

dougg0k commented Feb 23, 2024

I didnt mentioned anything about automatically, but the main gpu being used by the OS.

Not in my case, default has always been nvidia, nothing was set by me.

@chrisduerr
Copy link
Member

chrisduerr commented Feb 23, 2024

The default on multi-GPU systems is pretty much always the integrated graphics.

@dougg0k
Copy link
Author

dougg0k commented Feb 23, 2024

I suppose that can be so in some OSs in laptops, as they consume less power.

Maybe this is not a problem, since they didnt provide a way to disable the integrated using the bios setup.

@nixpulvis
Copy link
Contributor

Didn't we used to have a setting for this, or am I misremembering?

@kchibisov
Copy link
Member

kchibisov commented Feb 23, 2024

We never had explicit setting, but we always passed that we don't prefer any special config.

We can request faster OpenGL config, but it means just always pick more performant, which doesn't mean faster, since it can require GPU to GPU copies.

@kchibisov
Copy link
Member

The thing is that if you really want your nvidia card you can do that yourself with libglvnd's variable iirc, you don't even need anything special from alacritty, the mechanism to make alacritty use specific GPU are already in place.

@chrisduerr
Copy link
Member

Didn't we used to have a setting for this, or am I misremembering?

I believe there was something in the Info.plist on macOS, but IIRC that doesn't work properly? Certainly wouldn't work on Linux.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Development

No branches or pull requests

4 participants