Illustrator GPU Accelaration not working on newest AMD APU

[MajorSanyi]

So I wanted to use Illustrator, but currently unable to use hardware acceleration, because of the 1GB Vram hard limit. Sadly I can't assign more VRAM on my ASUS Zenbook bios,because the VRAM is locked by the vendor. It can use 8 gigs of shared memory though , so I don't think it should be the problem since the integrated memory on my laptop faster than some of the compatible GPUs. Funny thing is photoshop works fine and it has 1.5gbs of required VRAM on the system specs. I think it's bug, do you guys know any workarounds fo that?
My laptop is an 

ASUS ZenBook 14X UM5401QA-L7208W

[Anshul_Saini · Accepted solution]

Hi @MajorSanyi,

We are sorry for the inconvenience. Photoshop supports both VRAM & shared memory. On the other hand, Illustrator requires 1GB of VRAM at least. In your case, AMD integrated GPU only has 512GB VRAM, which doesn't meet the minimum system requirements for GPU performance. 

I hope this clarifies!

Regards,

Anshul Saini

[MajorSanyi]

Thanks for your clarification. Is there any plans to support shared memory? I think this is an arbitary limit, which is kinda dated with this architecture of laptops. I'm sure the gpu could handle it, there should be an option to force it, maybe with a warning message that this is against adobe recommendation. 

Again thanks for your clarification. 

[Anshul_Saini]

Thanks for getting back to me. I understand & I know that the current laptop & new GPUs are capable enough, but using shared memory will create a bottleneck and Illustrator performance will suffer, or the system can crash/ hang due to this. Let me explain this with a scenario. Illustrator will work perfectly for light work, even on a CPU on a laptop like this. It is when we use high-end rendering, 3D or graphic-intensive tasks we need GPU. Now, suppose we use GPU using shared memory in Illustrator. In that case, RAM will get overloaded, and GPU will not have enough shared memory left, which will create a bottleneck, or GPU will not get any shared memory at all as there is none available. Though I am not an engineer and not sure if it is possible or not but from my understanding, GPU on shared memory will not work unless the manufacturer assigns a fixed shared memory for a certain task, as Intel does for their Integrated GPUs.

I hope this gives more clarity.

Reagrds,
Anshul Saini