Hi,Â
i have a quite simple question, hoping someone of you has an answer to it:
Does Nokias S60 devices with PowerVR accelerator ( like N93, N95, N95 8GB ) its own graphics memory ( in hardware ) or are they sharing the memory with executables memory?
Best Regards,
Andi
They all use a unified memory architecture, RAM is shared between CPU and GPU.
Regards,
Georg
Thank you for the answer.
But anyway, GL is doing a copy of vertex data, to gl server. Right?
So with unified memory architecture, will it help to load the arrays to VBOs, to reduce memory consumption?
Regards,Â
Andi
It is generally a good idea to use VBOs for static vertex data. However you will not see much of a difference on devices with PowerVR MBX, even with UMA.