[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Public WebGL] context loss, slowdowns and aw-snap dialogs



On Thu, Apr 23, 2015 at 8:19 PM, Kenneth Russell <kbr@google.com> wrote:
Determining the amount of available "video memory" is an ill-defined
problem. Some GPUs have different memory pools for different object
types, and some use unified memory architectures. Some operating
systems deal better with paging resources off the GPU than others.
I'm aware of this, it's also the reason I bring it up again.

The WebGL working group has asked GPU vendors on multiple occasions to
define APIs to query the amount of video memory identically across
GPUs, but there has been no success.
It'll have to happen one way or another. GPUs are increasingly taking over data management and production (as opposed to merely consumption). Before long (certainly within the next few years) applications will emerge that host most data and logic on the GPU/vram rather than on the CPU/ram (there are already a variety of such applications cropping up all over the place).

If the WebGL WG has tried, and failed, to prompt a cross-vendor solution to this problem, doesn't mean it'll go away, like so many other issues of similar nature. It just means that the problem wasn't escalated properly/severly enough. I'd like GPU vendors to try to reach a solution to this before it becomes an issue of such prevalence that it becomes publicly embarrassing.