The (Incredible) Future of PC Gaming Panel Pax East 2014
https://www.youtube.com/watch?v=FYp38vTFpW8#t=767
Cheers shankly
Just finished watching it, very interesting video - there were many topics covered which are of key importance to the PC gaming industry. I'll break down my views of this into subsections.
Cloud Based Gaming
We've seen much more of this topic being discussed recently and it is starting to become a reality as companies such as Nvidia release new products such as their Grid computing "servers". Personally, I think that it will take a while for gaming to become mainstream over cloud based services (i.e. actually streaming everything to the user including all the graphics that we conventionally render to our monitors using local hardware - GPUs), there are many issues with this method of rendering and gaming as a whole as outlined by the whole panel, but indeed the end results of a good working system could be astounding; we could see games with insanely good graphics being rendered to our displays without having to have the local hardware to be able to run it.
However, the "traditionalist" inside me opposes this approach and I agree with the Oculus VR guy on the panel that I like to have my own rig in my house that I've built to provide me the firepower I need to run my games - taking this aspect of PC gaming away with a pure cloud computing based service would remove some of the "soul" of PC gaming in my opinion. With that said, I think that it will still be a superior experience to have your own hardware to run the most graphically intensive games and the technology required to be able to render such games over a cloud based service is still somewhat out of reach.
G-Sync and FreeSync
Whilst an industry open standard is usually the best result for consumers, I do agree with Pietersen from Nvidia that FreeSync is currently a big unknown and we don't really have much of a clue of what it can do and how it will work exactly yet. G-Sync on the other hand is here now, it is a tested, proven, working technology and I praise Nvidia for looking into the problem of tearing vs. input lag with current monitors and coming up with an effective solution (reviewers such as Anand and Linus have stated just how good they think the technology is as well as a plethora of other early users of G-Sync).
If FreeSync ends up offering a very similar solution and is made open standard then that is even more progress and I would welcome that greatly - and from what Pietersen was saying, Nvidia would too it seems. But I would still thank Nvidia for mobilising the industry to look into this problem and to create a solution, however this all speculation until we find out what FreeSync is actually all about. Until then, I will continue to praise G-Sync and might even purchase a G-Sync enabled monitor myself come summer time if there are particularly good models available.
Mantle and DirectX 12
This topic, as every regular here knows, has been the cause of many a heated debate on this sub-forum for quite some time now and will continue to be for the next year or 2 I suspect. On the whole, I agree with Chris (Star Citizen panel member) in supporting the direction Mantle is taking with low-level hardware access and removing the overheads used in DirectX 11 (a relatively high-level graphics API). However, what one must understand is that the very reason AMD are able to make Mantle work effectively as a low level API is that they know the architecture (GCN) that they are coding this graphics programming interface for and thus they can write (and have written) very specific code for their own GPU architectures allowing them to gain very low level access and to remove CPU overheads so that the processor can deal with more draw calls per frame than DirectX can deal with.
The whole reason why DirectX cannot achieve this low level access is because it is a general purpose API which is not coded for any specific graphics core architecture, there are hundreds of different configurations of computers running DX and the coding of the API cannot address each and every one of these individually (and if it did it would be a hell of a lot more bloated and inefficient than it is now). I am unsure of how this problem has been / is being solved with DirectX 12 and I would be very interested to find out the details, however that is the crux of the situation as it stands now and I hope people now realise that's why Mantle has been a relative success for AMD users (if it was coded as an industry open standard and AMD expected Nvidia to jump on it, then it wouldn't be as efficient and would suffer the same problems as DirectX 11 suffers now).
Gaming Operating Systems of the Future
Now this is a very interesting topic and one which I was glad to see being covered by this panel. In particular I agree with Pietersen in his opinion that Microsoft should release an almost "modularised" option based OS which allows users to switch between modes in the system giving them a gaming operating system when they want (turning off the enterprise services and processes) and then a low power, business based operating system at another time (switching off particularly CPU/GPU intensive tasks to allow for less power usage and more cycles dedicated to doing work and such). This is something I really would like to see in the future and I think it could revolutionise the way we use Windows and allow it become a much more streamlined, efficient and high performance operating system.
Linux is a no go for the near future in my opinion due to the fact that, as Matt Higby stated, developers would need to port their game (engine) code over to OpenGL which is a non-trivial task and is just not worth the time nor money given the relatively small customer base using Linux. The only situation in which I think this could happen is if there was a sudden mass migration to Linux from Windows which I cannot see happening any time soon, and even then it could be complicated given the various distributions of Linux that people can choose.
