Let's talk about potentially revolutionary hardware ideas we'll see in 1-5 years.

[RonsonPL]

edit 2 days later: OK. Let's not talk, then. ;]

Hi

In advance, let me explain: There's no info. I have no professional knowledge, I don't bring any news.

I wanted to talk with people who are (despite the age ;] ) still interested in observing how PC tech progresses.
If you happen to waste your time to google some phrases just to see if anything really, really interesing might happen in reasonably near term future, then you are like me and probably won't waste your time reading this thread. In any other case - no refunds for time wasted on reading, sorry. ;]


What do you think can change VR in revolutionary way? Let's find out all the possible scenarios of things/events that could propell VR gaming ahead with ludicrous speed. https://www.youtube.com/watch?v=ygE01sOhzz0 


What I expect to see believe isn't 100% impossible:

1. HMC memory. It has lower latency. AMD used mysterious "next-gen memory type" on their slide, after HBM2. Could be this or any of new types of HBM. It would allow for much faster graphics cards, and if used in APUs, for much better physics and many other things. HMC in consoles... yeah. I know the probability of that happening. A man can dream though, right? Well. Actually even HMC in GPUs cheaper than Titan is not certain. At least HMC is not dead and world will see a new generation of HMC announced within months. Figers crossed we'll see some good news. Latest unofficial quote I found is "compared to our HMC, HBM seems like a failed attempt to do HMC". This sounds like bragging and that might suggest they know they have something really good in their sleeves ;]

2. Budget HBM2 0,8TB/s memory. If this gets quickly popularized in PC gaming market and gets cheap and interesting for Sony/MS soon enough to be used in successors of PS4pro and Scorpio, let say 3-4 years from now, it could have a huge impact on the mainstream VR gaming quality level. edit: And also mobiles. Cheap HBM in mobiles can be significant for whole VR.

3. Intel entering gaming seriously. I know. The management of Intel has no clue about gaming market, much less about VR gaming, but things might change a lot within just a year, and 3-5 years from now we can have an Intel's hardware powering new generation of VR.

4. 7nm and such. Maybe a revolution will be possible at so small dimensions. Maybe some big on-die memory to help with graphics and latency-dependent calculations?

5. True VR design of GPU/CPU/APU. Who knows what solutions that monopolized CPU or GPU design, could surprisingly turn out to be wrong and should be removed to allow the VR-friendly design rule in it's place.

6. High bandwidth wireless data transfer for home usage never really needed low latency. It's not surprising then that we had no focus on that until now. But now we have people working on 4K120Hz wireless transmission with latency as insignificant as 4ms. Who knows how fast it will become popular and cheap. Getting rid of the cables will help a lot especially if light, small PC in a backpack starts to meet minimum requirements of CV1 (and most likely CV2 and 3). Better than 2016 mobile level of VR in open space, with no cable problem, is a half of something that will hugely help popularize VR. The other half is accurate tracking. You won't have Rift's camera or Vive's laser trackers when you walk in the park (or in dedicated spaces for VR usage, I hope 🙂 ).

7. All the things we already know about:
Curved non-LCD displays. No need to talk about it, Obvious benefits. 2018 or 2025? Who knows.
Foveated rendering, non-traditional approaches for display (still not being enthusiastic about Nvidia's idea, since there was no mention about low persistence yet). New types of lenses (especially the ones invented really recently, those which got Carmacks's attention (or was it Luckey? Anyway, someone important from Oculus commented about it saying that this looks really promising. I remember this much.)



PS. Miscelanous

PC racing games genre. Will 120-180Hz VR games set a new latency standard that could be fully utilized by developers to bring PC racing sims to a level which was never possible before? At 250MPH, it really matters if your controller motion-to-photon latency is decent or really practically non-existent. Will we see this becoming a huge e-sport mini-industry if suddenly you'll be able to have a almost pro-sim in your house if only you buy decent VR HMD and a good wheel? (every few years a bar is raised. Now we expect this one to bring it even higher: https://www.gtplanet.net/thrustmasters-new-gt-sport-wheel-to-include-direct-drive-motor/ and then highest-end wheels features appear in more mainstream wheels, and if VR causes huge boom in steering wheel sales for both PCs and consoles, good wheels will become even cheaper, since bigger production = lower costs)


This text was sponsored by the word "hope", with parts of words "naive" and "dreaming" visible behind it somehow. ;]
In co-operation with Prof. Eye M. Bored.