I think we're reaching a point where a system akin to the PS4 can pretty much produce any type of video game experience in a 2D or 3D space imaginable at a fairly high level visual fidelity at that.

The only thing you gain really from going higher is a prettier looking game at this point, but the higher end graphics of today are still very pretty and in some cases beginning to approach photo-realism.

If you go further than this level (other than just boosting the resolution/frame rate/adding some lighting/shader effects) the issue that you come into is budget.

Sure you can make a 25 Teraflop console with 32GB of HBM2 RAM one day for example .... but who's going to invest $100 million dollars per game to get the level of visuals necessary to push such hardware? Even the biggest publishers will start to balk once one or two of their games underperforms or flops.

Ofcourse it's hard to take advantage of more cores, yet tools will come to make this easier and hopefully we'll see more advances beyond shinier graphics, so every game that tries something more like Minecraft and No man's sky doesn't automatically get put in the just an indie pile.

Intersting.... those two games you mentioned. 

if innovations like procedural generation is what you are basing ur "more powah" push on.... then I fear you have got it all backwards. Especially if you think that we need tons more processing powah to pull them off. 

Make no mistake tho, the power in these consoles is more than enough to feed any kinda gameplay innovation and then some. The issue with that has nothing to do with the power at hand but that devs or publishers simply aren't willing to take the risk. 

Anyways, I e long since accepted that you and I would simply not agree on this matter. So let's leave it at that. 

First of all this is not the first time this is happening. The steam box or incremental console model was already tried by MS in 1983 with the MSX platform. Although MSX 2 was fully BC I don't think it was forwards compatible. However you could choose to spend as much as you wanted on a MSX computer, made by different companies with different features, even Sony made them. However the market chose a specialized console instead and MSX 3 never launched.

The risk is the same now. A specialized console can outperform a more general platform both in price and ease of development. The question is, will this really help tie your user base to you. When it comes time to upgrade, for wanting something new since it's all getting a bit stale, wouldn't it be much more enticing to try out a different platform with real differences instead of an incremental upgrade? I know I would be leaning towards an XBox or NX instead of Neo.2.

Another problem is that you're effectively creating an anchor with forced forwards compatibility. How long should the base model be supported? Instead of a ps5 in 2020 where developers can decide what features they still want to port back to the past gen, they'll be stuck with no option but to only have better graphics, while everything else is still tied to 5GB ram and the slow Jaguar processor from the base model.
As we moved further away from last gen you saw that cross gen games on ps3 and 360 got less features and started running worse. Just compare the features of FH2 or Shadow of Mordor. With this incremental model there would be no car tuning, drivatars etc or Nemesis system.

And how much can you realistically let 3 versions diverge. AAA PC games don't support 9 year old hardware, definitely aren't tied to having to run with 256mb video memory, 1GB system memory. It's always easy to say that we have plenty of memory now. 8GB seemed crazy 5 years ago, heck the ps4 almost launched with 4GB. Now the 5GB available is already cramped again. Not a good idea to stick with that until 2025 (NEO has the same memory, slightly faster speed) Nor the 8 core Jaguar at 2.1 Ghz.

Tech is slowing down is used a lot as an excuse. While that's true for processor speed, for parallel processing we're still in the early stages. GPU techology is still moving on, transitioning to finFET or 3D transistors which allows to go down to 14nm GPUs with more CUDA cores. 16 core CPUs are already here and AMD is also planning to use finFET for their 32 core Zen processor.  Anchoring to only being able to use 6 Jaguar cores is going to look really stupid in 2025.

Imo enforced forwards compatiility will slow progress down. Yet if you just want the same games in a higher resolution, I guess the incremental model works for you. I hope Sony makes it clear what their plans are at E3 going beyond the NEO.

True, but I'll keep on dreaming of dynamic evolving worlds, RPGs where the seaons change, weather makes an impact on the world and the world changes due to your actions, beit a fireball setting a forest fire or diverting a river to help a new settlement grow.

We do need more powah. The ps4 already struggles with Minecraft. Too many pistons in 1 place and the engine can't keep up. FPS is stable yet resolving my contraptions starts to fail with blocks not staying attached to sticky pistons, cpu limited. Fallout 4 dropped to a crawl with a bit too many monitors in 1 place. From dust was limited to tiny maps, sadly on PC too, cross platform parity :/

Anyway we'll see. I doubt we'll still be playing the same set piece heavy corridor shooters in 8K in 20 years. At least I hope not. They'll still be there, yet I hope they'll be seen and regarded as on rail shooters are now. I'm not ready to settle just yet :)

Sorry, getting off topic. Agreed to disagree. Still 6-7 weeks to E3 ugh!

Great read, 100% agree.

What people should more worry about is the game as a service trend instead of new more power consoles.

MSX computers were manufactered and sold by various companies, consoles have only 1 gatekeeper. The two models arent comparableat all. Sony, Nintendo and MS will decide how you are going to play on their platform.

And the bolded part is an example of why having this discussion with you is somewhat irritating. 

You keep flip flopping and tossing things around all to supposedly win. Even at the expense of contradicting yourself. 

with that bolded part. you seem to suggest that all the games we get from the current hardware is nothing more than on rsil corridor expericnes. 

but you mention games like mine craft, no man sky and let me add little big planet and dreams that are already doing what you are saying you want. 

then you talk about features that u want to see that already exist in tons of games today. 

all while conviently acting like we don't already have tons of great open world experiences. 

sheesh......