Mixed Information on Consoles or How I learned to loathe PR *spin off*

The more powerful it is, the less 'iterative' it is and the more of a new console. If you have 10 TF of power in Neo, and it's only playing PS4 games upscaled, it'd be a waste of power. The mild improvement model makes far more sense. PS4 games and enahnced on Neo. Then abandon PS4 when the PS5 comes out and provide Neo games and enhanced on PS5. And keep rolling with a low end and high end experience. But a whole new console in terms of poer, playing the same old games...it'd be like XB360 being hampered to playing XB ports.
 
The more powerful it is, the less 'iterative' it is and the more of a new console. If you have 10 TF of power in Neo, and it's only playing PS4 games upscaled, it'd be a waste of power. The mild improvement model makes far more sense. PS4 games and enahnced on Neo. Then abandon PS4 when the PS5 comes out and provide Neo games and enhanced on PS5. And keep rolling with a low end and high end experience. But a whole new console in terms of poer, playing the same old games...it'd be like XB360 being hampered to playing XB ports.
I don't think we will see a PS5. At least we won't see any more generation-changes like in the past. It will be iterative now, at least as long as x86 and the GPU are partially backwards compatible.
I really think that now we have Neo, in a few years we have Neo 2 and than the original PS4 won't be supported any longer. So a platform lives 6-8 years (neo after ~3 years, scorpio after ~4 years). That is good enough for a console. This also eliminates the problem, that there are no games at the launch of a new system. Developers can patch their games with a config-file to just increase the res, or texture filtering, or whatever to make the game look better on the new hardware.
This way publishers benefit from a permanent stable hardware-base. E.g. launch-games had a hard time in the past, just because of the small audience that exists on the launch of new consoles.
 
I don't think we will see a PS5. At least we won't see any more generation-changes like in the past. It will be iterative now, at least as long as x86 and the GPU are partially backwards compatible.
I really think that now we have Neo, in a few years we have Neo 2 and than the original PS4 won't be supported any longer. So a platform lives 6-8 years (neo after ~3 years, scorpio after ~4 years). That is good enough for a console. This also eliminates the problem, that there are no games at the launch of a new system. Developers can patch their games with a config-file to just increase the res, or texture filtering, or whatever to make the game look better on the new hardware.
This way publishers benefit from a permanent stable hardware-base. E.g. launch-games had a hard time in the past, just because of the small audience that exists on the launch of new consoles.
The developers knew which hardware to target and how when hardware was fixed. With each generation the developers could measure better whats new in the hardware and take advantage of it without having to worry about the previous hardware. Now, they are targeting multiple iterations simultaneously and with some rules on top.
 
Still a very limited set so specific optimizations remain possible?
 
I don't think we will see a PS5. At least we won't see any more generation-changes like in the past.

Actually, Shu (Shuhei Yoshida) hinted to this as well. It really makes no sense anymore on labeling generations - as these mid-cycle refreshes will carry legacy (backwards compatibility) games with them forward and phase out the older models when needed.

I believe Sony new goal (or maybe since PS4 launch), is to grow the PlayStation ecosystem with its current userbase for the forcible future and not start all over (again) with every new generation. You damn well can bet, that's what MS is doing with the convergence of XB1 and Win-PC users within the XBOX ecosystem.
 
The developers knew which hardware to target and how when hardware was fixed. With each generation the developers could measure better whats new in the hardware and take advantage of it without having to worry about the previous hardware. Now, they are targeting multiple iterations simultaneously and with some rules on top.

All you said can be documented / stated for every mid-cycle refresh. It's no different than iOS/android (software and hardware) on keeping apps/games developers aware on how each mid-cycle refresh effects their respective wares. IMHO, I hope generational (let's start all over) goes away... and makes Sony, Nintendo and Microsoft compete more at the userbase level (crafting better libraries, apps, features, etc...), than always relying on restarts. The mid-cycle refreshes will just help push along the needed ideals on keeping and growing the current userbases.
 
Last edited:
When you have new I/O (like a new controller or SD->HD), then you need a new console. When you don't have that you can do the iterative approach (unless you screw up on the SW/HW front).
 
It's more a case when you have new hardware that can't run the old codebase, you need a new generation, whether that's because the new paradigms are just better, or the old design was a dead-end. eg. PS1 wouldn't have scaled up to PS2 power, so PS2 needed to be radically different. PS2 was a bonkers solution so PS3 needed to be way different. PS3 was a bit of a cock-up so PS4 needed to be completely different to PS3. Each won had to divorce itself from platform legacy in order to progress. With a good architecture that works and scales, you can just scale the hardware and keep compatilbity, in theory.

You can always introduce a new IO system to an existing architecture and provide both classic play and new. And before everyone says that never gets adoption, Kinect got crazy adoption. Had it been an input system with any legs, it'd have carried on. PS5's 'Ubercontrol' system could work on Neo, and be a cross-gen interface - doesn't require a new gen, although having a new gen helps wtih marketing.
 
When you have new I/O (like a new controller or SD->HD), then you need a new console. When you don't have that you can do the iterative approach (unless you screw up on the SW/HW front).

As Shifty mentioned, you can introduce new I/O interfaces without uberly redesigning a previous architecture. You can phase out the older models when their life cycle is up... then continue on.
 
All you said can be documented / stated for every mid-cycle refresh. It's no different than iOS/android (software and hardware) on keeping apps/games developers aware on how each mid-cycle refresh effects their respective wares. IMHO, I hope generational (let's start all over) goes away... and makes Sony, Nintendo and Microsoft compete more at the userbase level (crafting better libraries, apps, features, etc...), than always relying on restarts. The mid-cycle refreshes will just help push along the needed ideals on keeping and growing the current userbases.
The fact that it happens on iOS and Androids doesnt make it right.
 
I'd add that the size of the cellphone market kinda lends itself to yearly upgrades, even if I don't buy a phone every year there are enough people who do or who are due an upgrade from a two year old phone to justify it.
So what? What happens in one market isnt necessarily a rule for the other market or should necessarily be applied for another.
 
I'd add that the size of the cellphone market kinda lends itself to yearly upgrades, even if I don't buy a phone every year there are enough people who do or who are due an upgrade from a two year old phone to justify it.
Hmm I think android's yearly upgrade is not as common as ios devices.

Because in Android developer panel, the majority of phones still in Android 4
 
A typical "generational leap" has classically been 10x-8x mote powerful overall to its predecessor. That was also the case with ps4bone and yet there were only a couple games that looked substantially better in the first year. There were months of talks and laments about diminishing returns and such. Only now are AAA devs starting to consistently deliver "true next-gen" games. 2018 seems early for a ps5 both from the standpoint of having hardware available to offer a true next-gen leap at a reasonable price, and from the standpoint of devs being ready to put it to use in convincing ways to the average consumer.
 
A typical "generational leap" has classically been 10x-8x mote powerful overall to its predecessor. That was also the case with ps4bone and yet there were only a couple games that looked substantially better in the first year. There were months of talks and laments about diminishing returns and such. Only now are AAA devs starting to consistently deliver "true next-gen" games. 2018 seems early for a ps5 both from the standpoint of having hardware available to offer a true next-gen leap at a reasonable price, and from the standpoint of devs being ready to put it to use in convincing ways to the average consumer.

I think this is correct, although VR potentially could change things a bit. We saw motion controls disrupt things too. Although VR does require a substantial amount of power to work correctly things could shift around somewhat if VR takes off or some other technology disrupts things.
 
A typical "generational leap" has classically been 10x-8x mote powerful overall to its predecessor.

I think a new generation is more defined by the deprecation of the predecessor. Regardless of power, if Sony sticks to their plan then it's not really a new generation. If we see a stratification of software then we have two generations.
 
Shu Yoshida said Neo is still a PS4. It's the same generation, and he repeated it will not shorten the generation. And again saying it's a high end version of a PS4.

So I'm wondering if PS5 is going to "reset" into a next gen, while hopefully still having full BC. Then a PS5 Neo mid-gen again. Some features would need a reset at some point, otherwise they cannot change them if FC is always mandatory and overlapping. For example, if PS5 allows high speed asset streaming with SSD, and Neo cannot (mechanica HDD), that means no PS5 game can be designed to rely on it, or they must make two different engines?
 
It seems the most likely ps5 is gonna be be sort of a reset. Substantially more powerful, yet not only fully bc, but also probably titles that were usually cross-gen on past generational transitions are gonna be a foward-compat ps4/5 with more substantial enhancements than what we will see between ps4-neo, perhaps with exclusive features to entice the migration. All of this is mere speculation though.
 
Shu Yoshida said Neo is still a PS4. It's the same generation, and he repeated it will not shorten the generation. And again saying it's a high end version of a PS4.

So I'm wondering if PS5 is going to "reset" into a next gen, while hopefully still having full BC. Then a PS5 Neo mid-gen again. Some features would need a reset at some point, otherwise they cannot change them if FC is always mandatory and overlapping. For example, if PS5 allows high speed asset streaming with SSD, and Neo cannot (mechanica HDD), that means no PS5 game can be designed to rely on it, or they must make two different engines?

They kind of have to break FC at some point. It's going to be really interesting to see how they handle a sliding window of support for these mid-gen upgrades, though. For example: If Scorpio and Neo become the baseline spec and PS4 and XBOne are deprecated, how do you handle re-defining the product family as Neo/Scorpio low-end and and Neo+/Scorpio+ high-end? Surely you can't move to PS5 and XBoxToo and cut off support for both prior SKUs, asking Neo/Scorpio buyers to pay a premium and then giving them an even shorter useful life for their consoles.
 
Back
Top