More than anything the question is what an AAA PS2 game made with modern software looks like.
What can your AAA studio do with 8.5GB worth of PS2 game, no patches, no updates ever?
Hey now - no patches is hardly an AAA game
But in all seriousness - a 'modern' retro console (with appropriate target specs - not just recycled off the shelf-GPU) is something I'd have liked to see too. The PlayDate is kind of the same concept, and the software for it has a charm that none of the modern consoles do. But not exactly mainstream appeal either of course.
Sony's had an updated
EE GS with 32MB eDram since 2000 but for obvious reasons they couldn't change PS2 HW.
Fixed that for you.
Anyway a 'PS2 Pro' was actually conceptualised in variety of prototypes. Some that even got made, sort of - like the PSX. And like the Pro's we eventually got over a decade later - they could have released something that would get bespoke 'pro enabled' versions of games, while still maintaining BC with the 'amateur'. Probably wasn't worth the hassle when already outselling all of competition combined by more than 3:1 though.
Cerny's points make no logical sense - PS1/2 are the obvious sweet spots for fastest game dev times at the lowest prices.
Cerny was referencing 'time to polygon' metric - not 'time to publish a 100M budget title'.
Sony could either buy or reverse engineer RenderWare and modernize it to make the world's best PS2 middleware engine. Making this PS2 engine available to developers for free would effectively be adding a 5% bonus to their earnings by making Unreal unnecessary while stopping games from becoming ubiquitous across x86 platforms. Sony owning PS2 development from the get-go would make it a platform built on PS2 exclusives where Unreal would never be able to catch up and syphon away games and profits.
Ok at this point I don't know if my sarcasm meter's off - but in the unlikely chance you're serious. Sony owns plenty of tech that could take that role if they wanted it to. But clearly it's not a strategy they chose to pursue. And given the historical precedent of companies that have failed to do this... repeatedly... over decades - I can't say I blame them really, it's a LOT easier said than done. Or rather - it's never been successfully done yet - period.