The PlayStation 2 provided backward compatibility with the PS1 by essentially incorporating an entire PS1 on a separate chip. It kept this arrangement permanently.

The PS3 started off providing backward compatibility the PS2 by the same method. However, the cost-reduced slim PS3 in 2009, dropped this feature.

This difference has been nagging at me for a long time. To be clear, either pattern – keeping the backward compatibility chip permanently, or dropping it after a few years – can be justified. What I'm still curious about is the reason for the difference. Whatever the reason for one behavior, why was it different in the other case?

I was reminded of this when I stumbled across an old discussion on the matter: https://web.archive.org/web/20090228051852/http://playstation.joystiq.com/2007/02/08/playstation-3-manufacturing-costs-to-go-down-way-down

The removal of EE+GS would probably be 50 bucks

Now that's just someone guessing, and surely he must be way over the mark. EE+GS were first released in 2000. This is 2009 we are talking about. The combined chip cannot possibly cost $50. I am very surprised that it was enough to be considered worth dropping. I would expect by that point it could just be fused with something else.

Just how much did the combined EE+GS chip cost, just before it was discontinued?

  • 1
    The production costs of the chip may have been small, but the licencing fees for the design may have been fixed. This something that became a big problem for Sony with the PlayStation 3 and it's Nvidia designed GPU, and why the PS3 had very little overlap with its successor during its lifetime compared to its predecessors. The PS2's lifetime almost overlapped with the PS4's. A similar problem happened with the original Xbox, where Nvidia stopped making GPUs for it months before the Xbox 360 even came out.
    – user722
    Feb 1 '20 at 5:28
  • @RossRidge Good point! I believe the PS2 chips were a joint project with Toshiba. Did they have to pay Toshiba for each chip manufactured?
    – rwallace
    Feb 1 '20 at 9:39
  • 2
    Sorry, I don't know. The problems Microsoft and Sony had with Nvidia were well known, and I'm just speculating based on that. There would've been fees for the MIPS core, and probably patent license fees for the GPU, but I have no idea how they were structured.
    – user722
    Feb 1 '20 at 9:47

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.