By using this site, you agree to our Privacy Policy and our Terms of Use. Close
padib said:

 

Hi Torok, I was going to write this on your wall but I think I want to reply here.

I think you brought so far the only valid point against the unified API.. The clock cycles that will obviously be needed when translating the API commands to hardware commands will cause an overhead. It will create a performance difference similar to Java versus C++.

However, I believe the risk of it drastically affecting performance is low. Given the Java vs C++ example above, and the newest iphone and ipad's recently shown ability to play advanced graphics, it's my expectation that this overhead will be real but not drastic, e.g. tolerable. Finally, taking into account Nintendo placing graphics as a secondary importance over game content, I think Nintendo will pursue the solution while mitigating this real issue.

That's my thoughts on it personally. Also, I believe Nintendo was firm about their announcement about the shared API and I don't see it as a debatable outcome, despite what many are arguing in this thread.

For that reason I expect this new device to launch soon, to be the first of the family of brother devices, and it will be a force to be reckoned with in the new generation, I think Sony and MS do well to pay attention to that force.

Let's remember, Sony, nay not one competitor was able to topple Nintendo in the handheld space. Translating the success from the handheld space onto all their devices is menacing to the twins.


Hi, padib! I wouldn't say it wouldn't be not even close to as bad as Java vs C++ (since Java is much slower because it isn't pre-compiled). I'm talking about something more like Unity: you write one code and it exports native code for several platforms. This way you have native code, pretty fast, and just wrote it once.

The only overhead I can see is that if the idea is: write one code and the SDK will simply export it for the home console and automaticaly export a downgraded port for the handheld, we would have a more high-level abstract code. It would potentially cut the punch of the console at half from what we know it. Basically, let's say that a console like that with a GTX660 wouldn't have advantages compared to a normal low-level console with a GTX650 that would normally give you half the fps a 660 would.

I believe we need to wait for more details about what a single API means. It could mean:

- Handheld and home console will have SDKs with the same programming language and similar APIs, easier to learn, easier to port.

- Both will have an Unity-like SDK that exports automatically the code for them. Developer will have to downgrade the game to run on the handheld and porting would be optional.

- The same as above but with automatic porting. It's a complex thing to do and the result won't be that optimized.

iPhone and iPad are good examples, but pay attention that a lot of Vita games have better visuals, even if the current iDevices use a way more powerful hardware. The last porting of Bioshock really showed a lot of weakness on these devices.