Frankly I don't care if Iwata told me the ps2 was stronger than gamecube, it's simply not true.
Not a great start in a technical forum thread ...
Resi 4 is a concrete example, and the decrease in complexity in the ps2 version is hugely noticeable. No one has any proof whatsoever that burnout, or any other game that doesn't exist on the gamecube would not run on it. That developer said it had abysmal triangle throughput, that can be debunked immediately.
I don't think you're in any position to debunk ERP's hands on experience with the system. Unless you have an experienced multiplatform developer source talking specifics, or leaked internal dev studio engine benchmarks, or you're a highly experienced developer adept a range of low level optimisations on a wide range of different hardware.
Lets start with: how did
you get past the culling performance hit on GC's hardwired T&L unit? Although ... given that you think "culling a z-buffer takes 16 MB" I'm guessing you haven't come up with a better solution than ERP.
And way to insult me saying implying I can't deal with being wrong.
*looks up*
*looks down*
So that developer had a hard time with gamecube, well that's one developer against concrete results showing otherwise, when resi 4's brought up the dev even says he's just sharing his own experiences.
No, he didn't have a hard time with the GC. He said GC is pretty straight forward, it's just that it didn't have much to give.
"Gamecube's Gekko is the most powerful general-purpose CPU ever in a console. The PowerPC alone is so much better and faster structurally that Gekko not only is much, much faster than the PS2's main CPU but every bit as fast as a 733 MHz Pentium," rebukes Eggebrecht. "Don't forget how extremely powerful the 256K second level cache in Gekko makes Gamecube. The size of a CPU's second level cache determines how fast general game-code runs. The PS2 has a tiny 16K of second level cache, and even the X-Box only has 128K." - Julian Eggebrecht, Factor 5
A Nintendo second party talking without having run anything on the Xbox, a year before the Xbox launched and before the GC launched. With years of hindsight, I've seen a number of developers state that Gekko was an excellent chip, but Xbox CPU was faster overall.
And given the chip was twice as big and 50% faster, I think the only person surprised by that would be you.
"The size of a CPU's second level cache determines how fast general game-code runs." ... well ... no. It's one factor amongst many, many other potentially much larger ones. So, no, basically. That's the kind of thing a none programmer would say.
Gekko
is much faster than PS2 main CPU, but PS2's strength lies in its vector units which are maths monsters and can far outperform Gekko in vectorised workloads.
What I care to discuss is the xbox and gamecube, the idea that ps2 was more powerful (overall) doesn't hold up with real world examples.
Burnout 3 and ERP's work are real examples.
Just as I don't see anything in those games that couldn't be done on the gamecube.
See, now this is a problem. We *know* the GC can't do what Doom 3 and Riddick did. We *know* it, it's a fundamental limitation of the hardware. You saying you can't see it doesn't change what the hardware can do.
We've even had JC say it can't do it - at which point you indicate that he knows less than you.
Not in the way those games do it, but as you say, approximated. Ps2 did not support per pixel dynamic lighting, while gamecube and wii can achieve it, though it wasn't as easy to do for the developer as on Xbox. Flashlights in games, as an example. See the silent hill games on ps2 vs. RE Darkside chronicles. Umbrella chronicles used per vertex lighting like the ps2. This is what I meant when I said per pixel.
You aren't talking about approximating here, you're talking about simply not doing something and taking the hit.
"Canned boss game studios"
Wow, you used a quote from a developer for whose game didn't even see the light of day, and whose studio was shut down, bravo.
Not only is that an insulting and fantastically weak argument, but earlier in the same post
you used a quote from Julian Eggebrecht.
While we ARE at it though, rebel strike is capable of even more polygons per second - 20 million as opposed to the 15 mil in Rogue squadron. It shouldn't be that hard a task to find an xbox game with as many polygons per second as Rogue Squadron if the Xbox is so much more powerful, and the gamecube has such terrible triangle throughput.
You just heard about a game hitting 30 million back in 2002. But you disregard that because [reasons].
Both Xbox and PS2 could hit higher poly counts than GC. But particularly on Xbox, high poly counts weren't necessarily the best way to achieve the best results.
No, you didn't think.