[steviep]

Originally Posted by Rangers

I have had a hunch or gut feeling Wii U wouldn't be even as powerful as PS360 practically all along.

Usually I would state I believed it would be ~30% more powerful, but I always had that hunch. My hunches are often right and later I wish I'd listened to them.

So yes, I believe the report. My thinking was that Nintendo is probably going to do something to horribly screw up the box related to bandwidth, so even if it's paper specs were better it would be crippled (I've thought this since DDR3 rumors).

I guess I just have a really really poor view of Nintendo engineering.

The Gamecube was a marvel of engineering in comparison to its competitors of that particular generation. The unified GDDR3 pool (be it 1GB or 1.5gb) will be helped out by the EDRam.

I'm not sure where the "weaker than PS360" crap is coming from, because at the very least the console has more ram and a much better CPU.

[ERP]

Originally Posted by steviep

The Gamecube was a marvel of engineering in comparison to its competitors of that particular generation.

Hmm wouldn't have been my assessment.
It was by far the worst performing of the 3 platforms. You should see the performance penalty when god forbid the GPU has to clip a polygon, it was so bad I actually wrote code to traverse triangle lists and clip tris with the CPU.

[ToTTenTranz]

Originally Posted by steviep

I'm not sure where the "weaker than PS360" crap is coming from, because at the very least the console has more ram and a much better CPU.

The other guys said the GPU was weaker, this one says it has more RAM and a better CPU.. so they're not really contradicting each other..

Originally Posted by ERP

Hmm wouldn't have been my assessment.
It was by far the worst performing of the 3 platforms.

Comparing Resident Evil 4 between Playstation 2 and Gamecube seems to suggest otherwise...

[ERP]

Originally Posted by ToTTenTranz

Comparing Resident Evil 4 between Playstation 2 and Gamecube seems to suggest otherwise...

perhaps but having actually written code for all 3 platforms in that generation that's my assessment.
You could get pretty good initial performance out of GC, but usually that was pretty much it, no amount of dicking around would get you more polygons or more pixels.
PS2 was a pain in the ass poor implementations were really bad.
Xbox was basically underutilized because no one could be bothered.

You might get prettier pixels out of a gamecube than a PS2, but usually only because you couldn't be bothered trying to figure out have to make the PS2 produce the better imagery.

Gamecube's "big win" was the processors relatively arge cache and the low latency main memory. I always felt it was designed that way as an over reaction to the horrible memory latency on N64 and all the developer bitching about it.
But the memory subsystem was over engineered, the large cache for the most part removed the advantage of the low latency memory...

[fearsomepirate]

Originally Posted by ERP

You might get prettier pixels out of a gamecube than a PS2, but usually only because you couldn't be bothered trying to figure out have to make the PS2 produce the better imagery.

I've got a pretty large library of games on both consoles, pretty much all the Cube exclusives, and many of the prettiest PS2 games, and there's always something off with even the best-looking PS2 games. Either the aliasing is out of control, or the textures are low-color, or the textures are low-res, or there's just not much geometry, or the water is a scrolling texture, or something. Some developers are really good at masking it, but there's always something.

That reminds me, I should finish Silent Hill 3.

[ToTTenTranz]

Originally Posted by ERP

perhaps but having actually written code for all 3 platforms in that generation that's my assessment.
You could get pretty good initial performance out of GC, but usually that was pretty much it, no amount of dicking around would get you more polygons or more pixels.
PS2 was a pain in the ass poor implementations were really bad.
Xbox was basically underutilized because no one could be bothered.

You might get prettier pixels out of a gamecube than a PS2, but usually only because you couldn't be bothered trying to figure out have to make the PS2 produce the better imagery.

Gamecube's "big win" was the processors relatively arge cache and the low latency main memory. I always felt it was designed that way as an over reaction to the horrible memory latency on N64 and all the developer bitching about it.
But the memory subsystem was over engineered, the large cache for the most part removed the advantage of the low latency memory...

What I take from your post is:

1 - PS2 has better paper specs than GC
2 - Developers were too lazy to take advantage from PS2's superiority
3 - GC shows better graphics than PS2
4 - GC is more developer-friendly (or at least lazy developer friendly)
5 - XBox had the best specs
6 - Developers were lazy with XBox

To be honest, there's nothing there that isn't already widely known..
I don't see anywhere that the PS2 is more capable than the GC in real life, and apparently there's no proof of it.
The best-looking GC games look way better than the best-looking PS2 ones, IMO.

[function]

I don't think you should take ERP's post as him calling anyone lazy. He's talking about his own has hands-on experience too so you can't just put his views down to him comparing paper specs like a typical forumite.

He's also not the only experienced developer I've seen place the PS2 above the GC in terms of overall capability.

[Brad Grenz]

Yeah, I mean the GC's advertised maximum polygon rate was something like 10 million a second, wasn't it? PS2 games could easily do 2-3 times that in the real world.

[fearsomepirate]

Isn't "overall capability" a little bit subjective when functionality isn't identical? Doing anything visually impressive on the PS2 required severely restricting the texture color depth, something the Cube wasn't limited by. And the best Cube developers did things with texel and pixel level effects that you just didn't see on the PS2. And yes, using 4-bit palletized textures does degrade image quality. So you had more fillrate on the PS2, you could do some neat things with the VUs, and you had a 32-bit frame buffer, but it also appears to have had some limits on the texturing side of things that the Cube didn't.

[fearsomepirate]

Isn't "overall capability" a little bit subjective when functionality isn't identical? Doing anything visually impressive on the PS2 required severely restricting the texture color depth, something the Cube wasn't limited by. And the best Cube developers did things with texel and pixel level effects that you just didn't see on the PS2. And yes, using 4-bit palletized textures does degrade image quality. So you had more fillrate on the PS2, you could do some neat things with the VUs, and you had a 32-bit frame buffer, but it also appears to have had some limits on the texturing side of things that the Cube didn't.

[ToTTenTranz]

Originally Posted by function

I don't think you should take ERP's post as him calling anyone lazy. He's talking about his own has hands-on experience too so you can't just put his views down to him comparing paper specs like a typical forumite.

He's also not the only experienced developer I've seen place the PS2 above the GC in terms of overall capability.

Ok, he said they "couldn't be bothered with". Perhaps he meant time constraints.

Nonetheless, if the PS2 is placed above the Gamecube, why do late Gamecube games look consistently better than late PS2 games?

If no developer ever figured out how to extract better visuals from PS2 than from GC, doesn't that mean that the GC is, in practical terms, the most powerful solution?

[ERP]

Originally Posted by ToTTenTranz

Ok, he said they "couldn't be bothered with". Perhaps he meant time constraints.

Nonetheless, if the PS2 is placed above the Gamecube, why do late Gamecube games look consistently better than late PS2 games?

If no developer ever figured out how to extract better visuals from PS2 than from GC, doesn't that mean that the GC is, in practical terms, the most powerful solution?

FWIW I don't ever consider any developer who ships anything lazy.
When you're building a cross platform game, there is always an element of lowest common denominator, it's about costs (and I don't just mean financial).
PS2 was often the "lead SKU" at big publishers because of the installed base, Xbox was a version you had to do, in most cases you could write a simple version of your renderer and just drop the assets on Xbox and they would usually run faster. So you'd increase texture quality and call it done.
Usually when you dropped it on gamecube it would run slower and you'd have no memory left, so you downsample to make things fit, figure out how you could use ARAM without crippling performance and ship it.

If you wrote an XBox exclusive with no intention of ever shipping on PC, and you actually spent time optimizing there was a lot of performance to be had, usually most titles were CPU limited because then the polygon indices had to be copied into the GPU ring buffer (which wasn't actually a ring buffer). If your app was pushing a lot of geometry it could literally spend 60% of it's time doing nothing but linear memory copies.
It was possible to place jumps into the ringbuffer, to effectively "call" static GPU buffers, but it was tricky to get right because of the pipeline and the fact you had to patch the return address as a jump into the buffer so you'd have to place fences between calls to the same static buffer.
If you did this however you could trivially saturate the GPU and produce something much better looking.

On GameCube the biggest issue is it was just had pathetic triangle throughput, the 10M polygons per second (I don't remember the real number) assumes you never clip or light anything.
GameCube was DX7 class hardware for the most part, albeit a more fully featured version than ever shipped in a PC. The GPU just wasn't very fast.
As I said it's real benefit was the memory architecture and I still feel it was over engineered.
On the whole it wasn't a bad machine, but I wouldn't have said it was "more powerful than PS2)

[function]

Originally Posted by ERP

PS2 was often the "lead SKU" at big publishers because of the installed base, Xbox was a version you had to do, in most cases you could write a simple version of your renderer and just drop the assets on Xbox and they would usually run faster. So you'd increase texture quality and call it done.
Usually when you dropped it on gamecube it would run slower and you'd have no memory left, so you downsample to make things fit, figure out how you could use ARAM without crippling performance and ship it.

Would it be fair to say that the PS2 having more main memory than the GC somewhat balanced out the lack of GPU support for S3TC texture compression?

[nomad]

Hi all,

http://www.segatech.com/gamecube/overview/

1 vertex color + 1 light + 1 texture20M polygons/sec
no vertex color + 1 texture26.4M polygons/sec
1 vertex color + no texture (gouraud shading) 32M polygons/sec

Nintendo said 6-12 Million polygons/sec real game.

[Grall]

Originally Posted by ERP

On GameCube the biggest issue is it was just had pathetic triangle throughput

Yet, the original Metroid Prime shows some incredibly detailed scenes for its time, running at 60Hz, and with soft-skinned - ragdoll, even! - enemies jumping about. Particularly the first world has a couple chambers that are just amazingly complex visually. Retro sure sunk a lot of effort into modelling all that stuff.

[Teasy]

Originally Posted by Grall

Yet, the original Metroid Prime shows some incredibly detailed scenes for its time, running at 60Hz, and with soft-skinned - ragdoll, even! - enemies jumping about. Particularly the first world has a couple chambers that are just amazingly complex visually. Retro sure sunk a lot of effort into modelling all that stuff.

That's because its incorrect (the pathetic 10m triangle throughput comment). I'm also surprised at the claim that you just couldn't get anything more out of GC with extra effort, stand out titles from Capcom and especially Factor 5 would completely disagree with that conclusion.

To be honest it seems from reading those comments that these conclusions are being drawn from developing PS2 games and then porting them over to GC, but surely that's no way to compare two systems with different architectures. I mean there are even PS2 to XBox ports that looked worse on XBox (surely nobody could argue with the fact that XBox was substantially more powerful than PS2?) because of the extent the developer went to to get the best out of PS2's exotic design and lack of time spent modifying it for XBox's own hardware. Yet if you look at most third party games graphically XBox was usually first, GC second and PS2 third even though those games usually had PS2 as the lead platform.

When it comes to games developed exclusively for GC they outclass the best PS2 offerings, and the few titles we saw developed for GC and ported to PS2 showed more than a small decrease in quality (Resident Evil 4 for example).

[Teasy]

Originally Posted by Brad Grenz

Yeah, I mean the GC's advertised maximum polygon rate was something like 10 million a second, wasn't it? PS2 games could easily do 2-3 times that in the real world.

GC's max polygon rate was 32 million for the GPU and of course the CPU could chip in for more. PS2 could in no way easily do 30m polys per second, though it could maybe just about do that if you focused on geometry at the cost of everything else.

[Sigfried1977]

Originally Posted by Grall

Yet, the original Metroid Prime shows some incredibly detailed scenes for its time, running at 60Hz, and with soft-skinned - ragdoll, even! - enemies jumping about. Particularly the first world has a couple chambers that are just amazingly complex visually. Retro sure sunk a lot of effort into modelling all that stuff.

Yeah, what a gorgeous game. I still have it, but I'm sort of afraid to fire it up again. Don't want to sully those precious memories with the harsh reality of what is without a doubt clearly outdated technology by today's standards. (also: the RGB signal quality of the GC was atrocious, so it'll look a lot worse than it could regardless)

[Grall]

Originally Posted by Sigfried1977

I still have it, but I'm sort of afraid to fire it up again. Don't want to sully those precious memories with the harsh reality of what is without a doubt clearly outdated technology by today's standards.

True, true... I have MP and MP2 for my GC still, even though I no longer own the cube itself. I sold all my other Cube games years and years ago to help pay for my PS3 which now stands cold and dead on my apartment floor. Bet that cube still works, if someone's still using it...

also: the RGB signal quality of the GC was atrocious, so it'll look a lot worse than it could regardless

Perhaps so, I last played the game on my Wii using component video, and that looked pretty sharp on the whole, considering it's only SD resolution of course. Actually bought MP Trilogy for the Wii, tried playing it but couldn't get through the first barrier (!) because I couldn't figure out how to activate the scan visor using the Wiimote...! (I'm a man - I don't need to read any manuals..........)

[Sigfried1977]

Component looks much better on the GC as well. Unfortunately the required cable never made its way over to Europe (at least not officially), and the amount of money people are charging for it on ebay is completely nuts. (like 100€+)

[Prophecy2k]

Are there any GC emulators for PC like the Wii ones, that can upscale everything? Just wondering really

[SedentaryJourney]

Originally Posted by Teasy

GC's max polygon rate was 32 million for the GPU and of course the CPU could chip in for more. PS2 could in no way easily do 30m polys per second, though it could maybe just about do that if you focused on geometry at the cost of everything else.

In which case the Graphics Synthesizer could render 75 million per second with 1 vertex colour. EE could transform 66 million polygons per second. So apples to apples you're looking at twice the geometry performance on PS2.

BTW who set the wayback machine to 2001?

[almighty]

Originally Posted by Prophecy2k

Are there any GC emulators for PC like the Wii ones, that can upscale everything? Just wondering really

Yep, I play the resident evil remakes rendered at native 1080p with 4xSSAA.

[Rodéric]

Nintendo numbers are conservative real world performance.
Sony's are ideal peak performance.

ie: not comparable.

[SedentaryJourney]

Originally Posted by Rodéric

Nintendo numbers are conservative real world performance.
Sony's are ideal peak performance.

ie: not comparable.

According to nomad's post, 32 million is for 1 vertex colour and no texture.