WiiGeePeeYou (Hollywood) what IS it ?

[ninzel]

Take this with a grain of salt since it is from Nintendo but Perrin seems pretty confident about the system's power.
At about 4:20, "it's a Ferrari under the hood."
Could I have come up with a more vague peice of evidence.
http://video.google.com/videoplay?docid=6133447032518637786&sourceid=docidfeed&hl=en

 

[Shifty Geezer]

Take this with a grain of salt since it is from Nintendo but Perrin seems pretty confident about the system's power.
At about 4:20, "it's a Ferrari under the hood."
Could I have come up with a more vague peice of evidence.
http://video.google.com/videoplay?docid=6133447032518637786&sourceid=docidfeed&hl=en

I refer you to Iwata saying Zelda already looks next-gen...

 

[Corwin_B]

now, what little of mp3 i've see until now demonstrates an upgrade in scene geometry complexity way above the dry 50% we could expect.

From what I gathered from various devs comments, what really hampered the NGC was the lack of RAM. Perhaps RAM limitation was the bottleneck in MP1/2, and as such corners were cut in scene complexity (I know polygons are not RAM-hungry, but you have to texture those).

Not saying there definitely isn't more to the Wii hardware than what was seen until now, but I'm not exactly getting my hopes up.

Nintendo said they invested a lot of money in the development of Hollywood and Broadway (I remember figures of over 1 billion dollars). There is no way (with them being very business-savvy) that they invested such huge sums of money and got fooled by IBM/ATI before ending with smaller and slightly faster version of the same chips. So there are 3 possibilities here :
- Nintendo is lying/misleading. For example, $1B could be the total amount of money invested in those lines of chips, counting development costs already accounted for in the GC (Flipper+Gekko)
- Hollywood and Broadway are, as many Nintendo fans hope, much more than overclocked GC chips, and their true power will be unleashed later (SDK update...). That could be the case, but I suspect a couple of titles could at least take advantage of some of this power. Right now, the best Wii games look good not from a technical standpoint, but rather from good artistic direction. Nintendo is taking a huge backlash on the technical side, and a couple of tech demos using the "true power" could really help.
- The chips in the final Wii are not those which were developed by ATI/IBM. Which leads us to...

My tinfoil hat theory is that development on the Wii first started with ambitious graphics/low power in mind, and work began on Hollywood/Broadway in this direction, coupled with experiments in innovative controls (which ultimately ended with the Wiimote), and forecasts for small/medium losses on HW at launch (classical razor blades sales model). But the immense success of the DS, coupled with the early launch of the 360 led Nintendo to rethink their approach. They saw how Backward Compatibility and small games could help a console (nice reception to the Zelda bonus discs on GC, GBA BC helped tremendously the early days of the DS in the US...), and how offering a "Virtual Console" and full NGC backward compatibility could help fill their release schedule (biggest problem of the NGC and N64) and their launch lineup (Zelda TP, I'm looking at you).

But with their power-saving approach and technical limitations, they couldn't include Flipper+Gekko in the Wii form factor in addition to more powerful Hollywood+Broadway, so they decided to go with upgraded/overclocked NCG chips instead, and rely on the controller (which is what worked so well for the DS). That would also allow them to keep costs in check.

 

[ninzel]

I refer you to Iwata saying Zelda already looks next-gen...

I reference you to the fact that Iwata is not Perrin Kaplan.

 

[fearsomepirate]

Wow! I'm glad I know now that in all those games where you play the game zoomed in so close on fire trails that you can see the individual pixels, Wii is really lacking. All this time I was looking at images and thinking "These are OK, I guess," but that was because I forgot that most games are played as extreme close up zooms on particles. "OK, I guess" has just moved down to "Sucks." I really dodged a bullet there.

I don't think Flipper could do the reflections we've already seen in Excite Truck at any speed.

 

[theafu]

I'd say developer apathy hampered the GC more than the ram did. Either way it's not an overclocking. Let's be real how many of us here OC'ed our PC CPU's and GPU... When was the last time you saw a benchmark where someone bumps the speed up about 50% and suddenly has 2x-3x the performance? or let's ignore a story matt put up stating the CPU Line I keep mentioning that suddenly got pulled. I've never seen an argument against Wii that contradicted what nintendo has said about the system or developers actually making games.

 

[darkblu]

Corwin_B, believe it or not i had the same theory as yours until not long ago.

i used to think that nintendo changed the course of their next console design somewhere half way in the product development cycle - that they realised they'd were being dragged into the vicious ms game called 'let's see who has more money to lose', and where hw efficiency was of little concern, as long as you ended up with a market share. i thought ninty decided they were not playing it and went for reiventing the rules. i was taking that theory as a very viable one, and i literary expected a moderately up-clocked cube, if upped at all.

up until i saw the e3'06 metroid trailers.

those trailers had the geometric scene complexity of a contemporary pc FPS title. no fancy shaders and over-the-top textbook fx, but the geometry alone was there neverthless. and what was more - the vertex lighting demonstrated was way heavier than what i had seen from retro before - every light source affected considerably more geometry than any of the previous metroids did. overall, i estimated the TnL work in the mp3 trailers as way higher. so i though 'wait a minute, either retro just figured out how to make use of the cube's TnL unit, and the rest improvements come from the increased texture fidelity, or what's running those is not a cube under disguise.' then the natural question followed 'what could be the chances that retro have been underdelivering all this time?' - 'not big'. of course, they never showed lightplay at the level of, say, RE4, but the latter never achieved the metroid series' rock-steady framerate, either. so unless retro decided this time around to shoot for unstable 30's, i could not see what could have given them the performance advantage allowing them to demonstrate what they were demostrating at e3. nothing but an advancement in the hw, that is.

so, that's my take on hollywood - something that started from a flipper, but is sufficiently evolved to allow things absolutely impossible on the cube. or at least, not demonstrated by anybody there (which would mean that flipper was one severely underutilised part, and i'd be only glad so see somebody finally achieving such things on it as recently demonstrated by retro ; )

 

[Teasy]

But with their power-saving approach and technical limitations, they couldn't include Flipper+Gekko in the Wii form factor in addition to more powerful Hollywood+Broadway, so they decided to go with upgraded/overclocked NCG chips instead, and rely on the controller (which is what worked so well for the DS). That would also allow them to keep costs in check.

Power saving, heat ect wouldn't come into it though. Since you wouldn't need to use Gekko & Flipper at the same time as the main CPU/GPU (and vice versa).

 

[OtakingGX]

I agree with the comments about Metroid Prime 3. It's the best looking Wii game so far, and it demonstrates that the hardware is capable of so much more than that of the Gamecube.

Retro is definitely doing something special with the zoom effect in the cinematics. And I almost wish the game were a third person adventure game, since Samus' suit looks so good in the brief glimpses you get of it.

It's a shame that the game was pushed back until early next year. I'm hopeful that means there will be online support.

 

[Megadrive1988]

Corwin_B, believe it or not i had the same theory as yours until not long ago.

i used to think that nintendo changed the course of their next console design somewhere half way in the product development cycle - that they realised they'd were being dragged into the vicious ms game called 'let's see who has more money to lose', and where hw efficiency was of little concern, as long as you ended up with a market share. i thought ninty decided they were not playing it and went for reiventing the rules. i was taking that theory as a very viable one, and i literary expected a moderately up-clocked cube, if upped at all.

up until i saw the e3'06 metroid trailers.

those trailers had the geometric scene complexity of a contemporary pc FPS title. no fancy shaders and over-the-top textbook fx, but the geometry alone was there neverthless. and what was more - the vertex lighting demonstrated was way heavier than what i had seen from retro before - every light source affected considerably more geometry than any of the previous metroids did. overall, i estimated the TnL work in the mp3 trailers as way higher. so i though 'wait a minute, either retro just figured out how to make use of the cube's TnL unit, and the rest improvements come from the increased texture fidelity, or what's running those is not a cube under disguise.' then the natural question followed 'what could be the chances that retro have been underdelivering all this time?' - 'not big'. of course, they never showed lightplay at the level of, say, RE4, but the latter never achieved the metroid series' rock-steady framerate, either. so unless retro decided this time around to shoot for unstable 30's, i could not see what could have given them the performance advantage allowing them to demonstrate what they were demostrating at e3. nothing but an advancement in the hw, that is.

so, that's my take on hollywood - something that started from a flipper, but is sufficiently evolved to allow things absolutely impossible on the cube. or at least, not demonstrated by anybody there (which would mean that flipper was one severely underutilised part, and i'd be only glad so see somebody finally achieving such things on it as recently demonstrated by retro ; )

if you're right, then Hollywood's equivalent of Flipper's "XF" block
(the geometry processor / T&L unit) is seriously upgraded, or replaced by some powerful vertex shader, or super geometry engine

 

[Fox5]

Corwin_B, believe it or not i had the same theory as yours until not long ago.

i used to think that nintendo changed the course of their next console design somewhere half way in the product development cycle - that they realised they'd were being dragged into the vicious ms game called 'let's see who has more money to lose', and where hw efficiency was of little concern, as long as you ended up with a market share. i thought ninty decided they were not playing it and went for reiventing the rules. i was taking that theory as a very viable one, and i literary expected a moderately up-clocked cube, if upped at all.

up until i saw the e3'06 metroid trailers.

those trailers had the geometric scene complexity of a contemporary pc FPS title. no fancy shaders and over-the-top textbook fx, but the geometry alone was there neverthless. and what was more - the vertex lighting demonstrated was way heavier than what i had seen from retro before - every light source affected considerably more geometry than any of the previous metroids did. overall, i estimated the TnL work in the mp3 trailers as way higher. so i though 'wait a minute, either retro just figured out how to make use of the cube's TnL unit, and the rest improvements come from the increased texture fidelity, or what's running those is not a cube under disguise.' then the natural question followed 'what could be the chances that retro have been underdelivering all this time?' - 'not big'. of course, they never showed lightplay at the level of, say, RE4, but the latter never achieved the metroid series' rock-steady framerate, either. so unless retro decided this time around to shoot for unstable 30's, i could not see what could have given them the performance advantage allowing them to demonstrate what they were demostrating at e3. nothing but an advancement in the hw, that is.

so, that's my take on hollywood - something that started from a flipper, but is sufficiently evolved to allow things absolutely impossible on the cube. or at least, not demonstrated by anybody there (which would mean that flipper was one severely underutilised part, and i'd be only glad so see somebody finally achieving such things on it as recently demonstrated by retro ; )

Err, current game polygon counts haven't increased much since the gamecube was launched, they've mainly added shaders. Not too impressive if it can keep up those polygon counts, and the lighting could just mean they've moved away from fixed function or considerably beefer up the t&l abilities.

 

[Refreshment]

Anyhow, the $250 price tag is a complete turn off. The Virtual Console may be a plus for some... but I already own a GCN, N64, SNES, GB, and NES. Why would I re-buy games I already own, especially on a budget console? So it all comes down to games: Do I want GCN+ level games (level = AI, Graphics, Sound, etc) and get a unique free hand controller, or do I spent $50 more and get a platform with 10x+ performance, significantly more game support, and a proven online network?

While i understand your reasoning (and it is valid) i dont think the price reference is "fair".

The 300 dollars Xbox360 doesnt exist, but the 340/350 does. Leaving games aside, the memory card is needed to properly use the console and yet some of its functions remain gimped. So 50 more wont buy you a more advanced console.

Kind of makes you think that questionable MS strategy works some times

 

[darkblu]

Err, current game polygon counts haven't increased much since the gamecube was launched

sorry, Fox, but my eyes tell me otherwise. we're talking of mp1-2 vs mp3, in case you missed that, and the increase there is multifold - 2x-4x times at least. i have not tried to count the polies but i know from experience what it takes for the human eye to start noticing better curvatures and details in similar models. and yes, the metroids share similar models and art concepts.

they've mainly added shaders.

how could you tell that? or you know something we don't?

Not too impressive if it can keep up those polygon counts

not too imressive in the context of what? gamecube? when was the last time you saw mp1-2?

and the lighting could just mean they've moved away from fixed function or considerably beefer up the t&l abilities.

i don't see why they should have moved away from fixed function. i'm observing a significant increase in scene complexity, but i can't tell one zilch what unit is doing that (they could use some sw assist just as well), and i don't think anybody can tell the presence of vertex shaders just by looking at footage. at least not from such one like the demonstrated in the mp3 trailers.

 

[Fox5]

how could you tell that? or you know something we don't?

I'm referring to the poly difference we've seen in PC games in that time. In 5 years, we've gone maybe 2x to 3x increase in poly count, but shader complexity has increased far more than that. Didn't mean to say the Wii has increased over gamecube, just that the focus shifted away from polys years ago and it's not very impressive that 2006 hardware can handle what everything else is doing, minus the primary visual effects games have focused on for the past few years.

 

[pc999]

The 300 dollars Xbox360 doesnt exist, but the 340/350 does. Leaving games aside, the memory card is needed to properly use the console and yet some of its functions remain gimped. So 50 more wont buy you a more advanced console.

Taking in acount that MS is probably getting a few hundred% of proffit with each memory card the cost of the console (if they included it on the console) would probabily about 310-315$.

Anyway you can think that once they are selling the GC with proffit ata lot less than 100$ (I can find it at 65euro) and by Moores Law they could have at least a CPU of 84M transistores (for reference it is a lot more than a 970FX) and a GPU of 204M transsistores (a lot more than a X1600/X800) and keep it at the same price, unless the flash ram (certainly cost them less than 20euros as even I can find at a lower price)+ the wireless+ the controler cost more than 150$ (very hard to belive IMO) then it is very probably overpriced even with 4x the silicon on it, a but really overpriced if it is just a for what they show.

 

[darkblu]

I'm referring to the poly difference we've seen in PC games in that time. In 5 years, we've gone maybe 2x to 3x increase in poly count, but shader complexity has increased far more than that. Didn't mean to say the Wii has increased over gamecube, just that the focus shifted away from polys years ago and it's not very impressive that 2006 hardware can handle what everything else is doing, minus the primary visual effects games have focused on for the past few years.

ah ok, you were talking about the general advancements in the industry, i misread your post, my bad.

btw, believe it or not, but the 'primary visual effect' in the industry for the past years remains, low and behold - textures. in all varieties, skillfully used in multiples. only now we are starting to see more and more emphasis on realistic dynamic light contributions, manly through the advent of HDR. but i'm still waiting for the day when lighting will be the predominant artistic visual means, and we'll have a single normal map, at most, possibly with a diffuse map.

btw, geometry never stopped to matter. and it never will. other things on top of it may come and go, but don't expect to see voxels anytime soon ; )

 

[Ooh-videogames]

It's not odd if the prime concern was hardware BC with GC, although that decision might be considered odd if it sets Wii a few years behind the curve!

That would be...peculiar marketting. 'Let's create hardware that can do this, but won't show it for the first year. Let them see worse games to begin with and yet not tell them that Wii can do better. That's bound to get us sales!' I don't see the sense in that. Tell people what your system can do, and give them every reason to think well of it. Do everything you can to break down reservations for buying it. If someone wants good graphics, and your system can do good graphics, shout about it. Don't hide that for a year to surprise them! What you're suggesting would be the same as releasing screenshots of a game without any AA or texture filtering, and then when the game's released, add AA and filtering. It'd be a nice surprise for those who bought the game, but by then I expect most people will have stopped paying attention to the jaggie mess you were showing earlier.

To me, the reason for secrecy is because you know comparatively those that talk about such things will make a lot of noise about your less impressive technical details. If everyone else is using FP16 HDR, and you can't use HDR, that'd be reason to keep quiet and leave people guessing. Whereas if you have FP16 HDR but don't say so when present titles make it look like your system doesn't support it, keeping that info secret only hurts your image (to that 0.1% who care about such things )

Incidentally, that patent talked of YUV colour modes. I'd be very interested if that was proper NAO32 like colour-space supported in hardware. That'd be a real boon for the system IMO, and a first for the industry?

IMO, MP3, Excite Truck, Red Steel, SMG, all look great. All of those titles development started on OC' GCs. So to judge them as what Wii is capable of is immature. It was easier to know what 360 is capable of, because we have all the specs.

Launch title PD0 can't be classified as a accurate reprensentation of what the 360 is capable of, so these launch games appearing on Wii shouldn't be as well.

Also what I meant is not knowing what to expect, none of the titles were built on PC hardware to spec. They all began life on OC' clocked GCs. So we want know until games are developed from the ground up on Wii hardware. Many here and abroad, that frequent these boards already have low expectations. Mainly because of leaked specs, comments by devs who had yet to have complete devkits, suggesting that the console could and likely retail for $150. Instead its retailing for a hundred dollars more.

 

[swaaye]

In my fabulous opinion, MS's $300 Xbox is just a marketing ploy to allow them to say they cover that price range. And Sony is doing the same thing. It was, however, stupid IMO because they fragmented their install base with a cripple console. This limits what devs can do. Either devs risk alienating buying customers (which would also backfire for mind share on the entire 360 product) or they don't develop with that hard drive in mind. Ridiculous when you realize it's just a crap 5400 RPM 20 gig notebook drive, it's 2006, and a 250 gig 3.5" 7200 RPM 16 meg cache drive goes for $70 on Newegg. And even worse when you realize there was no such limitation with the older console.

 

[Ty]

Launch title PD0 can't be classified as a accurate reprensentation of what the 360 is capable of, so these launch games appearing on Wii shouldn't be as well.

While I agree with this overall, I think the comparison isn't quite that accurate since it seems that Wii is far closer architectually to GCN than the 360 is to the XBox.

Mainly because of leaked specs, comments by devs who had yet to have complete devkits, suggesting that the console could and likely retail for $150. Instead its retailing for a hundred dollars more.

Well we also know they are making a profit from the hardware from day 1, unlike 360 and PS3 - so who knows exactly what the true cost of materials is for the Wii.

 

[Ooh-videogames]

While I agree with this overall, I think the comparison isn't quite that accurate since it seems that Wii is far closer architectually to GCN than the 360 is to the XBox.



Well we also know they are making a profit from the hardware from day 1, unlike 360 and PS3 - so who knows exactly what the true cost of materials is for the Wii.

Well, its obvious to me from playing my copy of PD0, that its a Xbox port with visual upgrades. A ports a port, architecture differences aside.

How much did GC cost to manufacture? I assume it wasn't much.