Nintendo Wii Matches the Power of the Original Xbox
By William on Friday, June 15th, 2007 at 12:41 AM PST In Nintendo
According to a FIFA producer, the Nintendo Wii matches up to the original Xbox as far as power goes. Of course the Wii is an easy system to develop code for, but matching up to the power of a last generation system isn’t exactly a big deal in my opinion. People need to keep in mind that the Wii is great simply because it is original. That originality could wear off one day, but certainly doesn’t look like it’s slowing down any time soon. I honestly wish people would stop comparing the Wii’s actual system details to Sony and Microsoft’ systems. Honestly, the Wii cannot compete on a hardware level and Nintendo knows it.   The point of the article is that the Wii is pretty much what the original Xbox was. Is that really the point we need to care about? Probably not.
FIFA producer Tim Tschirner has said that he believes the Nintendo Wii matches up to the first  in terms of power.
Tschirner’s comments came in an exclusive interview with GamesIndustry.biz’s sister site, Eurogamer.net
“It’s about as powerful as the original Xbox,” he said. “The video hardware unfortunately is not as powerful. There’s just a couple of key things that you can do on Xbox like shaders which you just cannot do on the Wii… Overall though it’s pretty much what the original Xbox was.”
@xboxmodder = “why look at clock numbers when the arcitexture clearly proves you wrong”
At what point did I use clock numbers to claim anything? Did I bring up the fact that Xbox’s cpu was clocked nearly twice as fast? Did I bring up clock frequencies of ram? Nope.
You’re using typical responses designed to counter people using generic figures such as those, in order to defend the Gamecube’s honor. I used it’s architecture to imply that it does nothing to “blow the Xbox out of the water”.
If this were a situation where Gamecube and Xbox were relatively equal on paper, only Gamecube was customized, and XBox was “off the shelf”, it would be easy to say Gamecube > Xbox. But that is NOT the case here.
I even said specifically, that overall, Gamecube is more efficient, but that that efficiency is what allowed it to keep pace, not overtake Xbox.
The architecture “clearly” proves me “right”, in that you’re numbers out of context, and there’s nothing you can say to prove otherwise.
And it matters not to me, which cpu was more powerful. I did say, Xbox cpu + two programmable vertex shaders > Gamecube’s cpu + generic T&L. That is reality.
(and keep in mind, that I’m not disagreeing with Wiiboy, for the sake of fanboyism. He simply made statements that I find to be incorrect, from simply checking up on what he claimed. (in a very condescending, fanboyish way)
I’m not interested in pumping up one company over another. If I had a choice, the Wii would have been powerful enough to run modern game engines as is, but targeted at ntsc. They could have gotten all the ports from the other systems, and still had it’s own unique games as well. I would likely have been happy with JUST that, for the remainder of this generation, as I don’t mind the standard tv at the moment.
As to Wiiboy, every gpu and their momma does indeed use HSR. None of them draw textures and effects to all pixels. Xbox had hardware occlusion query, occlusion culling, compression, etc.. Not as if it didn’t.
All of your posts are hyperbole, designed to worship the capabilities of the Gamecube and Wii. I care not for any of these companies. You assume anyone that doesn’t feel the same way about the Gamecube and Wii, to be fanboys positioned on the opposite side of the isle. (If that were the case, the entire gaming industry is on the same side, including Nintendo to some extent)
@the nerd comments. This is no different than a typical baseball or football discussion. You don’t have to be a loser to wonder and investigate how things work. Especially if you were looking into a job that might employ such things.
And as to fill-rate and texture rate, etc.. Most developers I have seen comment, the one I know personally, (whom only worked on Gamecube directly, but had all the internal Xbox docs, and was in a position to comment) have said the same thing. And there are plently of developer comments that mirror that. The most you will get, is Gamecube = Xbox, and that will be a very specific narrow circumstance. Not an overall statement. And of course you’ll find a few things that Gamecube could do better.
But overall, developer consensus is Xbox > or = Gamecube. With most just going with the >. On paper, in practice, in theory, etc.. And none of completely “out of context” specs quoting will change that.
There’s nothing wrong with that. Nintendo competed well in all areas. Gamecube was a fine system. (for last generation hardware)
And to add to the response to Xboxmodder, if I were using straight clock frequencies to determine fill-rate, I would have said 933 million pixels per second, for 1 or 2 texture layers. But I didn’t. I used Wiiboy’s own quotes, and benchmarks I’ve seen posted, to get the 350-750 million number I listed for Xbox. Not theoretical maximums. I just assumed full, theoretical maximums for Gamecube.
From Erp, known developer on B3d.
http://forum.beyond3d.com/archive/index.php/t-2168.html
Keep scrollong through the thread. I would mark it up with Google’s cache, but it won’t let me post the link here, and I don’t think this place allows html.
Plenty of other known developers that imply the same things, (as I’ve already done in this thread) They all come to the same conclusions regarding Xbox and Gamecube.
And just a side note: I would assume the Fifa developer in the original topic of this thread, described Wii’s video hardware as, “not being as powerful as Xbox”, in regards to capability, rather than performance.
Sort of like saying, Photoshop is a “powerful” tool in editing photos. Or 3dmaya is a powerful modeling tool, etc.. He’s just using “powerful” in the context of what it can and cannot do, rather than actual performance with what it’s capable of.
I doubt he means the Xbox could outperform the Wii, at something made specifically within the parameters of the Wii hardware.
So, don’t imply that I have said Xbox = or > Wii. I “clearly” agree that it isn’t. Just based on ram alone, and ignoring anything else.
GAME OF THE YEAR 2007 GAMESPOT/EDGE MAGAZINE/GAMETRAILER ALL OTHER RESPECTED REVIEWERS ETC WIL ALL FOLLOW SUIT
GAME OF THE YEAR 2007 OVERALL ALL FORMATS
M-A-R-I-O G-A-L-A-X-Y
OUCH DOSNT THE TRUTH HURT
THE 1ST XBOX ZOMBIE TO SAY MARIO GALAXYS ZERO LOADSCREENS AND HIGH LEVEL PHYSICS AND GRAPHICS ARE CAPABLE ON A XBOX 1
IS A BORN LYER
THE PHYSICS ALONE IN GALAXY ARE AMASING
NO LOADTIMES AND 60 FALTLES FRAMES A SECOND AND CLAss leading platforer graphics
on a so— called xbox level machine
what a long winded thread when all alone post 1 was the truth EA LIED it was a marketing/policical statment
NOT A STATMENT OF FACT ON WIIS POWER
PLEASE LEARN THAT EA FEAR NINTENDOS INDUSTRY LEADING POWER
ZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZ XBOXFANSZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZ
FUNNY M,ARIO GALAXY LOOKS 100% NEXT GEN AND RUNS AT A SOLID 60 FRAMES AND IS THE BEST VIDEO GAME THIS YEAR AND A GENUNIE TOP 5 OF ALL TIME GAME
WERE IS THE X360 ALL FORMAT WINNER UPS THERE AINT ONE
I thought Mario Galaxy was an awesome game. Not all from a technical perspective, but in general. It likely deserves game of the year.
But it’s graphics and physics, aren’t a testament that the Wii is as powerful as a PS3 or 360, just that a game that’s well made and designed is a good game, regardless of power. (no one is disagreeing with that)
And the physics and graphics are far more interesting, than they are amazing, as far as processing power goes.
Xboxmodder – technically, (if I recall correctly), it’s a 777mhz Celron with the same 8-way associative cache used in Pentium III. (according to the Linux profile on modded XBoxes)
Of course, you only have 128kb cache, etc.. But it’s not the same as a stock Celeron either. As it is, most hyped comparisons between G3s and Pentiums,was between G3 and Pentium II anyway. “up to” twice as powerful at the same clock rate. With particular emphasis on “up to”. i.e, in certain cases they cherry picked, other cases not so much, and in a few others, the other way.
But, then again, Gamecube used a custom modified G3, so it may very well have been 2x at the same clock rate in many, if not most cases. Bringing it’s cpu up to at least Xbox, if not superior.
But, as I think has been stated, Gamecube used tradition T&L, XBox had vertex shaders, which puts more load on Gamecube’s cpu. (and bus)
@Xboxmodder–”"effective in game numbers are real the xbox fans are basing there opinions on …on paper numbers and clockspeeds..wiiboy cubeboy who ever he is is basing the argument on ingame “effective performance” the performance that we all will acualy see in are games.
sorry he clearly wins this”"
Lol, you clearly didn’t read many of his posts. It’s quite clear he’s the one shoving the paper specs and numbers around “out of context” to prove some kind of point. Tech discussions are fun and informative, fanboy driven arguments are not. Too many people with agendas to be interesting.
@Swp64-”Tech discussions are fun and informative, fanboy driven arguments are not. Too many people with agendas to be interesting.”
Lol, “said the pot to the kettle(s)”. Mr. “Nintendo’s new found market leadership + lack of processing power = crippled programming innovations from Japan = Seppuku. (Hara-kiri)”
But yeah, much of what can be done with Xbox’s vertex shaders, would have to be job shared between the geometry engine and the CPU on Gamecube to achieve the same thing.
And it wasn’t anywhere near powerful enough to do the same things most of the time.
(but to be fair, overall, there are some of things you could do on Gamcube, that you couldn’t do on an Xbox, but that list was very short and specific. The list gets pretty long the other way around)
wii and gamecube physics in the lead
YOU TUBES ALL FOLLOWING VIDEOS
PS3 PHYSICS DEMO /PS3 INGAME PHYSICS
VS
GAMECUBE MARIO 128 DEMO / ELIBITS WII /MARIO GALAXY
GAMECUBES INTERACTIVE MARIO 128 A.I /PHYSICS DEMO IS MORE IMPRESSIVE THAN PS3S NON INTERACTIVE DEMO
ELIBITS IN GAME PHYSICS MORE IMPRESSIVE THAN PS3S NON INTERACTIVE DEMO
MARIO GALAXY IN GAME PHYSICS AGAIN MORE IMPRESSIVE
COMPARE PS3 TO WII VIA YOU TUBES WATCH GC AND Wii HAND ON HEART KICK SOME physics ass
BROADWAY IS A RISC OUT OF ORDEWR SUPER SCALLER OPTERMIZED FOR GAMING FUNCTION ADD TO THAT ATIS GPU BASED PHYSICS = WII KICKS OUT HIGH END PC PHYSICS
WATCH AND LEARN CELL IS A INLINE OVERHYPED DUDU PILE
did you realize wii has dual geomitry and light engines (T&L) ENGINES…..
hollywood on board T&L fully supports compression and real time decompression 4to1 data compression
broadway cpu on board support for fully programmable geomitry and lighting (T&L)allso fully supporting real time compression decompression at 4to1
gpu t&l plus custom effects plus TEV unit programmable 16 stage shader blender
cpu fully programmable (T&L) CUSTOM GEOMITRY AND LIGHTING SUPPORT
thats some serious t&l g&l hardware suppot
Yeah, I’ve seen their whitepapers explaining the cheap (inexpensive, not crap) lighting in Gamecube, and heard developers comment on it. It’s probably the same for Wii, and they did some cool things with it. Resident Evil 4 for example.
Gamecube could blend lighting into a pixel, in a texture stage. So, you could read a texture, blend it with the next texture, then blend in a lighting calculation, then blend a texture with that.
Since it writes the result to its on chip cache, and reads it back for every stage, that’s cheap, since you don’t have to re-render the polygon.
Xbox calculates it’s lighting, reads all of it’s textures at once, and runs combiner ops on those, which is pretty fast, and you’re not re-calculating the polygon to do it.
But if you wanted to blend texture layers, then blend in a lighting calculation into that, and blend the resulting texel with another texture, you’d have to resort to multi-pass rendering. i.e, you’d have to write the resulting pixel back to ram, and read it back, after re-rendering the polygon. So in that sense, Xbox could probably not run something like RE4 in the same way Gamecube did, with any kind of speed.
But at the same time, there’s a list of situations where Xbox could do things Gamecube couldn’t. The level of normal mapping in Chronicles of Riddick for example, wouldn’t be doable on Gamecube. Or the per-vertx lighting in Splinter Cell, or the texture mapping in Doom 3, etc..
And I still don’t get what you are referring to with 4to1 compression. Lots of things are considered compression. Z-compression, DXT, etc.Rendering a curve using B-spline in the gpu would be considered considerable compression, since you aren’t reading all the polygons that make up a curve, you’re just tessellating them based on an equation on the gpu itself, rather than storing them in ram, and moving them over a bus, or tessellating on the cpu, and moving them over the fsb etc..
Xbox could compress volumetric textures, Gamecube couldn’t, etc.. But they both comress and decompress everything they move from one place to another on the fly, in hardware.
It’s all relative as it is.
Not degrading your argument btw, but I’ve never heard of gpu based physics on Wii, outside of a bunk developer interview, of a guy who didn’t fully understand what he was talking about. Ati could do physics in unified shader gpus, because the shaders are more general purpose, and can run math related to physics, because they’re more versatile, and can compute a wider variety of computations. From all I’ve seen and read, that’s not the case on Wii, that’s 360 and up.
And really, physics is one of those things cell is actually really fast at. And they had some interesting demos. “Cups”, the fish demo, the one showed involving Heavenly Sword, The water demo Highmoon studios had and made a minigame out of etc.. And it can do those physics, in addition to other things.
Calculating bones, mesh deformations, blending one animation with the next, etc.. Those are things Cell could outrun Wii at considerably, especially on things like http://www.youtube.com/watch?v=hdImUIhbG9E&feature=related.
Not to say Wii is week at physics, I’m sure it’s capable enough, it’s just not gonna outrun something like Cell.
Yeah, I recall the “textured lighting” they did for RE4. That was pretty cool.
Overall, I think if Nintendo had updated to more conventional hardware, and targetted NTSC, they would have been destroying PS3 and 360 right now. (more so than they are)
Plus, it would have had the potential to run things like Bioshock, straight ported. And pretty much any other multi-platform game would have the potential to be ported with little effort. Then, toss in Nintendo first party stuff, and it’d have been game over.
HD won’t really matter much until next gen. Especially in Japan. Hopefully their next hardware is Gamecubesque level. (in terms of relative power to their competition)
That would be awesome.
*I’ve seen many of those physics demos. It performs the way I’d expect a decent cpu to do, given there’s no OS running in the background, etc.. Not bad, not awesome.
This demo was impressive though.
http://www.youtube.com/watch?v=Jd3-eiid-Uw
Pfff too much to read since the last time I posted! Technical discussions are fun indeed. Thanks for all the answers and insight. Still have too read all though.
Some comments from my side. In game performance doesn’t say anything unless the very same game is run on multiple platforms with the least and most performing code behind it. Rogue squadron does 20M polies? Ok, but how many pixels does it render? Horsepower is torque times speed, so (bare) GPU horsepower should be polies times pixels. Only knowing the number of polies doesn’t say anything.
Still some questions though: We discussed fillrates but how about texture reads? Multi layer textures do require multiple texture reads per pixel don’t they? So how can a Wii achieve 2Gpix while has to read pixels from memory and its bus is only 240Mhz and how does the 360 do that? Or does 2Gpix only apply if it has those textures in texture cache (or “catch” as some try to refer to it:)?
@swp’s last post: Yeah, I think having more memory (256M external) would have been enough already. I would buy a game such as assassins creed even if it doesn’t have real bump mapping or realistic lighting (though Zelda has very good lighting IMO (looks like specular), much better than RE4).
@BranH: I think you are right about that fifa developer. Either he developed the J2ME version of fifa and doesn’t know anything about hardware or performing code (like a lot of java devs:), or the writer of this article misunderstood what he exaclty meant. It is quite obvious that the Wii has better memory speed, better CPU speed, better (bare) fillrates so claiming it is on the same level as XBOX wouldn’t be too bright. Though I still wonder, the Wii has newer hardware, so can’t it be possible that it supports the same shader technology as XBOX? I mean, ATI must have based the Cube’s GPU on technology that existed those days. Perhaps they were able to extend the Wii’s GPU with newer technology? Or would that require a completely different architecture?
@havoc, perhaps I was a bit harsch. BTW I already live next to Germany so it wouldn’t make any sense to move (other than cheaper BMW’s haha). But my point is, if you are not interested in this topic, you are free not to post at all instead of calling people nerds (unless you add a smiley to that ofcourse
).
Well, as it was explained to me:
On 360, you have 48 unified shaders, in 3 mimd blocks of 16. Capable of processing either pixel shader, or vertex shader instructions.
A block of 16 texturing units, decoupled from shaders, so they can fetch and filter, without stalling the shaders. A thread arbiter that keeps hundreds of threads in flight, that organizes threads of pixel shader, vetex shader, or texturing tasks.
You have 8 rops, located in a block of 10 megs of edram. Rops can process 8 pixels per clock with both color and z values, or, they can do 2x Z only. (i.e, no color, just “depth” value)
An ideal game engine on xenos, would run a z-only pre-pass first. In other words, it would process nothing but the geometry of a scene. Perhaps pre-cull, run back face culling, clipping polygons outside the view, etc..
Anyway, you write all geometry to the scene first. All 48 shader alus are capable of process geometry, so that runs several times faster than it would on a gpu that only had 8 vertex shaders for example. In older gpus, the pixel shaders sit idle, while vertex shaders work, because there’s no pixel shaders being run in this pass.
The polygons are made up of pixels with only z values, no color, so rops can process them twice as fast. (16 pixels per clock)
When you;re done, you have a frame with the top most pixels as determined by their z-values, (so there’s no overdraw) which then get processed in a second pass by pixel shaders. (in this case, the same processors that just ran the geometry)
It adds textures, runs shader programs, etc to those. (you don’t add complex texture and shader effects to any pixels that can’t be seen in a frame)
You have 48 shader alus, and 16 texture units. The thread arbiter, can keep 64 threads in flight, and hands out tasks to the different blocks of processing units. So, if a pixel shader needs textures for something, that haven’t been fetched from ram yet, the thread arbiter can swap out that thread, and give it a math process that needs done anyway, until the textures are there to process.
In other words, hide texture latency with math operations.
That’s why ati gpus are good at folding at home, while Nvidias haven’t been. They don’t stall as much with texture fetching from ram or filtering, because those tasks are decoupled from the shader units, and they can just do something else while they wait.
Of course, that’s simplified, and ignores how you get 4x msaa over the top of full fill rate, ignores tiling because a 1280×720p frame with 4x msaa won’t fit into edram at once, etc..
I also left out the filling the Hierarchical Z-Buffer in that first pass, the fast Zclear, z-compression, and other hardware functions.
Then there are things like, the fact that you can get HDR at the same cost, bandwidth, and storage as a normal pixel, because it can process a native FP10 HDR format.
Which, allows pixels to have a 10bit value for red, green, blue, and 2bits alpha. = still 32bit pixels.
10.10.10.2, rather than having to double the pixel’s color value with fp16, etc..
And other theoretical things you could technically do with memory export, or with the tessellation unit, etc..
Point is though, that’s where they get the theoretical, 8 pixels per clock, that includes a z-buffer, 2 texture layers, 6 shader ops, and 4x mass.
in other words, there are 8 rops, that can do 2x z only, 16 texture units, 48 shader alus, and hardware 4x msaa that doesn’t cut into fill-rate.
(of course, the mix of instructions would likely be far different overall, but it’s a generic benchmark)
Block diagram:
http://img253.imageshack.us/img253/9563/xenoszf6.jpg
IGN GAME OF THE YEAR MARIO GALAXY
NO LOADSCREENS NO LOADING WAITS NO GLITCHES NO PATCH REQUIRED 60 FRAMES GOUGOUS GRAPHICS
XBOX 1 RAN HALO 2 LIKE SHIT AND AVERAGED 14 FRAMES A SECOND PLUS LONGGGG OUTRAGUS LOADTIMES POP UP TEXTURES MANY BOTTLENECKS
YET WII IS A XBOX
IT AINT SINKING IN IS IT
xbox 360 10mb off chip edram buffer 32gb read 50 nano seconds latency
wii 3mb on chip embedded edram 1tsram-r 28gb bandwidth plus great compression decompression 28gb @ 480p vs 32gb @720p wii clearly far more capable per pixel and resolution to fillrate/bandwidth ratio
32gb buffer @720p poor latency issues vs
28gb catch/buffer @ 480p amasing latecy perfromance of 5 nano seconds wiis catch buffer is better @ 480p than x360s is @ 720p FACT
MAIN RAM LETS SAY 48MB FOR TEXTURES 48 x 6 COMPRESSION =288 MB OF TEXTURES thats 288mb of sd resolution textures 3 times smaller than 720p
so 288mb of textures @ 480p = shit loads of textures and texel fillrate at 480p
wii 1mb texture catch = 16gb bandwidth 16gb x 6 for compression= 96gb texture virtual bandwidth @ 5 nano seconds ON CHIP
TEXTURE READS OF 96GB BANDWIDTH VS AROUND 3.2 GB BANDWIDTH XBOX 1
CLEARLY WII IS A 480P 360 NOT A 480P XBOX
DO THE MATH
TEXTURE READ WII 1MB TIMES 6 = 6MB VIRTUAL TEXTURE CATCH WITH 24 MB OF ON DIE 1T SRAM-R INSTANTANIUS READ VIA SUPER FAST 1T SRAM 10 TIMES FASTER THAN 360S RAM FACT
HD TEXTURES TAKE 4 TIMES MORE RAM SPACE FOR THE SAME TEXTURE AS A 480P TEXTURE
40 MB OF TEXTURES WII X 4 = 160MB ADD 6 TIMES COMPRESSION YOU GET THE PICTURE WIIS A 480P BEAST
DO THE MATH
WII FILLRATE 2XPLUS XBOX ///WII LATENCY 10TIMES PLUS XBOX ////WIIS BANDWIDTH 5 TIMES XBOX//// WIIS DISC TWICE SIZE OF XBOX PLUS GREATER COMPRESSION AND MUCH FASTER LOADING SPEED
WII CLEARLY NO XBOX
xbox sdram slowwwwww frontside bus = 133 mhz = 1gb bandwidth
wii 1t sram-r edram 1tsram-r die embedded ram plus clock synced gddr3= dam fast
wii front side bus 243mhz plus 4to1 compression = 8gb bandwidth
again wii is noooooooo xbox
Wiiboy, you’re a delusional, irrational fanboy, and you don’t even try to understand even a fraction of what you type about. It’s obvious you have no interest in this discussion, or the subject in general, outside of fellating Wii hardware.
I would hate to see the forums you frequent, as I’m sure you think you “know stuff”. The ones in which you aren’t banned from of course. And I’m sure people just blow you off as I’m about to do, or just don’t start tech discussions, just to avoid your cap-locked bs.
But I’ll help you with a couple of your “issues” since I’ve typed this much.
***32gb buffer @720p poor latency issues vs 28gb catch/buffer @ 480p amasing latecy perfromance of 5 nano seconds wiis catch buffer is better @ 480p than x360s is @ 720p FACT ***
the 32gb’s bus, is write only. It’s not addressable by the gpu. It just moves data across it to rops (located on the edram die) in one direction, it makes no requests from it directly. The bus between rops and edram, is a separate 256 gigabytes per second. You don’t store textures there, or game code, etc..
Your latency claims are wrong, in addition to being meaningless.
As is the vast majority of what you write. It’s not even worth discussing.
FACT
FACTOR 5 working on wii exclusive ps3 dropped like hot potato reports of crying at nintendos door tail between legs not confirmed but likly
WIIADD FACTOR 5 = XBOX GRAPHCS BLOWN AWAY
GAMECUBE ROGUE SQODRON 2 15 MILLION POLYGONS 60 FPS 8 TEXTURE LAYERS REALTIME SHADOWING SHADING BLENDING
GAMECUBE ROGUE SQODRON 3 NEAR 20 MILLION POLYGONS NEAR SOLID 60 FPS LIGHT SCATTERING HDR ETC
BOTH GAMES SUPPORTED 512X512 TEXTURES AND 5 TO 8 TEXTURE LAYERS INCLUDING DOT 3 BUMP MAPPING
HALO 2 XBOX MAXED OUT 4 TEXTURE LAYERS ZERO 512X512 TEXTURES SUPPORT BUGS GLITCHES ETC LONG LOADTIMES CRAPPY SUB 30 FRAMES
NBO XBOX 1 TITAL SUPPORTED 512X512 TEXTURES ONLY 256X256 “”"FACT”"”
NO XBOX GAME BETTERED 15 NMILLION POLYGONS AT SUB 30 FPS
XBOX SUB 30 FPS AND SUB 15 MILLION POLYGONS
GAMECUBE NEAR 20 MILLION POLYGONS NEAR SOLID 60- FRAMES HOW IS WII A XBOX
WATCH FACTOR 5 ON WII WATCH CAPCON ON WII MONSTER HUNTER
YOU BELIEVE XBOX CAN DO WHATS COMING “”"”FANBOYS TALKING CRAP THE LOT OF YOU”"”
GAMECUBE OUT GRAPHIC=-ED XBOX MANY TIMES OVER
WIND WAKER TWILIGHT PRINCESS
PRIME 2/3
RES EVIL 4
ETC
ALL NO LOADING TIMES
WII SHALL KICK OUT 2.5 TIMES THE VISUAL POWER AT 480P THAN XBOX OR CUBE EVER DID AND STILL RUN LOADSCREEN FREE
DONT BELIEVE MII PUT YOUR MONEY NAME AND PHOTO WRE YOUR NET POSTS ARE IM WILLING CUZZ I KNOW ILL WIN
SILLY XBOX FANS TALKING CRAP
COMING WII GAMES IN 2008 WILL PROVE WII AQ 480P OPTERMIZED 360 LIKER MACHINE NOT A XBOX1 OR SUB XBOX 1
FANBOY ALERT SOUND THE XBOX FAN ALARM THERE TALKING SHIT AGAIN
lennel- yes, i agree partially with that. And I pointed that out in this wall of text thread. I did say, and this is agreed upon by nearly all developers, that overall, Xbox > Gamecube, in both flexibility and overall power. It doesn’t have to be superior at “everything” to be superior overall.
I am not of the opinion that Wii is no better than an Xbox. It’s probably not as flexible at some things, but overall it’s more powerful. The only one that’s passionate about it is wiiboy. I really wouldn’t mind if Nintendo hardware were proven to be awesome, I just don’t see that.
lol, Wiiboy calling other people fanboys, hilarity ensures.
Anyway, you keep comparing x-game to Halo 2, of which you know little about as it is.
Gamecube would choke and die running Halo 2. It couldn’t keep up on the bump-mapping alone.
And you throw around goofy polygon numbers for what? You do know that for every vertex attribute you add to a polygon, the fewer you could actually transform, right? Your assumption that in any of the more ambitious Xbox games, the geometry processed would somehow run better on a Gamecube is flawed. Keeping up with bones, and weights, and animation, etc. would be too much for it.
On a Gamecube, you’d simply not try to run such things as much, nor would you attempt the level of bump-mapping, etc..
You’d do what they did in Prime, and use a bunch of generic effects, with good art direction. So what if you don’t have bump mapping anywhere, on walls, characters, etc., as long as it looks good. (better than most) And there are fewer glitches involved in the “generics”.
And so what if in Nintendo games, the polygon count on character drops when they run, or you lack aa, or are you’re forced to use lower color precision at times and you see ugly banding everywhere, as long as fans don’t complain too much, and the frame rate stays ok, it’s fine.
And I didn’t say cube couldn’t do dot3 bump mapping, just that it doesn’t have efficient hardware to compute it, and couldn’t use it anywhere near as much as an Xbox could. Gamecube is good at older multi-texturing, but it’s not going to keep pace with an Xbox in that area.
And you’re mentioning things like Windwaker, RE4, and Prime as examples of what?
They’re very nice looking games. But they don’t do much to prove graphical superiority, considering an Xbox fan could just as easily list Riddick, Doom 3, NG Black, Splinter Cell PT, DOA:U, Wreckless:Yakuza (which was on Gamecube, sans bump-mapping) etc..
As well as having superior looking versions (and sometimes at much higher resolution), of nearly all multi-platform games.
And you continue to misunderstand texturing. Gamecube can do “up to” 8 texture layers. Sure, you could pile up 8 onto a single pixel, on some of the surfaces in a game. But you’ll notice the comments about limiting a metatile to 3 texture layers because of performance and memory reasons, was from Factor5. In addition to other devs mentioning that,in practice, you’re pretty much stuck to ~2-3.
And your assumption that Xbox “maxes out” at 4 textures is wrong. It maxes out at 4 texture layers “per rendering pass”.
No rules that say you can’t do a second or even third render pass to add more textures, blendiing, etc..
Doom 3 did several render passes on geforce3/4, and had 7+ texture layers, and a list of blending effects. It would simply cost in things like geometry processing. (which is more powerful than Gamecube’s, especially at dx8+ level programming)
In closing, Gamecube/Wii 4Life, yo.
how come mario galaxy has twice the visuals and twice the frame rate of any xbox game iv ever seen and does it all without bugs glitches tearing or crashing loading screen loading break free and lets not forget the physics….
excite truck a gamecube dev kit developed game slapped on wii at lournch has better graphics and framerate than the xbox could ever run that fast sweet arcade racer and allso supports perfect motion stearing and envirament morphing texture mophing
how come twilight princess on wii has identical graphics to the gamecube version only upgrading interlace to proscan and 4:3 to 16:9 yet has way better graphics than any xbox game and allmost no loading times….
how come visually fzero gamecube matches any xbox racer yet its frame rate motion blur etc make fzero clearly the fastest racer game ever created its insanly fast and glitch free
how come when ubisoft started optermizing there ps2 to gamecube ports insted of the lazy ports of earlier games the gamecube clearly out done the xbox even on ported ps2 code
jade engine beyond good and evil was pushed harder and better on gamecube than xbox it had better colour better effects better water and physics better loadtimes and more debugged same applys a few other cube optermized games
how come wind waker by far had the greatest cell shading last gen it clear shit on any cell shaded xbox game and again ran allmost compleatly free of load breaks
how come prime 2 on gamecube had zero loadtimes screens ran perfect at 60 frames a second had no bugs glitches or issues and had amasing graphics
yet halo 2 had long loadtimes bugs glitches plain colour borring flat textures pop up graphics pop up textures glitches bugs and very very very long loadtimes
the xbox fans and the nintendo fearing 3rd partys have some explaining to do
Who cares about power when arguably the best game of the year is out exclusively on Wii?
Also, it looks damn good, better than xbox game I’ve ever seen. Sure, Wii is far less powerful than 360 or PS3, but as long as it has the potential for amazing games, who really cares? Not me. I’ll play amazing games on any console.
c`mon guya ffs! who cares? choose the system you like or buy them all! all i know is micro ,ninty, and sony love all this stuff you put out on the net etc, just shut up and play , or make somthing better yourself , peace!!
@chicothespic, branH and swp64 already explained a some of your questions. Read all from the start of this thread.
@branH: You certainly know your stuff. Thanks for the link, I’ll look into it. BTW I should have thought of your explaination about texturing myself, it is the same solution as for fillrate you explained earlier. After the GPU finished geometry calculations, it only needs to read the visible texels from memory and write ‘em to the framebuffer. I feel like I’m up to date now
@baco, you are right, but it is quite interessing to know how stuff works so we at least know what to expect from systems.
Perhaps time for a conclusion as I look at it at this point:
The Wii is faster than XBOX, but may lack some of the modern pixel mapping techniques such as bump/normal mapping. It also lacks newer technologies as seen on the 360 such as parallax mapping and light effects. (I’m sure these effects could be simulated on a Wii somehow, since it does support displacement mapping, but it would be at a great performance penalty). Fillrate wise @480p it may be on par with the 360 @720p though. I would suppose that polygon calculation would be on par with XBOX (I think it is assumable that is can calculate 120M polies a second).
So, what to expect: games with RE4’s texture resolution or better, twilight princess quality of lighting, perhaps some bumpmapping, twice as detailed environments/characters and more/better AI. I hope monster hunter will prove this to be right, especially since it was to be a PS3 title.
chicothespic = Simple, most games optimized for Gamecube, have a good chance of looking better on Gamecube than it will on XBox.
That’s not so hard to figure out me thinks. If it’s “Gamecube optimized”, it’s set in stone, to use the types of multi-texturing effects the Gamecube is good at. Porting anything that takes a dx7 or lower code path wouldn’t give an Xbox an advantage, outside of perhaps aa, or higher color precision, etc.. Many areas of the system would go under-utilized, and would only use the limited functionality that the Gamecube is capable of. It doesn’t magically compute things like normal maps that weren’t designed into the game in the first place.
However, developing something designed for Xbox, and trying to port it visually as is on Gamecube, with all of the bump or normal mapping details, vertex and pixel shaders the XBox could muster, would simply not run. At all. Emulate what the Cube’s gpu can’t do, on the cpu. And watch it be nothing more than a frame by frame slide show, if it can render a frame at all. Everything would have to be cut back to dx7 ish level, and just cut back in general.
That’s just the way it is.
Not sure what you guys are so puzzled about. It doesn’t have to be more powerful than the Gamecube at everything, to be more powerful overall.
And you seem to think that you can counter the Xbox > Gamecube stuff, by making worthless points about Metroid Prime. It uses zero bump mapping of any kind.`It’s not an ugly game, or a primitive game overall, but it runs well because they don’t try to do things they aren’t capable of doing.
And your frame rate, depends entirely on the amount of work you;re doing per frame.
You can make an Xbox game that runs at a hundred frames per second, if that’s what you were shooting for when you started. It’s not as if any of these system aren’t capable of it. They intentionally shoot for ~30 frames per second, and design the game around that.
Nintendo first party especially, does a good job of toning down what they need to, in order to get their ~30 frames.
Even when in Mario Sunshine, it took using grainy, redundant, low resolution textures over most everything, in order to have higher res, higher quality on a few key things, and not have to load as you progress through the (small) levels.
They didn’t mind if some things looked pretty crappy, to get other, more important things looking the way they wanted, and keep a decent ~30fps rate.
That’s a good philosophy to have if you ask me.
@not a fan- I was playing devil’s advocate to Wiiboy, I don’t think of the Gamecube as weak compared to an Xbox. It was competitive from all I’ve seen. Wii should be able to top it.
Wii displacement mapping, is just someone misunderstanding a Nintendo patent for emboss bump mapping. Not sure if i could post the url here properly, but here it is:
http://appft1.uspto.gov/netacgi/nph-Parser?Sect1=PTO2&Sect2=HITOFF&p=1&u=%2Fnetahtml%2FPTO%2Fsearch-adv.html&r=3&f=G&l=50&d=PG01&S1=%22emboss-style+bump+mapping%22&OS=%22emboss-style+bump+mapping%22
It’s just hardware function to compute the displacement of an embossing map, based on a supposed light source, without involving the cpu. More or less fixed function pixel shaders.
Like this:
http://www.delphi3d.net/articles/bumpmap_02.jpg
If you copy that into photoshop or studio, then copy the top center square, paste it, make it 50% transparent, and move it over the top of the bottom center one.
Then move it around according to an imaginary light direction.
The only mention of geometry in that patent, is in regards to a typical it’s t&l unit, since that effect would be tied in with the movement of a polygon.
They could have added some functions to compute something more complicated, but i don’t really see much evidence of that yet. (in practice, or developer comments)
And I wouldn’t expect 120 million polygons out of Wii (or Xbox). AFAIK, it would be half that at most in just vertices (about 50 million vertices). A vertex being the corner of a triangle, (shared with its neighbor where possible), so depending on what type of game, and what the meshes look like, you might get over half that.. >30-35 million. But that’s nothing but transforms. Then, you can start cutting that, depending on a list of other effects you might do.
@not a fan
it was partly intended with as a kind of homer simpson quip, but put it this way, would any one who wasnt any kind of nerd talk about polygon counts ETC.
no offence, by the way.
also, i was interested from a “would i want to buy some thing that says its new but actually has antique technology in it” point of view, if you get what i mean.
and considering its supposed to be part of the “current” generation of consoles.
@BranH = Yeah. Even the ~35 million per second figure, would be really generic polygon work.
Wii = a few hundred thousand polygons in any given frame, tops.
And yeah, that was what I was referring to when I said people were grasping at straws when the Wii launched. Nurbs, displacement mapping, etc..
@not a fan -Although they can do things like EMBM (environmental bump mapping), I see no evidence of any extensions to the gpu itself, for any of the more modern types. They could have, (this is ati after all), but I don’t see it, and most developers have implied (and complained) that it doesn’t. Supposedly, it’s not much more than an over-clocked Gamecube, with more ram.
It’d be hard to believe that that’s ALL it is, and perhaps developers hate on it so much, because they can’t just drop their game off onto its hardware, and run it as is, the way they could for a low end pc for example.
We’ll see I guess.
broadway cpu well over 20 million trancisters xbox1 celeron 9 million as you can see theres more than double the logic count on broadway than there is xbox combine that with risc/copperwire/embedded ibm mpu design/silicon on insulator= a celeron destoyer pure and simple……
gekko was more powerfull than xbox celeron FACT FACT FACT it had copperwire risc and customizations and was a shrunk tightly compact MPU design plus a 162mhz front side bus with 4to1 compression = virtual bus of 648 mhz (xbox 133mhz hmmmm)
risc beats cisc
copperwire beats aluminium
silicon on insulator beats non silicon on insulator
mpu embedded design beats off the dam shelf
256k level 2 catch plus 4to1 realtime compression decompression beats 128k level 2 catch nooooooooo custom compression
broadway cpu = 2x gekko = 2.5 x celeron FACT
XBOX BUS 133MHZ =1 GB BANDWIDTH
WII BUS 243 PLUS REALRTIME 4TO1 COMPRESSION DE-COMPRESSION = 8GB BANDWIDTH
XBOX FANS DO THE RESERCH DO THE MATH
ADD TO THAT DIE EMBEDDED 1TSRAM-R IS AS FAST AS LEVEL 2 CATCH 5 NANO SECONDS 24MB VIRTUAL LEVEL 3 CATCH DESTROYS XBOX RAM
WII SRAM = 64K LEVEL 1 AND 256K LEVEL 2 ALL SUPPORTING REALTIME 4TO1 COMPRESSION CPU AND 3.2MB GPU CATCH PLUS 24MB VIRTUAL LEVEL 3 CATCH (DIE EMBEDDED 1TSRAM-R)
XBOX SRAM= 32K LEVEL 1 128K LEVEL 2 NO CUSTOM COMPRESSION AT ALL IN HARDWARE CPU AND A 256K GPU CATCH NO COMPRESSION SUPPORT IN HARDWARE
A INSANE DIFFERANCE
DO UR RESEARCH WAKE UP WII IS A HIGHLY CUSTOM ON THE FLY HIGH SPEED COMPRESSION DE-COMPRESSION CUSTOM CONSOLE XBOX IS A OFF VTHE SHELF BASIC PC
ALLMOST 28MB OF SRAM PLUS REAL TIME CUSTOM COMPRESSION OF DATA AND TEXTURES IN REALTIME
VS WELL UNDER 1MB FOR XBOX NO SUCH COMPRESSION TRICKS
DO THE MATH
WII TEXTURE READ 1MB TIMES 6 = 6MB TEXTURE CATCH
WII TEXTURE READ 16GB TIMES 6 FOR COMPRESSION =96 GB BANDWIDTH
XBOX TEXTURE READ ERMMMM ERMMMMMM ABOUT 3GB IF YOUR LUCKY AND A GPU CATCH OF ERMMMM 256K NO COMPRESSION INTO CATCH IN REALTIME SUPPORT WHATS SO EVER
HOLLYWOOD READS 6MB TEXTURES IN CATCH IN REALTIME ITS LIKE MANY MANY TIMES THE PATHETIC DRAM AND TINY CATCH OF XBOX
XBOX RAM 70 NANO SECONDS LATENT
WII RAM 5 NANO SECONDS READ LATENT THE MAIN MEMORY FEEDS THE CATCH IN REALTIME GIVING UNLIMITED ROOM FOR TEXTURE FEEDING
FROM DISC THRU SYSTEM TO THE SCREEN WII NUKES XBOX AT TEXTURE FEEDS AND FILL RATES AQND EFFECTS “”"”"”"”"”"”"”FACT”"”"”"”"”"”"”"”"”"”"‘
EVERY 1MB OF RAM IN WII CAN HOLD 4MB OF DATA////// EVERY 1MB OF TEXTURE RAM CAN HOLD 6MB OF TEXTURES ALL DECOMPRESS AT THE POINT OF PROCESSING WITH IN ETHER THE CPU OR GPU ON THE FLY IN REAL TIME NO OTHER CONSOLE CAN DO THAT
243MHZ BUS BECOMES 4X THAT=972 MHZ VIRTUAL BUS
256K LEVEL 2 CATCH CPU BECOMES 1MB VIRTUAL CATCH
1MB TEXTURE CATCH BECOMES 6MB
GUESS WORK MAIN RAM
24MB DATA BECOMES 96MB DATA
SAY 40MB TEXTURES BECOMES 240 MB
SOUND 4 MB TIMES 4 COMPRESSION BECOMES 16MB
ALL ON THE FLY HIGH SPEED LOADING REAL TIME DECOMPRESSION NO PC OR COMPEATING CONSOLE CAN DO THIS “”"”"”"”"”"”"”FACT”"”"”"”"”"”"”"”!!!!!!!!!
WII=480P OPTERMIZED X360 LIKE PERFORMANCE NOT XBOX 1 AT ALL
Lol. Wiiboy, drowning in his own “facts”. Your like a bot, with an inferiority complex on Nintendo’s behalf.
**EVERY 1MB OF RAM IN WII CAN HOLD 4MB OF DATA////// EVERY 1MB OF TEXTURE RAM CAN HOLD 6MB OF TEXTURES ALL DECOMPRESS AT THE POINT OF PROCESSING WITH IN ETHER THE CPU OR GPU ON THE FLY IN REAL TIME NO OTHER CONSOLE CAN DO THAT**
No it doesn’t. Chips have used color and z compression between rops, and frame buffer. All of them do. Every last one of them. You’re confused, simply because the Gamecube and Wii read and write to it’s on-chip ram.
Everything gets decompressed “at the point of processing”, all of them do it.
You continue to confuse yourself, by comparing bandwidth and compression of caches and buses internal to the chip, that are required for Gamecube and Wii, for all the reading and writing that they have no choice but to do, between its ram and the frame buffer.
An Xbox reads its textures (in happy 6:1 compression), decompresses them “at the point of processing” and can combine two layers at a time, loop those pixels back, (internal to the chip) and add two more, then write the result back to ram. (where its frame buffer was stored) (It’s 30 gb per second cache serves its purpose)
Gamecube, has to read and write to and from frame buffer every single time it wants to add a texture, and cuts into its fill-rate for each and every one. It needed it’s on chip ram, it’s part of the design.
So, doing bandwidth comparisons in the manner you are, tells you nothing.
And it’s just funny how you continue to think you can magically apply your “4to1 compression” to everything.
And transistors don’t tell you much either, considering the embedded ram adds a large number of them. And you can get benchmarks of the G3 chip that came after Gecko. It’s not an unknown quantity.
And it’s really funny that you think compression (or anything for that matter) is faster on a Wii, than it is on anything current, pc or otherwise.
A 360 has all the compression Wii has (heck, an Xbox did), in addition to newer, more advanced formats, for things like normal maps.
Wii can do nothing more than standard s3tc.
They’re not in the same league, ballpark, or heck, their barely in the same sport.
XBOX CANOT DECOMPRESS IN REAL TIME XBOX CANNOT DECOMPRES IN CATCHES FACT FACT FACT XBOX DOES NOT SUPPORT CUSTON DATA COMPRESSION FACT FACT FACT
DATA COMPRESION 4TO1 GC AND WII 243 FSB = 8GB BANDWIDTH
XBOX NO SUCH TRICK 1GB BANDWIDTH FACT
XBOX SUPPORTED COMPRESION “”"”"”"”"”"”"”"NOT REALTIME DE-COMPRESSION FACT”"”"”"”"”"
WII ON THE FLY DECOMPRESION OF DISPLAY LIST/SCIPT/GEOMITRY ETC ETC ETC ON THE FLY FROM DISC THRU SYSTEM TO PROCESSOR THEN DECOMPRESED AT THE POINT OF PROCESSING
XBOX HAD NO SUCH DATA COMPRESSION AND ONLY COMPRESSED TEXTUES IN RAM
I REPEAT AS ITS FACT ONLY WII SUPORTS SUCH CUSTOM TRICKS
Wow, so you can compress, just not de-compress. That makes so much sense dude.
Stop making shit up WiiCube. I’m not guessing, nor am I speculating.
Just with an Xbox 1.
Any time you read or write anything to and from ram, it gets compressed and decompressed in real time. You never read or write anything uncompressed. It reads textures (in compressed form) from ram, over it’s memory bus. (It’s still compressed).
They get decompressed after they’ve been read, then process in its pixels, then re-compresses the color and z values of those pixels, then they travel (compressed) over the main memory bus, and are written into ram, where they are stored, compressed.
That is the entire point of texture compression when reading textures,
and color and z-compression for the framebuffer, WiiCube.
That was the PS2’s problem, in that the GPU couldn’t deal directly with compressed textures, it could only save storage space in ram, because textures would have to be decompressed (by the cpu), before they were passed over the bus to the gpu.
DXT 1-5 saves in texture storage and texture bandwidth, Lossless Z-compression saves in frame-buffer storage and bandwidth. (In addition to occlusion culling and query, etc, etc.)
That is the point of texture compression. That is the point of d3d compression period. That is the point of decoding and decompressing audio streams.
It saves bandwidth, and storage space.
And you did read the erp quote, right?
http://72.14.205.104/search?q=cache:RffI1PGW2mgJ:forum.beyond3d.com/archive/index.php/t-2168.html+%22DXTC+textures+and+that+reduces+texture+bandwidth+significantly%22&hl=en&ct=clnk&cd=1&gl=us&client=firefox-a
Which is exactly why your multiplication skills go to waste, when you attempt to inflate Gamecube / Wii’s effective storage capacity and bandwidth numbers, to make it easily comparable to next gen hardware, especially when you ignore similar and better capacity on other hardware. It must be really traumatic for you to learn otherwise.
http://72.14.205.104/search?q=cache:f-hqNPlNSIIJ:www.microsoft.com/presspass/press/1998/mar98/s3pr.mspx+%22by+a+factor+of+four+to+six%22+%22texture+storage+and+read+bandwidth%22&hl=en&ct=clnk&cd=1&gl=us&client=firefox-a
oh thers more DEDICATED SOUND PROCESSOR BUS
XBOX SOUND CHIP WAS SENT DATA OVER A ALLREADY KRIPPLED BUS SHARED FOR THE WHOLE SYSTEM HMMMMM
GAMECUBE AND WII HAVE DEDICATED SOUND BUSES 8 BIT AND CLOCK BALANCED TO THE CLOCK SPEED OF THE HIGH PERFORMANCE SOUND PROCESSOR A DEDICATED BUS TO GO WITH THE DEDICATED CHIP ONLY GAMECBE AND WII HAVE THIS
XBOX SHARED SOUND DATA ON ITS SHARDED FSB PS3 AND X360 DONT EVEN HAVE SOUND CHIPS OR BUSES THEY HAVE COUGH COUGH CPU SOUND AND SHARED BUS WITCH IS SHIT
ALL GAMING PCS AND WII HAVE DEDICATED SOUND CHIPS X360 AND PS3 DONT EVEN HAVE SOUND CHIPS
WII HAS 2X MAIN MEMORY POOLS 24MB 1TSRAM-R DIE EMBEDDED ULTRA FAST RAM 2X 64 BIT BUS AND A SECOND POOL OF 64MB GDDR3 DUALCHANNEL 32 BIT BUS
PLUS ACCESS MEMORY VIA FLASHDRIVE ON DEDICATED FLASH MEMORY BUS (POSSABLY 16 BIT)
XBOX HAD ONE SINGLE POOL OF SHITTY SDRAM AT 200MHZ AND THATS IT CLOCKED UPTO 400MHZ IN MEMORY CONTROLLERS HARDLY WII LIKE IS IT
OH YES XBOX REQUIRED MAIN MEMORY FOR FRAME/Z BUFFERS UPTO 16 MB BEFORE YOU EVEN START USING MAIN MEMORY FOR ANYTHING ELSE
WII DOES ALL FRAME/Z BUFFERING ON CHIP IN ITS ULTRA FAST ULTREA HIGH BANDWIDTH GPU CATCH
3MB ULTRA HIGH SPEED ULTRA HIGH BANDWIDTH REAL TIME COMPRESSION DECOMPRESSION GPU CATCH ALL CATCHING AND BUFFERING DONE ON CHIP 3MB
BACKED UP BY 24 MB 1TSRAM-R EMBEDDED RIGHT NEXT TO GPU ON DIE (27MB OF SRAM BEFORE U EVEN TOUCH EXTERNAL MEMORY)
XBOX EXTERNAL RAM 64MB SOOOOOOOOOO WERES THE SRAM CATCH THERE AINT ANY AT ALL ONLY THE STANDARD GPU TINY CATCH AND THE CELERON TINY CATCH THATS LESS THAN 1MB
NOT ONLY IS WIIS 27MB IN SIZE ITS GPU AND CPU CATCHES ARE ULTRA HIGH BANDWIDTH AND THE WHOLE SYSTEM SUPPORTS COMPRESSION REALTIME DECOMPRESSION
WII CLEARLY 3 TIMES A XBOX 1 IN MEMORY TERMS IT CLEAR KILLS XBOX
XBOX 1 200MHZ SDRAM @ 200MHZ 64MB
WII 24MB 1TSRAM-R FAST RAM ON DIE @ 486MHZ 2.5 TIMES FASTER IN NATIVE CLOCKSPEED OVER 10 TIMES FASTER IN NANO SECONDS LATENCY
EXTERNAL 64MB GDDR3 @ 486MHZ NATIVE CLOCKSPEED 2.5 TIMES FASTER THAN XBOX SDRAM CLEARLY AND ALL RAM IN WII IS CLOCK SYNCED AND BALANCED GDDR3 IS FASTER AND SWEETER THAN SDRAM CLEARLY
2 POOLS OF BALANCED CLOCK SYSNED HIGH PERFORMANCE RAM VS 1 POOL OF SHITTY SDRAM
I REST MY CASE YOUR HONOR
so we have establihed that xbox fanboys and anti nintendo statments from within the industry by jelous EA games believe that
1, 64 mb of rubbish sdram beats 88mb of dam fine and fast ram
2 a off the shelf aluminium celeron beats a risc copperwire gamecentric optermized embedded highly efficent powerpc cpu
3,xbox tiny sram beats a 27mb pool of great sram and a large 256k l2 cpu catch
a sharded and krippled fsb beats a clock balanced higher speed bus and a dedicated sound bus
4, shitty sdram beats 1tsram-r and gddr3
5, high speed optermized priority disc drive is beaten by standard dvd drive
6, off the shelf aluminium beats copperwire silicon on insulator micro embedded design
7,133mhz bus beats 243 mhz bus with 4to1 compression (virtual bus of 4x 243mhz)
8,standard inefficent xgpu beats a highly fine tuned ultra efficent kinda tile rendering custom design gpu
8, tiny gpu catch that carnt read compressed data beats a 3.2mb high bandwidth super fast sram catch…
i think xbox fans are somewhat confussed
xbox 12 million polygons @ 30 frames a second
gamecube near 20 million polygons @ 60 frames a second
wii 50 million if not more @ 60 frames a second (thats 360 level)
but aparantly wiis a xbox for gods green earth only an idiot fanboy would believe that bull shit wii is as was leaked at e3 a gamecube 2.5 witch is roughly a xbox 2.5 so how on my big phat cock is wii a xbox
Lol, Wiiboy, you’re not informing me of anything.
I believe I’ve mentioned, that Gamecube has its on-chip caches, for the reading and writing it does between them. It’s inefficient with bandwidth, and therefore needs more of it, hence, its on chip frame-buffer, and frame-buffer bus.
It writes, reads back, write, reads back, writes, reads back, etc, etc, etc, for adding addition texture layers. So the bandwidth stays internal to the chip, rather than hitting main memory’s bus.
An Xbox can combine its texture layers internal to the chip, using its internal bandwidth, caches, and loop-back. It only writes to main ram again (compressed), once it’s been combined, into pixels.
And you point out the Gamecube’s crappy A-ram. That usually meant that you’re limited to 24 megs of ram for higher function graphics data, like textures. The only other things A-ram was used for, was menu functions, and to help with load times by storing things there like a cartridge would.
But ask any developer, if they’d rather have an additional 16 megs of fully functional, main ram to do with what you want, or a pool of crippled a-ram, with negligible bandwidth, see what they tell you.
You’re confusing Gamecube and Wii btw, Gamecube’s was 81 mhz, 8-bit bus. A whole additional 81 megabytes per second of awesomeness.
Wii replaced A-ram with GDDR3 in Wii, last I checked.
And it’s nice that you constantly ignore 128 bits wide data bus, when doing your “awesome comparisons” Cherry pick much?
And the 360 and PS3 don’t need sound chips, when their cpu’s are capable of devoting more processing to sound, than the entire Wii cpu. Calling its buses or bandwidth, or processing ability “shit” while fellating Wii hardware, is nothing but pure, hilarious fanboyism.
***OH YES XBOX REQUIRED MAIN MEMORY FOR FRAME/Z BUFFERS UPTO 16 MB BEFORE YOU EVEN START USING MAIN MEMORY FOR ANYTHING ELSE***
A 640 x 480 frame. 32bit color and Z = 2.4 mb.
Which is then, on top of that. We don’t technically have to include AA in that figure, (we could) but considering Gamecube would need to drop color precision, and sometimes cut its fill-rate, it wouldn’t make much sense to even throw that in to bloat the frame buffer size. (we could also consider that aa pixels compress much higher too)
(or, unless you want to do your comparisons assuming Xbox is taking a multi-platform game like SC3, and using its power and resources to render in high definition, and compare that, to the PS2 and Gamecube at standard res. Then, we could inflate it to 16 megs, but even then there’d still be color and Z compression)
Anyway, you’d still have to convert the back buffer to a front color buffer, but so does Gamecube. The video DAC reads the image data from main ram.
And, last I remember, since the on-chip back buffer on Gamecube never becomes the front buffer, the Gamecube had to triple buffer (doubling the amount of ram for front buffer)
Back buffer, “middle buffer”, and finally front buffer.
So, what other bullshit assumptions do you meed disproven WiiCube?
And keep in mind, that I am aware of your attempt to switch back and forth between Gamecube and Wii, knowing full well that most people here agree, and have even explained why they believe Wii > Xbox and Gamecube overall.
That part of your fanboyism isn’t being questioned. It’s all the Gamecube = “awesome” era bullshit, you’ve resurrected from your fanboy wars with Xbox, to somehow prove the Wii’s awesomeness that was being questioned, and picked apart.
I guess you do that, so you can think you’re being informative, or pretend you’re arguing the obvious, and “fighting the good fight”, “fending off the fanboys”.
The only one acting as a fanboy here is you, and of course, the random Nintendo defense force members who occasionally drop in, of whom, all seem to share a brain with you, and continue to recycle their ignorant polygon count and frame-rate numbers.
I would love to find out that Wii is magically awesome. I would enjoy reading the technical explanations of it, or point out where my assumptions are wrong. I have no problem with that.
Only problem is, that isn’t happening. Sorry, Wii is what it is, and it seems you’ll have to wait till next gen, to see a Nintendo console compete graphically again.
(of which I more than expect them to do, not the best, but good at most things)
@modderuk – Just goes to show, you know absolutely nothing about what you just copy and pasted. Congratts.
If I thought you knew what you were talking about, I’d ask you how you arrived at your figures, but I won’t.
And not even remotely, is Wii capable of anywhere near the geometry of a 360. (nor is PS3 for that matter)
*Minor typo correction.
**”which is then **compressed**, on top of that”
broadway is in intel terms a celeron 1.5 mobile with 1mb l2 catch and 400mhz buz brodway is 2 times a intel in clock for clock power then add all the custom upgrades like compression silicon on insulator gameing instructions etc custom fpu and superier fsb its clearly on par with a intel mobile celeron 1.5ghz with 1mb catch
now imagine a 1.5 ghz 1mb catch version of xbox one i can clearly see broadway hitting 2.5 times the performance of xbox 1 cpu you clock brainwashed baffoons and all at a capped 480p dooooooooooo theeeeeeeeeee mathhhhhhhhhh