Slashdot Mirror
Next-Gen Console CPUs Not Up to Hype
rAiNsT0rm writes "Anandtech follows up their initial in-depth coverage of the Xbox 360 and PS3 CPU with the real truth about the next-gen consoles' Poor CPU Performance. From the article: "Speaking under conditions of anonymity with real world game developers who have had first hand experience writing code for both the Xbox 360 and PlayStation 3 hardware (and dev kits where applicable), we asked them for nothing more than their brutal honesty. What did they think of these new consoles? Are they really outfitted with the PC-eclipsing performance we've been lead to believe they have? The answer is actually quite frequently found in history; as with anything, you get what you pay for."" Update: 06/30 21:11 GMT by Z : The original article disappeared from Anandtech, so I've changed the link to point to the story as hosted by Google Groups.
1. With the next generation of consoles becoming nothing more than computers, what becomes the purpose of having two separate machines? Or perhaps the real point is, why use your computer for gaming?
2. What will the next generation of consoles actually do to improve the quality of games? Polygon technology has reached an apex whereby increases in graphical quality are hardly noticable in most cases. What about the *fun* factor? Early generation consoles used increases in technology to give us better gameplay than before. This is easily visible in going from Atari 2600 -> NES -> SNES -> N64. The Atari was actually capable of very little (but was fun), while the NES had full graphics capabilities, but low color support. Jumping to the SNES provided tons of color, scaling, rotation, and other features that made games more fun. The N64 proved that 3D environments didn't have to be boring, linear, or only for shooting zombies (or demons as your preference may be). For example:
Zelda -> Zelda III: A Link to the Past -> Zelda 64
Contra -> Contra III
Super Mario Bros. (I-III) -> Super Mario World -> Mario 64
StarFox -> StarFox 64
Today's games, OTOH, are mostly just regurgitations of the FPS. Doom was a lot of fun when it came out, Quake was a hackers dream, and Quake III made blasting your buddies the best thing since sliced bread. (Unreal Tournament wasn't bad either.) But it really gets old after awhile. How many times can you run around shooting the same bad guys with the same tired weapons? Where's the new game play frontiers? While consoles were screwing around, I had fun playing RTSes on my computer. Or flying a starship in Bridge Commander. Or driving mechs around. i.e. Varied and interesting game play. Sadly, even that has disappeared on the PC.
Where's the gaming goodness? Where's the pointy sticks? Where is the Coconut Monkey!?!
While I realize that the gaming industry thinks that games are Hollywood productions, I honestly think fun games require nothing of the sort. Sure, I'd love to see another Wing Commander game with Mark Hammil and Tom Wilson, but that's not what the gaming industry is producing. What we need is for games to again break out of the mold and try new things. Keep riding the bleeding edge of gaming. It doesn't have to be an expensive game, just a *fun* one.
Tell me something: Why do games today *have* to be something I can't let my 5 year old son play? He still plays the old Nintendo games I used to play as a kid. He thinks they're a lot of fun. Yet do you think there's a chance in hell that I'm going to sit him in front of Doom III or an X-Box? No way! Why have we eschewed Gaming Goodness(TM) for violence and call it fun?
Maybe it's just me. Maybe I'm getting old.
Javascript + Nintendo DSi = DSiCade
I just want a lazy-susan thingie on the bottom of my new XBOX360 so I can rotate it the promised 360 degrees...
You can't talk about Wikipedia's flaws on Wikipedia
This just in: PC Hardware site blasts consoles while citing anonymous "sources" and blatant factually incorrect claims (for instance, PPE core = Xenon core).
Developers atuned to developing for PCs with their out of order execution and high general-purpose performance port their code quickly to these in-order CPUs that rely on multiple threads for performance, and find that the performance isn't blistering!
It turns out they'll need to make more efficient code, as Xenon/Cell forgo lots of transistors that make horrible code perform better.
Gag me...
And this is news? The console makers have been doing this for years. Remember when the PS2 was announced and we were told of its "Toy Story Quality Graphics Rendering"? Same thing with the infamous "Mode 7" in the Super NES system. Who can forget the So called 16 bit TurboGrafix 16? As I stated above, the console makers have hyped up every system that has ever been released and all have failed to meet the hype that preceeded them...
[n8.r0n] http://petesweb.spymac.net/
Physics processors came too late for this generation of consoles. This will really put PCs over the top. This should be coming out by the end of the year.
Transcend Humanity. Please.
Slashdot needs emoticons, if just so we can pretend to be shocked.
Pulp Audio Weekly - Geek News and Reviews
We don't know what any system will cost.
I think the real problem is each time you push for more improvements, the more complex the architecture gets. The article said that most developers would be using only one of the PS3's processors for most operations. Well, when you're used to designing for one processor, you tend to continue designing for one processor.
Each new feature added to the console requires learning that developers for past consoles, who have been used to the last console, will do slowly, and maybe reluctantly.
What developers really want is the *exact same* architecture, but much faster, more memory, etc. No more processors, no more complex ways of addressing different caches. Just make the thing the same, only faster, and developers would love it. Initially...
However, a year from now, the developers will learn the basics of the new consoles, and want something more. Then they will get into all those features that the new architecture gives them, and be excited to be the first to make a game that has realistic crumbling concrete when the tank slams into a wall, or whatever else they decide to do.
But asking a developer now about how their next gen console devkit performs is premature.
"Scientists don't change their minds, they just die." -- Max Planck
We need a 'Duh.' rating.
According to the article, both console's CPUs will be, for real-world applications (and not silly benchmarks) about that speed. Twice as fast as the Xbox.
Interestingly though, the article also says that the two GPUs (which are again nearly the same in performance) will be much better than their predeccesors. The other components will be fairly improved as well, so overall the consoles will be over 2x as fast as Xbox 1. Not as powerful as the manufacturers claim, of course, but still a good improvement over the last generation of consoles.
On the other hand... Now Nintendo's claims that its Revolution will be "only" two or three times more powerful than the Gamecube don't seem so bad. I always root for the underdog, and I like their lack of crazy hype so far.
The article said that most developers would be using only one of the PS3's processors for most operations. Well, when you're used to designing for one processor, you tend to continue designing for one processor.
/.; I forget where I first saw it).
Not really surprising; at any rate, it may be essential to get used to this type of architecture/programming, as The Free Lunch Is Over, if this article is to be believed. (This may have featured in
"Slashdot - News and Chat Sites Deviant". (Click "homepage" link above for details).
Congratulations, we've all fallen victem to the same cruel joke every time some company decides to release a new console, product, etc.
And in the end, none of it matters if the games aren't very good anyhow. Now that we've gotten over the dick measuring contest to see who can spit out the most flops, maybe we can all get back to enjoying the games on our current generation systems, and hoping that we'll see even better games in the future.
I don't mean to sound like a troll, flamer, asshat, or other nasty forum lurker, but does the computational power of a console make or break it? I've got a cell phone that probably has more computational power than an SNES, but Legend of Zelda: A Link to the Past was a damned good game. Let's step away from our obsession with graphics and the raw-power of the machine and worry about other things.
Here's a few I can think of right now:
1) The Xbox 360 will still be using DVDs. Guess what, we have already managed to fill up a full DVD with some games. Because we're going with HD games now, that'll take up more space (or processor time if we compress it to save space) which we don't have.
2) The Xbox 360 is using 2.4GHhz wireless controllers last I heard. Not a bad concept, but what happens when the battery dies mid-game? What about the cost of batters that add up over time? What happens if I have some other 2.4GHz device such as a phone or wireless router in the near location? I'm not the most knowledgable about wireless communications, but could this cause some interference?
3) Backwards compatability might not be included. Every day I hear a different story. Please, someone tell me it's going to be there for sure. Shouldn't Microsoft be more worried about pissing off the installed customer base that they had to fight to get than trying to get a few more flops out of a processor?
Just my opinion, but let's focus more on the games than the hardware.
I know some people who run current-gen consoles thru scalers (or use their HD set's scaler) have issues with lag: microseconds between when a controller is actuated and when the effect is displayed onscreen.
Scaler folks have had issues with HD upconversion lag when it comes to, say, DVDs. However, many HT receivers will let you customize your audio delay to compensate since lag should be fairly consistent. There's really no compensation for gaming, unless you're psychic.
Presumably, the next gen of consoles (along with decent GPUs in general purpose computers) will not have this issue since their output resolutions bypass scalers. However, some of the upcoming 1080p sets (Samsung at least) will not take 1080p via their HDMI inputs, so they'll deinterlace 1080i internally, and beyond picture quality concerns this may impact when it comes to lag. Or, use their RGB ins and suffer from D->A->D conversion.
That's what really scares me about the current game industry. Everyone's so focused on their pretty graphics that the rest of the games components, like physics, AI, or even gameplay are taking a backseat. They might as well put "produced by Jerry Bruckheimer" on half the shit games that come out..
:P
*P.S. Sorry to any Bruckheimer fans out there...but I couldn't think of a better all flash, no substance Hollywood guy at the moment
"A truly wise man realizes he knows nothing."
However, take the Anandtech article with a smaller grain of salt, too. I'm not sure which quotes from the article were attributed to final hardware and which were talking about the development kits (we already know that the Powermac xbox devstation is slower... or at least that's what one of the EA guys told me at E3). There was this quote: My guess is same can be said for CPU as well as GPU but that's a hunch.
Besides that, realize that the developers get much, much better at maximizing the hardware over time. When the SNES came out, developers complained that the extra colors and memory were pointles because the cpu was too damn slow (3.5 mhz, right?). 1st wave games had smallish sprites, tons of slowdown when things got busy, and many arcade ports only had a single-player option because 2-player bogged the hardware). Towards the end you had near-perfect ports of streetfighter 2, and full-color, parallax scrolling games with several large sprites like Donkey Kong Country. My hunch is that the 2nd wave games for 360 and ps3 will have similar gains.
It's still a really good article and worth checking out, but I'm not surprised in either direction.
Seems to me like all the games that were first out of the gate for the PS2 and XBOX were designed to wow with graphics. Great visuals, but weak and one-dimensional gaming.
Problem is, it seems to have shifted the whole mentality of game developers. Games seem to look good first, but play good second. On a whim i put away some of my PS2 titles and dug out the old PS1 stalwarts. The original Driver was still a kick in the ass. Breath of Fire III was amazing. FF7 was good, Grandia was good. For kicks i fired up my old K6-II and played older versions of Sim City (2K and 3K), Stronghold, Age of Empires, C&C were all so much more fun. It wasn't nostalgia either.
Paper Mario seemed like a great game too. The graphics were nice and clean, but not overly extravagant. But it was still a great game build up from many simple concepts. Just like the old days.
I hope that the hardware *does* stagnate, and maybe devs will stop writing 500 lines of code to control breast jiggle in the next Dead On Arrival and instead brainstorm some ingenuity into the games instead.
It doesn't have to wow me with graphics. Wow me with fun!
</rant>
do() || do_not();
Not so much because of its average-case power, but because of what happens when you pull out all the stops and optimize some game like crazy for it. Look at the PS2-- really weak machine in a lot of ways, but when someone who knows how to really harness the hardware makes a game for it you periodically get an Ico or Metal Gear Solid 3 or something where the graphics just absolutely blow you away. The Cell looks to have the same tweakability features of the Emotion Engine, only times like a thousand.
:P
I also want a Cell just, like, to play around with. They say Linux is running on this thing? Awesome. I just want to play with the microchip and see what I can get it to do. OK, yeah, I'm something of a compiler junkie. Blah
Irritable, left-wing and possibly humorous bumper stickers and t-shirts
for the lazy:
d =140
http://www.pcper.com.nyud.net:8090/article.php?ai
From the link:
---
What AGEIA and even game developers envision a PPU will enable for a gamer is a world with physics unlike anything we have seen in a real time game before. We are talking about thousands of rigid bodies, real flowing water, hair simulation, avalanches of rock, clothing simulations and more. Even more impressive is the idea of a universal collision detection system that allows you to interact with absolutely ANYTHING in a game world. All of it calculated in real time with nothing scripted in the game engine.
Sure you might have seen some explosions in a game you have played before, ones that might destroy an entire building. In nearly all cases, those have been scripted, meaning the debris and fire and dust were all created specifically for that explosion scene. Their motions and reactions were probably all scripted so that they went in a particular direction at a particular time and a particular speed. But what if you could have the option of changing that? What if you could have the explostion of a dam on a river be changed in real time depending on YOUR placement of the explosives? You might place them on the very center of the dam, creating a big hole that water rushes through, or instead you might only use a small amount of explosives to destory a small side portion and let water move out more slowly and let the water pressure be the force that eventually destroys the entire dam.
Damn. That would be a cool scene, and I didn't even see a demo of that -- just made it up!
---
(end of snip)
Nintendo has been pretty honest in the past as to their actual performance specs... and if what they say about being roughly 2 to 3 times more powerful than the cube is true, that puts them neck and neck with the XBOX360 and PS3.
That along with the ability to download old games makes me, if anything, more excited for Nintendo's new offering than the phony specs for XBOX and PS3 ever did.
now we just have to hope that they don't pull.. well a Nintendo and do something totally freaky with their controller. To be honest, I have high hopes.
While I realize that the gaming industry thinks that games are Hollywood productions, I honestly think fun games require nothing of the sort.
Urgh. Never understood why people thought Hollywood was glamorous or in any way desirable.
But that's beside the point, which is that those in The Industry want it to be like Hollywood, because somehow that's Grown Up. This Shows that The Industry Has Matured. They want their prestigious awards. They want to be Just Like Movie Directors. It all smacks of insecurity.
It also smacks of driving themselves into a bloated hole where they now can't *afford* to take risks because the costs of game development are so high.
There will always be a market for unimaginative, glossy games, and there will always be the bottom line. But to treat this as an ideal is frankly twisted.
Games are *not* (or should not be) like films. Films are not interactive. Games are. Imagine what the film industry would have been like if Directors had been in thrall to still photography.
"High production value" cut-scenes are bullshit. They aren't interactive, and they jar with the style of the rest of the game; but they let bloated-ego software developers Compare Themselves To Hollywood.
If you want to apply production values like that, apply them to the game itself, not to cut-scenes, no matter how well-made.
Instead of playing wannabe Scorsese, those in the industry should be concentrating on the potential of *their* medium; to allow the player more freedom to do what they want to do (the path it would have been interesting to see them go down), to choose new and different styles of gameplay, rather than the same restricted gameplay in progressively better-rendered worlds. Cut scenes, by their very nature, are going to force gameplay through predefined points. It's all so..... old-fashioned.
Anyway, enough... yeah, I'm probably getting old, but this isn't so much about romanticisation of the past. It's criticism of the way that, rather than focusing on the way technology could open up exciting new avenues in gameplay, the Industry has concentrated on turning out (basically) the same old stuff, but with ego-bolstering production values.
"Slashdot - News and Chat Sites Deviant". (Click "homepage" link above for details).
AnandTech is talking like they've had access to both consoles and have tested extensively when it's all hearsay. You don't say things like "Although both manufacturers royally screwed up their CPUs..." on hearsay. It is extremely unlikely that MS and Sony would both be stupid enough to "royally screw up" on something so important to them. They also imply that IBM is stupid (or evil?) for selling MS and Sony on their inferior product. I find it extremely unlikely that one person over at Anandtech is smarter than Sony, MS, and IBM.
Also, as the article stated, the platforms were designed for extensively multi-threaded games, but no one is writing games that way. So... why are they surprised that it's (supposedly) slow? If I put the bread on top of the toaster it takes a lot longer than if I put it in the slots. That doesn't make my toaster slow, though, it makes me an idiot.
This article really seems to take the wind out of their sails regarding what's being boasted 'under the hood' and what it's actually capable of doing.
:)
But I look at a game like doom3 running on a xbox. Yes it's low res and yes I read their changed some of the levels so there isn't as much draw distance (like removing a window from a corridor etc).
But still, it's doom3 running on what is a 733 mhz cpu with ONLY 64 megs of ram and doing a pretty good job of it.
Whereas my p4 1.6 with only 128 megs of ram (really need to upgrade) and a gf4ti4200 runs doom3 like shit. Downright unplayable. Heck I wish I could have the xbox version of doom3 to run on my pos system.
My point? Well, history has shown that the developers will eventually make these systems do tricks that no one initially thought the systems were capable of. But the pc is such a moving target with so many configurations that we don't see near as much optimizations.
But I'm a pc gamer for life and mainly cause I hate exclusive agreements and would love to see these systems be a disappointment.
I miss the days (snes/genesis) where only 1st party titles were exclusive (mario vs sonic) and with pretty much all other titles it was may the best console win.
How much do they offer these developers to only play on one side of the fence? I think one of the biggest first exclusive agreements was tombraider on the ps1. But what I always liked was the pc was ignored in these agreements. Doesn't seem to be the case these days. Cough, halo, cough. And I'll never forgive the developers dropping the pc with the oddworld series. Ok way OT now I'll stop rambling.
One of the biggest limitations ended up being the meager 64MB of memory that the system shipped with.
One of the most important changes with the new consoles is that system memory has been bumped from 64MB on the original Xbox to a whopping 512MB on both the Xbox 360 and the PlayStation 3. For the Xbox, that's a factor of 8 increase, and over 12x the total memory present on the PlayStation 2.
One of the biggest limitations was the 64mb of memory - clearly too little. Now, five years later, they've increased that by a factor of 8.
*quickly does sums on fingers*
4.5 years = 18 months x 3
Didn't some guy come up with a rule about this? (My local library was all out of copies of that issue of the magazine)
2^3 = 8
So, five years on, they've managed to about keep pace with historic advancement, being relatively no better than the 64mb that was widely regarded to hamstring the last generation of consoles?
Sure, right now, 512mb sounds great... But then 64mb sounded good five years ago too.
HalfLife2's High Dynamic Range lighting model is expecting to need one to two gigabytes of system RAM to work properly. Sure, PCs run with a clunky OS but it's not that bad. Battlefield 2 needs 512mb minimum and prefers 1gb.
Five years ago, console fanboys dismissed PC gamers when they pointed out 64mb might be nice now but would barely cut it in two years and seriously hamstring the console in 4-5 - the lifecycle of a typical console. They were wrong then.
Now, five years later, all they've done is up that hamstrung amount in accordance with Moore's law and, once again, it seems fine for a console's release and is going to be a major issue well within the system's lifespan.
"If Sony and MS are overhyping their systems then so is Nintendo."
False! I don't know how old you were last console cycle, but Nintendo was very realistic about the Gamecube's abilities before it was released. They said, "it can render 9 million polygons per second under realistic conditions." Cue Sony: "Well ours can render ONE HUNDRED BILLION polygons per second!" and then microsoft: "We can do INFINITY BILLION TO THE INFINITY POWER!!!". So it isn't at all clear to me why the fact that Sony and MS overhype indicates that Nintendo overhypes.
Now, I'm no hardware wiz, so I can really only comment on this from the perspective of the average non-techie gamer, but... I've played the new (ie. unreleased) Need for Speed on the thing, and I must say that it looks damn sweet. Sure, maybe the article's right and the machine doesn't perform as well as it should, but as a gamer, am I going to notice the limitations? Is my gaming experience going to be impacted by this? Probably not.
Basically what I'm trying to say is that while the article is certainly interesting to the geek in all of us, saying that the processors are "Not Up To Hype" seems a bit too sensational given that the only people who will notice these minor failings are the developers who, one would hope, already know about them.
"What developers really want is the *exact same* architecture, but much faster, more memory, etc. No more processors, no more complex ways of addressing different caches. Just make the thing the same, only faster, and developers would love it. Initially..."
While we have yet to see exactly what Ninty has up their sleeves, this sounds like the Revolution. Same API, using IBM CPU's and ATi GPU's again just like last time...as long as they don't do some exotic IBM CPU like the Cell (co-developed by Sony so unlikely) or three pathetic CPU's working together to be okay like the XBox 360, it'll be basicly just a much much faster GameCube.
With wireless, free online play, thousands of downloadable games and game demos, and a new 'revolutionary' controller added, of course.
The fact that people are excited about Battlefield 2, which is yet another FPS war sim army-style, just blows my mind. I have a friend who's trying to justify it to me.
"No, it's great. See, the graphics are amazing, and the netplay is wonderful. Now, you spawn on your team leader, and you all work together. It's brilliant!
My response, "So it's yet another Doom clone with new spawn rules and a graphics update. Yee-haw. Know what I was playing? Katamari Damacy and Way of the Samurai 2." Trying to explain to him these games, let alone show them to him, is an utter waste of time. He walks out at the title screen, claiming he can't stand graphics so "old".
It's really depressing, because as long as there are people like him, we're going to see more games like EAInsert-Sport-Here 200X, Halflife 2 (Just like Halflife 1, but more so).
Slashdot. It's Not For Common Sense
A couple years ago Jason Rubin (of Crash Bandicoot/Jak&Daxter fame) gave a speech about Hollywood that seems to have been wildly misinterpreted. He likened the current state of the video game industry to the packaged goods business. People aren't buying the content of games, they're buying the box. They're buying the marketing, the [evil] publisher. The [evil] publisher wants it that way, they want to remove the public's association with talent from the purchasing of the game... they want consumers to think that all developers are the same and let hype take care of the rest (at this pointed he pointed out that Crash games are still being made, but not by Naughty Dog).
He then mentioned that if the top 300 game developers got in an airplane and it crashed, the industry would be set back a decade. If the top 300 marketing people fell into the same misfortune, the Industry wouldn't miss a beat. People hooted and cheered at this irony... laid out so eloquently, between where the publishers place the importance of moving products with where the real importance was.
He then confused a lot of people, talking about Hollywood is the future and getting invited to parties, and so that is what a lot of people walked away with... However, the real crux of the passionate speech was that Game Companies, not publishers, belong in big bold letters on the box. Game development is a talent industry, not a packaged good... Game Designers who consistently design good games deserve the same name recognition and the same selling power as the equivalent Hollywood celebrities, Robert Deniro, Kevin Spacey, etc. with their name Right There on the Box in the same way that Hollywood movies are marketed (And that there are more people making good games than just Will Wright and Miyamoto). Until developers make those demands, publishers will feel free to keep marketing and unloading the same crap on the unsuspecting public.
It's too bad you're making this stuff up. There are various rumors going around (quad core 2.5 GHz X360-type processor, dual processor 1.8 GHz G5's, etc, etc) but there is absolutely NO reliable info on the Revolution's processor. None.
Please provide a link if you're absolutely sure there is. I've been mongering every scrap of information released on the next gen console and I can guarantee Ninty isn't "expected" to have any kind of processor on the Revolution- Nintendo has kept it completely quiet.
So basically while seeing a lot of "this is bullshit" comments, we're not seeing any comments from anybody who really knows or has worked with either of these two platforms. Instead, we're seeing people more willing to believe MS and Sony who have everything to gain from lying about their products vs. a more realistic view of two over-hyped machines by a website who will attract viewers to their article whether they say good things about these two consoles or not. It really will make no differenc to Anandtech. People will come to read their articles because they've earned a readership so they've no real motivation to make stuff up or distort things.
Admit it people, some of you just don't want to hear what they're saying. Had they said that the PS3 does put out 2 teraflops and the XBox 360 only one, then you could have simply continued on with the normal console flame war which has been going on since E3 ie 3 cores vs 7 SPEs, etc. Then of course, there'd be doubters from the other side accusing Anandtech of being on the payroll of MS and Sony.
Look at the motivation people. Think about who's really got cause to BS the console gamers.
"It is accurate that at this time we will not support high-definition [on Revolution]," Kaplan told IGN.com.
It's really hard to tell what will happen by the time it's released. The Gamecube is theoretically capable of 720p output, though the games only utilize 480p. Considering the video hardware that is being used, it's safe to assume that the Revolution is at least as capable as the Gamecube It's not going to matter all that much, because we're still going to be stuck with 480p DVD movies for a while. 480p is a form of SDTV. Even if it's not "HD", it's still much higher quality than any analog television. Your comment about the RCA analog television is grossly exaggerated.And let's be honest... All three systems will have hardware that's paractically the same, regardless of these cracked out specs and numbers (ironic isn't it that all three are using what is essentiall a next-gen Gamecube with PowerPC and ATI graphics). What it will really boil down to is the games.
Here's your source!
http://cube.ign.com/articles/522/522559p2.html
Q: Is Revolution "two-to-three times more powerful than GameCube"?
A: USA Today reported this news based on a comment from Nintendo of America's vice president of corporate affairs, Perrin Kaplan. The information was later determined to be false. We do not yet know how much more power Revolution wields over its predecessor.
Developers pushed the GC to over 14 million shortly after it was released. (I think it was in one of the star wars games). The numbers that Nintendo was putting out were not only realistic; they were slightly conservative.
Contrast with Sony and MS whose claimed performance numbers for the last 6 years have been pure fantasy and/or hype.
"PS3 article is pulled for now because Anand is worried about MS tracing his anonymous insider."
(minus page 6 about the GPUs, it got squased in my cache when I tried linking back after it was pulled)
In our last article we had a fairly open-ended discussion about many of the challenges facing both of the recently announced next-generation game consoles. We discussed misconceptions about the Cell processor and its ability to accelerate physics calculations, as well as touched on the GPUs of both platforms. In the end, both the Xbox 360 and the PlayStation 3 are much closer competitors than you would think based on first impressions.
The Xbox 360's Xenon CPU features more general purpose cores than the PlayStation 3 (3 vs. 1), however game developers will most likely only be using one of those cores for the majority of their calculations, leveling the playing field considerably.
The Cell processor derives much of its power from its array of 7 SPEs (Synergistic Processing Elements), however as we discovered in our last article, their purpose is far more specialized than we had thought. Speaking with Epic Games' head developer, Tim Sweeney, he provided a much more balanced view of what sorts of tasks could take advantage of the Cell's SPE array.
The GPUs of the next-generation platforms also proved to be quite interesting. In Part I we speculated as to the true nature of NVIDIA's RSX in the PS3, concluding that it's quite likely little more than a higher clocked G70 GPU. We will expand on that discussion a bit more in this article. We also looked at Xenos, the Xbox 360's GPU and characterized it as equivalent to a very flexible 24-pipe R420. Despite the inclusion of the 10MB of embedded DRAM, Xenos and RSX ended up being quite similar in our expectations for performance; and that pretty much summarized all of our findings - the two consoles, although implementing very different architectures, ended up being so very similar.
So we've concluded that the two platforms will probably end up performing very similarly, but there was one very important element excluded from the first article: a comparison to present-day PC architectures. The reason a comparison to PC architectures is important is because it provides an evaluation point to gauge the expected performance of these next-generation consoles. We've heard countless times that these new consoles would offer better gaming performance than anything we've had on the PC, or anything we would have for a matter of years. Now it's time to actually put those claims to the test, and that's exactly what we did.
Speaking under conditions of anonymity with real world game developers who have had first hand experience writing code for both the Xbox 360 and PlayStation 3 hardware (and dev kits where applicable), we asked them for nothing more than their brutal honesty. What did they think of these new consoles? Are they really outfitted with the PC-eclipsing performance we've been lead to believe they have? The answer is actually quite frequently found in history; as with anything, you get what you pay for.
Learning from Generation X
The original Xbox console marked a very important step in the evolution of gaming consoles - it was the first console that was little more than a Windows PC.
The original Xbox was basically a PC
It featured a 733MHz Pentium III processor with a 128KB L2 cache, paired up with a modified version of NVIDIA's nForce chipset (modified to support Intel's Pentium III bus instead of the Athlon XP it was designed for). The nForce chipset featured an integrated GPU, codenamed the NV2A, offering performance very similar to that of a GeForce3. The system had a 5X PC DVD drive and an 8GB IDE hard drive, and all of the controllers interfaced to the console using USB cables with a proprietary connector.
For the most part, game developers were quite pleased with the original Xbox. It offered them a much more powerful CPU, GPU and overall platform than anything had before. But as time went on, there were definitely limitations that developers ran into with the first Xbox.
One of the biggest limitations
On a purely hardware level, ATI's Xbox 360 GPU (codenamed Xenos) is quite interesting. The part itself is made up of two physically distinct silicon ICs. One IC is the GPU itself, which houses all the shader hardware and most of the processing power. The second IC (which ATI refers to as the "daughter die") is a 10MB block of embedded DRAM (eDRAM) combined with the hardware necessary for z and stencil operations, color and alpha processing, and anti aliasing. This daughter die is connected to the GPU proper via a 32GB/sec interconnect. Data sent over this bus will be compressed, so usable bandwidth will be higher than 32GB/sec. In side the daughter die, between the processing hardware and the eDRAM itself, bandwidth is 256GB/sec.
At this point in time, much of the bandwidth generated by graphics hardware is required to handle color and z data moving to the framebuffer. ATI hopes to eliminate this as a bottleneck by moving this processing and the back framebuffer off the main memory bus. The bus to main memory is 512MB of 128-bit 700MHz GDDR3 (which results in just over 22GB/sec of bandwidth). This is less bandwidth than current desktop graphics cards have available, but by offloading work and bandwidth for color and z to the daughter die, ATI saves themselves a good deal of bandwidth. The 22GB/sec is left for textures and the rest of the system (the Xbox implements a single pool of unified memory).
The GPU essentially acts as the Northbridge for the system, and sits in the middle of everything. From the graphics hardware, there is 10.8GB/sec of bandwidth up and down to the CPU itself. The rest of the system is hooked in with 500MB/sec of bandwidth up and down. The high bandwidth to the CPU is quite useful as the GPU is able to directly read from the L2 cache. In the console world, the CPU and GPU are quite tightly linked and the Xbox 360 stands to continue that tradition.
Weighing in at 332M transistors, the Xbox 360 GPU is quite a powerful part, but its architecture differs from that of current desktop graphics hardware. For years, vertex and pixel shader hardware have been implemented separately, but ATI has sought to combine their functionality in a unified shader architecture.
What's A Unified Shader Architecture?
The GPU in the Xbox 360 uses a different architecture than we are used to seeing. To be sure, vertex and pixel shader programs will run on the part, but not on separate segments of the hardware. Vertex and pixel processing differ in purpose, but there is quite a bit of overlap in the type of hardware needed to do both. The unified shader architecture that ATI chose to use in their Xbox 360 GPU allows them to pack more functionality onto fewer transistors as less hardware needs to be duplicated for use in different parts of the chip and will run both vertex and shader programs on the same hardware.
There are 3 parallel groups of 16 shader units each. Each of the three groups can either operate on vertex or pixel data. Each shader unit is able to perform one 4 wide vector operation and 1 scalar operation per clock cycle. Current ATI hardware is able to perform two 3 wide vector and two scalar operations per cycle in the pixel pipe alone. The vertex pipeline of R420 is 6 wide and can do one vector 4 and one scalar op per cycle. If we look at straight up processing power, this gives R420 the ability to crunch 158 components (30 of which are 32bit and 128 are limited to 24bit precision). The Xbox GPU is able to crunch 240 32bit components in its shader units per clock cycle. Where this is a 51% increase in the number of ops that can be done per cycle (as well as a general increase in precision), we can't expect these 48 piplines to act like 3 sets of R420 pipelines. All things being equal, this increase (when only looking at ops/cycle) would be only as powerful as a 24 piped R420.
What will make or break the difference between something like a 24 piped R420 and the unified shaders of the Xbox GPU is ho
:).
:)
:) I never have written a single statement for a console, but reading about how they're programmed it's similar to old amiga hardware as in: utilize the different hardware to get as much out of it as possible. That wasn't hard, it was FUN :). Good to know there are still people out there enjoying that kind of work :)
Nothing is more rewarding than fiddling with hardware registers, parallel execution lists and then... finally... get something visually on the screen
reading your post made me think back to the old demoscene days on the Amiga 500.
Never underestimate the relief of true separation of Religion and State.