Got this from another BBS who was listing this stuff as the gospel, but there were some things that seemed to me technically incorect, but I'm not a hardware guy. Anyone know if any of this is true or false? Aren't they comparing the PS2's graphics processor to the XBox's genral processor instead of the XBox GPU? I understood the 733mhz CPU just left to do AI and whatnot. Is this, or any of this article entirely accurate or is it just some biased PS2 fanboy stuff?
------------------------------------------------------------------------------------------------------
What is important here is to explain what each part of an
architecture actually does and what it results into. "Call it whatever you want, but the 4mb on the GS performs all the
duties of vram.[08-01]". Well, as I will show further down this isn??t right.
How ever, what is much more important is to understand why I say that. For that we need to take a little look on how
the videoRAM and bandwidth structure is modelled on traditional architectures contra the new PS2 architecture.
A traditional rendering architecture has one memory bank known as videoRAM. Tasks of a traditional videoRAM
architecture:
1) To send and receive textures to and from the graphics chip (rendering).
2) To send and receive Z-coordinates to and from the graphics chip.
3) To store permanent textures.
4) To store the framebuffer (the picture we see on screen).
5) To store the Z-buffer (information the graphics chip needs to render).
The down side to the videoRAM is that it has a relatively slow bandwidth. It can??t ship and receive information very
fast. How ever if we could turn of all the render effects and Z-buffer and simply let it concentrate on streaming the
textures to the graphics chip it would save trumendeous amounts of bandwidth. For instance. The PSX has quite high
polygon performance but very few rendering effects and no Z-buffer. It only needs 130Megabytes of bandwidth.
Nintendo 64 has less polygon performance but a lot of rendering effects and Z-buffer. As a result it has to use
500Megabytes of bandwidth.
So what if you could create an architecture that actually could turn on all render effects plus Z-buffer (aka Nin64) but
never the less use as little bandwidth as a system that doesn??t have this (aka PSX). The saying goes you can??t eat the
cake and keep it at the same time. Normally this might be true. But not for PS2 because this is exactly what it does.
So how does it do it? I??ll come to that but first a little vital explanation. The ideal videoRAM is supposed to have
enormous bandwidth as well as huge storage capablilities. The problem is it isn??t cost effective to combine these
basic factors. So the higher storage capabilities the less bandwidth capabilities and vice versa. Never the less this
technology is what all computers and videogames are using because it is the most simple and cost effective one
around. This gives moderate storage and bandwidth capabilities. In the case of the PS2 this is not enough. This way
the texture as well as render capabilities would be very limited. What Sony did instead was to devide the videoRAM
architecture into two separate parts: One memory block with huge storage capabilities and quite high bandwidth
capabilities (This is the part that all computers and videogames have). But then Sony also placed one small 4
Megabyte block of memory directly onto the graphics chip with enormous bandwidth capabilities. The small block
was made big enough just so it could take over the tasks that demand high bandwidth. This way Sony got both high
storage capabilities and speedy bandwidth.
The huge memory block can now dedicate all its bandwidth to simply send textures to the 4Megabyte buffer. Now,
this means a lot of textures. Then the 4Megabyte buffer takes over all the rendering and Z-buffer check, something
the huge memory block is quite bad at. This means very advanced rendering. As I stated before (07-28) the PS2
textures are stored in the 32Megabytes of mainRAM.
Tasks of the PS2 videoRAM (the huge memory block):
1) To send textures to the 4Megabyte buffer (streaming).
2) To store permanent textures.
Tasks handled by the 4Megabyte buffer (the small memory block):
1) To send and receive textures to and from the graphics chip (rendering).
2) To send and receive Z-coordinates to and from the graphics chip.
4) To store the framebuffer (the picture we see on screen).
5) To store the Z-buffer (information the graphics chip needs to render).
Peter said the 4 Megabyte buffer "performs all the duties of vram". That is almost true, but not completely. Instead it
takes over the exact tasks that it is best suited for. That includes almost everything that needs speed. But as I said
above, you can't make a videoRAM both fast and store a lot at the same time. So it skips the storage... and leaves that
for the much bigger main videoRAM block.This way the PS2 can devide the burdens more accurately to the exact
parts most potent to handle them. In other words the PS2 has a more refined and specialized architecture. It is
somewhat complicated but it makes the PS2 a very powerful texture machine with enormous rendering capabilities.
"People, as much as it hurts to hear it, the fact of the matter is that the PS2 architecture was poorly designed. [08-
01]". The PS2 is as poorly designed as a NASA space shuttle ready for launch. The traditional way of programming,
handling and managing a game doesn??t work on the PS2. Never the less this is exactly what the programmer
community is used to. The whole architecture has been rearanged to combine extreme computer power with very high
efficiensy. It has made it very alien to most programmers. If a programmer never the less choses to use the PS2 the
usual way it becomes a very unbalanced architecture. To program like for a PC kills most of its advantages. You
would then have 32 megs of main RAM combined with about 4 megs of videoRAM. It would be just as efficient as
using a hammer for a screw driver. It doesn??t translate into a good game.
Daniel
More technical talk
The PS2's architecture is not poorly designed. You should believe everything you read in Microsoft press releases. Is
it difficult to program for? Yes, because it's different from what was up until now "the norm". But why is it different?
That's the important question, why did Sony go with bandwidth over size? Because the PS2 was engineered to do one
thing, and one thing only, media processing. Most every other computer system out there, including the Dreamcast,
the Xbox and the Dolphin/StarCube, are built around computing principals developed in the early days of personal
computing when all that mattered was how fast you could run a word processor or spreadsheet. That's what we call
baggage. Such systems were rapidly approaching a ceiling of performance in 3D/media applications because that's not
what they were, really, designed to do. Media processing is fundamentally different from such static applications. In a
word processor you perform lots of different operations on a relatively small piece of data. Conversely, in media
processing (3D rendering, video/sound manipulation, compression/decompression), you're basically doing the
opposite, performing the same few operations over and over again on a huge set of data. And that is what the
engineers at Sony had in mind when they came up with the Emotion Engine and Graphic Synthesizer. The PS2 is
uniquely adept at media processing in ways that no traditional architecture is. Sure, there are some pretty fancy 3D
accelerators and sound chips out there in the PC world and in the Dolphin/Xbox/Dreamcast, whatever, but they, and
things like SIMD instructions tacked onto Pentiums and Athlons and PowerPCs, are simply symptomatic of, and
stop-gag solutions to the core problem, that the traditional PC architecture, used in basically ever system out there,
save for the PS2, was never intended to do such media processing feats.
To suggest that Sony just blew it on the PS2's design is not only naive, but preposterous. They didn't accidentally end
up with a machine hard to program for. And is it really hard to program for, or just different? We know Sony can
make an easy to develop for system, the original Playstation was just such a platform. They chose to go with a non-
traditional style of system to avoid running strait into the world of diminishing returns.
But, I digress. Peter posed a question. Would programmers prefer a large amount of slow memory to store textures
in, or a small amount of incredibly fast memory? Right now, if you ask them, most will probably indicate the former,
but only because that's what they're used to, not because it is inherently superior. Truth be told, I can think of many
situations where high band, small space wins out over low band, big space. Again, in the example Pete gives us. You
can load all 64MB of textures at one time, but that takes time. I thought load times were a no-no... And what happens
when you move to a new area, the game engine screeches to a halt while new textures get slowly uploaded to the
VRAM. That isn't a problem in a properly implemented texture streaming scheme on a system like the PS2. And it
never will be.
So which is really the better way of doing things?
Brad Grenz
YOU GUYS STILL WORRIED ABOUT THE LACK OF VRAM ON THE PS2!
I've read plenty of letters from people whining about the fact that the PS2 has 'only' 4 MB of
texture RAM as opposed to DC's 8 MB, while ignoring the fact that the PS2 is capable of
rendering 75million polys/sec, while DC can only do 3 million - obviously fanboys who've
realised that the anti-aliasing thing is a non-issue, because the ps2 is capable of handling all 3
methods of anti-aliasing, the only reason why developers have not put this feature in there games
yet are because they were completly rushed so that they can get if done in time for the japanese
ps2 launch, but dont be surprised if they implement AA in there games for the North American
ps2 launch. However, it doesn't seem as if any of those making these complaints have bothered
to read the official PS2 specs, which can be found on Sony's North American Playstation site; if
they had, they'd have seen that the hardware MPEG decoder - used for decompressing DVD
images - can also be used for compressing/decompressing static bitmaps - Sony state themselves
that this can be used for textures. Presumably, it will be a similar process as that used for JPEG
images - though via hardware, not software - which means that the compression/image quality
ratio, if applied correctly, could effectively give a texture storage capacity twice as great as that
of DC, and would look better then textures compresses in dolphin with S3, and with minimal
compression artefacts present! I expect we'll see the greatest compression on 'natural' textures
(such as rock faces, sand, carpet etc), where loss of quality will be none. Remember, also, that
PS2 will be able to stream textures from the CD/DVD in realtime, without a processor hit.
Sounds good to me, that's all I can say!
SCEI recently unveiled a PlayStation2 demonstration described as the 'Spirit of the Orient.' The
demo included several character models wearing various styles of clothing, and illustrated the
realistic attributes that can be attributed to garments on the PlayStation2. According to company
representatives, future PS2 titles will feature clothing that can be creased and even torn in-game.
Furthermore, the PS2 Emotion Engine will calculate natural factors such as wind and their
effects on clothing in real-time. Thereby clothes will react dynamically with their surroundings.
Insiders suggest the aforementioned effects will be incorporated into the 'second-generation' of
PlayStation2 titles later this year.
I was reading an article saying that the X-box's main processor was 2x more powerful than the
PS2's. This is VERY wrong. The PS2's processor is over twice as powerful than the x-box's
because the ps2's is equal to about 2 733 pentium3 processors and the x-box's is comparible to
about a p3 650mhz. Besides, the 800mhz p3's can barely do a gigaflop and the ps2's emotion
engine can do 6.2gflops! And the XBOX uses a NON-custom made chip by intel which is a P3
600Mhz, lets not even talk about dolphin while it doesnt even compare to XBOX and the ps2 is
much better in HARDWARE QUALITY than the XBOX!!!
1) The PS2 300 MHz Emotion Engine is much more powerful than 2 PIII 600 in float point
calculation (about 4x faster) because there is 2 Vector Units in the EE and each of them have 10
float point units within. So I really wonder how a PIII 600(xbox) can catch up with the Nvidia
video processor, which can push 300,000,000 triangles per second. Of course, Intel can create a
special version of PIII for the X-box but that will never happen.
2) XBOX has 64MB of unified ram, which means that's all the XBOX has to offer in terms of
ram, so it will be used for sound, textures (using unified ram for textures is not as good as using
seperate VRAM for textures even if you do have 4MB), calculations, OS (YES AN
OPERATING SYSTEM!! HOW STUPID IS THAT, AND MICROSOFT SAYS THEY ARE
DEVOTED TO MAKING XBOX A VIDEO GAME ONLY CONSOLE) etc.Share the video
RAM with the Main RAM bus is not a good idea. Actually, it is not new to the PC. Check some
motherboard's with video card's built-in, they use the main RAM as video ram and it will
actually slow down the performance. It is because the CPU and the Video cannot access the
memory bus at the same time, and that can cut the 400 MHz Direct Rambus RAM bus
performance to half when both CPU and Video chip need to get data at the same time. In the
other hand, although PS 2 is using 200 MHz version of Direct Rambus RAM, which is half the
speed of XBOX, the real world performance will not be 1/2 because there is 4 megs of video
Ram built-in in the Video chip (Graphics Synthesizer) with a Ultra wide internal bus (2560
bit/s). That means there will be no performance hit at all and both the EE and GS can run in full
speed without spending the time for waiting for the RAM bus to get ready, where is XBOX will
have to wait for its ram to free-up, but if developers program their games to wait for the ram
then you cant get as detaile textures, sound, etc. but if they program the game to not wait for the
ram you will see a huge performance and processor hit and the game will run at a much lower
frame rate then programmed for. So, the RAM performance should be really really close
between X-box and PS2 in overall, PS2 mite have the upper hand. So the 64MB's of unified ram
is not a problem at all to the PS2.
3) X-box will use a special version of window 2000 as the OS for X-box. That will be a very big
mistake. X-box gave a huge 64 MB's of RAM, but this is base on the standard of a game console,
not for the unit to run windows. It can easily eat up 16 Megs for memory just for the core
system.
So, will X-box doom PS2? NO!!!
OTHER WEAKNESSES IN "XBOX"
IGNPS2 INTERVIEW WITH NAUGHTY DOG - THIS IS JUST SOME OF THE QUESTIONS
-CHECK OUT THE WHOLE INTERVIEW HERE
Question: Do you think there will be any major advantages to the Dynamic hardware setup of the
PS2 as opposed to traditional PC-esque hardware, it seems to making mainly problems for
developers?
Jason Rubin: Yes, I do think that there are advantages to the hardware setup of the PS2. The
multiprocessor setup of the system allows the PS2 to do so many things at the same time (if
programmed well) that the hardware is actually far more powerful than its spec. They say the X-
Box is going to be either 600mhz or 1000mhz. The PS2 is only 300mhz. But if you can have
your background engine running in 10% CPU and 50% VU1 time, then you have far more
power.
That is how we are doing our programming for the next game. I think that the problem with the
current games (the ones you have seen) is that they have opted to take the easy way out and they
are only doing CPU code.
So think of it like this. They are actually leaving a significant part of the power of the system on
the table. We plan to bring that power into the game.
And yes, of course there will be more powerful systems in the future. That is how technology
works. But the X-Box may not be that system.
Question: In what ways do you think the PlayStation 2 will change gaming as we know it today?
Do you think that it'll be revolutionary?
Jason Rubin: The PS2 will first be evolutionary... and then later, revolutionary. First, you will
see games that are similar to the past, but a whole lot better looking and smoother. Later, you
will see games that do things that have never been done before.
Broadband will have a lot to do with the Revolutionary part of the cycle. We will go beyond
deathmatch, and beyond the glorified chat room of EverQuest and get true multiplayer online
gaming with interaction. That is something we haven't seen yet.
The Xbox is actually slower than the PS2, if you think about it.
They conveniently forgot to mention the vector units on the PS2 when they compared the stats.
You know, that's like comparing cars to each other, but not mentioning one of them has a
Corvette engine and the other an old piece of junk engine.
Anyone who looks at Xbox 700mhz and PS2 300 mhz and immediately thinks "Oh, xbox is
faster!" is not thinking clearly. Xbox doesn't have vector units like PS2, and it's slow, pokey 6.4
GB/sec unified RAM used as VRAM can't compete with PS2's lightning fast 48 GB/sec RDRAM
(yes, PS2's VRAM is 7.5 times as fast as the Xbox's). Okay, yes, it has 4 MB, but I read in an
interview with Hideo Kojima that said the RDRAM is so fast, it really doesn't matter.
Know how many gigaflops PS2 can do? 6.2 gflops. A 700 mhz Pentium III can't get close to that
(what was it, like 3 gflops on a lucky day?). Don't forget the Xbox uses a **PC** processor, not
a console's! Just look at the N64 (100 mhz) and PSX (33 mhz) and compare them to computers
of the same speed. If you take the same mhz::speed ratio between PC CPU's and console CPU's,
the PS2 mops the floor with the Xbox. As you can clearly see, when it comes to consoles, MHZ
is NOT everything. There are plenty of other factors in there. Unfortunately Microsoft thought
mhz was everything, and made a terrible mistake.
Sure the Xbox has all those pretty specs, but are they too good to be true? Of course, remember
that it uses unified memory, so that memory has to be made to work with everything: sound,
system, video. It is un-specialized for specific purposes. On the Xbox, there is no DEDICATED
video RAM sitting right next to the graphics processor. There is no DEDICATED system RAM.
No DEDICATED sound RAM. It's all kind of in the middle, and it's too slow for the video.
That's the same problem the GeForce 2 has: Fast graphics chip but the slow memory bottlenecks
the whole thing and you don't nearly reach those specs nVidia was hyping up. Don't try to argue
that Xbox will be leaps and bounds faster than any normal 700mhz PC just because Windows
isn't there. Windows better be pretty freaking unbelievably sluggish if it's going to do that. If it
really is, why don't developers still make DOS games these days? Windows is a little sluggish,
but removing it is not going to make a PC 700 mhz CPU suddenly behave just like a console's.
Xbox is just a low price gaming PC, and thus it will have PC bottlenecks. The PS2 is custom
designed for the specific purpose of games.
PS2 will be a year ahead of Xbox, giving it plenty of time to build up a large library of games
and user base. Don't forget it can run hundreds of legacy PSX games in enhanced mode as well,
and functions as a DVD player. It's incredibly useful right out of the box. When the Xbox comes
out, it will only have a few launch titles, and that's it. DVD? Okay, I've heard that a lot of people
are buying the PS2 just for the DVD player, so why would they want to get an Xbox for
something they already have? There's no doubt the PS2 will reach $200 or $250 by the time
Xbox is out, so if they wanted a 2nd DVD player, getting another PS2 would be more logical.
After all, it'll have way more games and peripherals. Why would you want to spend $300 on a
new console with few games after you bought a highly useful one for $300 just one year ago? Or,
if you're choosing between two, why buy one that's ::ahem:: LESS powerful, has a tiny library of
games, and is more expensive? Exclusives, what exclusives? Ported PC games? Sheesh.
And let's not forget the target markets. PS2 is going for the general market, a wide range. Xbox
is going for the "hardcore" audience (who are the minority and very likely would have a PS2
anyways). Nintendo will pick up the kids market easily with the Gamecube. It's not like Sony
won't have their share in kids, though. Even a lot of young kids these days have Playstations and
know and respect the name.
Brand names are another thing. Ask any person on the street if they've heard about "Playstation"
or "Nintendo". Very likely they would have. Then try Xbox and watch the confused look on their
faces. MS is launching the Xbox right when Playstation will be more popular than ever, and
Nintendo's bombshell Gamecube will steal fire from Xbox, since they're coming out very close
to each other. Really, when you think of a game console, do you think of Microsoft first? Well,
in MS's case you might think of the DOJ trials and a Windows computer crashing.
$500 million marketing blitz is not going to work. True, after being bombarded with
commercials for Xbox, it will generate some interest. But that interest will be quickly wasted
away when the gamers (or parents) go to the store to check it out and sees a pitiful rack of Xbox
games compared to the huge PSX/PS2 library. And just because it's Microsoft doesn't mean
they're going to win. Remember the battle between PalmPilot and WinCE? Who won that battle?
How about Quicktime vs. AVI, who won there? Realmedia vs Windows Media? ActiveX vs Java
and Macromedia? AIM and ICQ vs MSN Messenger? Oh yeah, where are the EXCLUSIVE
GAMES, Microsoft? Where are your Zeldas, Crash Bandicoots, and Marios? Halo certainly
won't be, in fact it's coming out on Mac and PC. It might be old news by the time Xbox comes
out, unless MS ::cough:: "recommends" ::cough:: to Bungie that they delay it for computer
platforms until then. MS almost had a reason for computer gamers to buy an Xbox when they
bought Bungie and put Halo's availability into question, but they lost it when legions of angry
Bungie fans poured thousands of signatures into a petition saying "I'm NOT going to buy an
Xbox!" Speaking of games, where the heck are they? Have you seen the list on xbox.ign?
Pathetic. Even the developer list is meager.
In Japan, it's as good as dead. I was talking to a Japanese friend of mine over ICQ recently, he
said that nobody in Japan really cares about Xbox or Microsoft, they're all about Playstation.
I've seen a lot of Nintendo supporters bashing at PS2 recently. However, I think PS2 and
GameCube will go nicely together, just like PSX and N64 did. Stop fighting, and concentrate on
the true enemy: MicroBorg. They've assimilated your computers, and now they want to
assimilate your living room.
One last thing: How many of you, and how many people you know, held off on buying a
Dreamcast because the PS2 was on the horizon? Quite a bit I imagine. Now how many are doing
the same to the PS2 because of the Xbox? "Uh, X-what?" There ya go. Enjoy.
------------------------------------------------------------------------------------------------------
What is important here is to explain what each part of an
architecture actually does and what it results into. "Call it whatever you want, but the 4mb on the GS performs all the
duties of vram.[08-01]". Well, as I will show further down this isn??t right.
How ever, what is much more important is to understand why I say that. For that we need to take a little look on how
the videoRAM and bandwidth structure is modelled on traditional architectures contra the new PS2 architecture.
A traditional rendering architecture has one memory bank known as videoRAM. Tasks of a traditional videoRAM
architecture:
1) To send and receive textures to and from the graphics chip (rendering).
2) To send and receive Z-coordinates to and from the graphics chip.
3) To store permanent textures.
4) To store the framebuffer (the picture we see on screen).
5) To store the Z-buffer (information the graphics chip needs to render).
The down side to the videoRAM is that it has a relatively slow bandwidth. It can??t ship and receive information very
fast. How ever if we could turn of all the render effects and Z-buffer and simply let it concentrate on streaming the
textures to the graphics chip it would save trumendeous amounts of bandwidth. For instance. The PSX has quite high
polygon performance but very few rendering effects and no Z-buffer. It only needs 130Megabytes of bandwidth.
Nintendo 64 has less polygon performance but a lot of rendering effects and Z-buffer. As a result it has to use
500Megabytes of bandwidth.
So what if you could create an architecture that actually could turn on all render effects plus Z-buffer (aka Nin64) but
never the less use as little bandwidth as a system that doesn??t have this (aka PSX). The saying goes you can??t eat the
cake and keep it at the same time. Normally this might be true. But not for PS2 because this is exactly what it does.
So how does it do it? I??ll come to that but first a little vital explanation. The ideal videoRAM is supposed to have
enormous bandwidth as well as huge storage capablilities. The problem is it isn??t cost effective to combine these
basic factors. So the higher storage capabilities the less bandwidth capabilities and vice versa. Never the less this
technology is what all computers and videogames are using because it is the most simple and cost effective one
around. This gives moderate storage and bandwidth capabilities. In the case of the PS2 this is not enough. This way
the texture as well as render capabilities would be very limited. What Sony did instead was to devide the videoRAM
architecture into two separate parts: One memory block with huge storage capabilities and quite high bandwidth
capabilities (This is the part that all computers and videogames have). But then Sony also placed one small 4
Megabyte block of memory directly onto the graphics chip with enormous bandwidth capabilities. The small block
was made big enough just so it could take over the tasks that demand high bandwidth. This way Sony got both high
storage capabilities and speedy bandwidth.
The huge memory block can now dedicate all its bandwidth to simply send textures to the 4Megabyte buffer. Now,
this means a lot of textures. Then the 4Megabyte buffer takes over all the rendering and Z-buffer check, something
the huge memory block is quite bad at. This means very advanced rendering. As I stated before (07-28) the PS2
textures are stored in the 32Megabytes of mainRAM.
Tasks of the PS2 videoRAM (the huge memory block):
1) To send textures to the 4Megabyte buffer (streaming).
2) To store permanent textures.
Tasks handled by the 4Megabyte buffer (the small memory block):
1) To send and receive textures to and from the graphics chip (rendering).
2) To send and receive Z-coordinates to and from the graphics chip.
4) To store the framebuffer (the picture we see on screen).
5) To store the Z-buffer (information the graphics chip needs to render).
Peter said the 4 Megabyte buffer "performs all the duties of vram". That is almost true, but not completely. Instead it
takes over the exact tasks that it is best suited for. That includes almost everything that needs speed. But as I said
above, you can't make a videoRAM both fast and store a lot at the same time. So it skips the storage... and leaves that
for the much bigger main videoRAM block.This way the PS2 can devide the burdens more accurately to the exact
parts most potent to handle them. In other words the PS2 has a more refined and specialized architecture. It is
somewhat complicated but it makes the PS2 a very powerful texture machine with enormous rendering capabilities.
"People, as much as it hurts to hear it, the fact of the matter is that the PS2 architecture was poorly designed. [08-
01]". The PS2 is as poorly designed as a NASA space shuttle ready for launch. The traditional way of programming,
handling and managing a game doesn??t work on the PS2. Never the less this is exactly what the programmer
community is used to. The whole architecture has been rearanged to combine extreme computer power with very high
efficiensy. It has made it very alien to most programmers. If a programmer never the less choses to use the PS2 the
usual way it becomes a very unbalanced architecture. To program like for a PC kills most of its advantages. You
would then have 32 megs of main RAM combined with about 4 megs of videoRAM. It would be just as efficient as
using a hammer for a screw driver. It doesn??t translate into a good game.
Daniel
More technical talk
The PS2's architecture is not poorly designed. You should believe everything you read in Microsoft press releases. Is
it difficult to program for? Yes, because it's different from what was up until now "the norm". But why is it different?
That's the important question, why did Sony go with bandwidth over size? Because the PS2 was engineered to do one
thing, and one thing only, media processing. Most every other computer system out there, including the Dreamcast,
the Xbox and the Dolphin/StarCube, are built around computing principals developed in the early days of personal
computing when all that mattered was how fast you could run a word processor or spreadsheet. That's what we call
baggage. Such systems were rapidly approaching a ceiling of performance in 3D/media applications because that's not
what they were, really, designed to do. Media processing is fundamentally different from such static applications. In a
word processor you perform lots of different operations on a relatively small piece of data. Conversely, in media
processing (3D rendering, video/sound manipulation, compression/decompression), you're basically doing the
opposite, performing the same few operations over and over again on a huge set of data. And that is what the
engineers at Sony had in mind when they came up with the Emotion Engine and Graphic Synthesizer. The PS2 is
uniquely adept at media processing in ways that no traditional architecture is. Sure, there are some pretty fancy 3D
accelerators and sound chips out there in the PC world and in the Dolphin/Xbox/Dreamcast, whatever, but they, and
things like SIMD instructions tacked onto Pentiums and Athlons and PowerPCs, are simply symptomatic of, and
stop-gag solutions to the core problem, that the traditional PC architecture, used in basically ever system out there,
save for the PS2, was never intended to do such media processing feats.
To suggest that Sony just blew it on the PS2's design is not only naive, but preposterous. They didn't accidentally end
up with a machine hard to program for. And is it really hard to program for, or just different? We know Sony can
make an easy to develop for system, the original Playstation was just such a platform. They chose to go with a non-
traditional style of system to avoid running strait into the world of diminishing returns.
But, I digress. Peter posed a question. Would programmers prefer a large amount of slow memory to store textures
in, or a small amount of incredibly fast memory? Right now, if you ask them, most will probably indicate the former,
but only because that's what they're used to, not because it is inherently superior. Truth be told, I can think of many
situations where high band, small space wins out over low band, big space. Again, in the example Pete gives us. You
can load all 64MB of textures at one time, but that takes time. I thought load times were a no-no... And what happens
when you move to a new area, the game engine screeches to a halt while new textures get slowly uploaded to the
VRAM. That isn't a problem in a properly implemented texture streaming scheme on a system like the PS2. And it
never will be.
So which is really the better way of doing things?
Brad Grenz
YOU GUYS STILL WORRIED ABOUT THE LACK OF VRAM ON THE PS2!
I've read plenty of letters from people whining about the fact that the PS2 has 'only' 4 MB of
texture RAM as opposed to DC's 8 MB, while ignoring the fact that the PS2 is capable of
rendering 75million polys/sec, while DC can only do 3 million - obviously fanboys who've
realised that the anti-aliasing thing is a non-issue, because the ps2 is capable of handling all 3
methods of anti-aliasing, the only reason why developers have not put this feature in there games
yet are because they were completly rushed so that they can get if done in time for the japanese
ps2 launch, but dont be surprised if they implement AA in there games for the North American
ps2 launch. However, it doesn't seem as if any of those making these complaints have bothered
to read the official PS2 specs, which can be found on Sony's North American Playstation site; if
they had, they'd have seen that the hardware MPEG decoder - used for decompressing DVD
images - can also be used for compressing/decompressing static bitmaps - Sony state themselves
that this can be used for textures. Presumably, it will be a similar process as that used for JPEG
images - though via hardware, not software - which means that the compression/image quality
ratio, if applied correctly, could effectively give a texture storage capacity twice as great as that
of DC, and would look better then textures compresses in dolphin with S3, and with minimal
compression artefacts present! I expect we'll see the greatest compression on 'natural' textures
(such as rock faces, sand, carpet etc), where loss of quality will be none. Remember, also, that
PS2 will be able to stream textures from the CD/DVD in realtime, without a processor hit.
Sounds good to me, that's all I can say!
SCEI recently unveiled a PlayStation2 demonstration described as the 'Spirit of the Orient.' The
demo included several character models wearing various styles of clothing, and illustrated the
realistic attributes that can be attributed to garments on the PlayStation2. According to company
representatives, future PS2 titles will feature clothing that can be creased and even torn in-game.
Furthermore, the PS2 Emotion Engine will calculate natural factors such as wind and their
effects on clothing in real-time. Thereby clothes will react dynamically with their surroundings.
Insiders suggest the aforementioned effects will be incorporated into the 'second-generation' of
PlayStation2 titles later this year.
I was reading an article saying that the X-box's main processor was 2x more powerful than the
PS2's. This is VERY wrong. The PS2's processor is over twice as powerful than the x-box's
because the ps2's is equal to about 2 733 pentium3 processors and the x-box's is comparible to
about a p3 650mhz. Besides, the 800mhz p3's can barely do a gigaflop and the ps2's emotion
engine can do 6.2gflops! And the XBOX uses a NON-custom made chip by intel which is a P3
600Mhz, lets not even talk about dolphin while it doesnt even compare to XBOX and the ps2 is
much better in HARDWARE QUALITY than the XBOX!!!
1) The PS2 300 MHz Emotion Engine is much more powerful than 2 PIII 600 in float point
calculation (about 4x faster) because there is 2 Vector Units in the EE and each of them have 10
float point units within. So I really wonder how a PIII 600(xbox) can catch up with the Nvidia
video processor, which can push 300,000,000 triangles per second. Of course, Intel can create a
special version of PIII for the X-box but that will never happen.
2) XBOX has 64MB of unified ram, which means that's all the XBOX has to offer in terms of
ram, so it will be used for sound, textures (using unified ram for textures is not as good as using
seperate VRAM for textures even if you do have 4MB), calculations, OS (YES AN
OPERATING SYSTEM!! HOW STUPID IS THAT, AND MICROSOFT SAYS THEY ARE
DEVOTED TO MAKING XBOX A VIDEO GAME ONLY CONSOLE) etc.Share the video
RAM with the Main RAM bus is not a good idea. Actually, it is not new to the PC. Check some
motherboard's with video card's built-in, they use the main RAM as video ram and it will
actually slow down the performance. It is because the CPU and the Video cannot access the
memory bus at the same time, and that can cut the 400 MHz Direct Rambus RAM bus
performance to half when both CPU and Video chip need to get data at the same time. In the
other hand, although PS 2 is using 200 MHz version of Direct Rambus RAM, which is half the
speed of XBOX, the real world performance will not be 1/2 because there is 4 megs of video
Ram built-in in the Video chip (Graphics Synthesizer) with a Ultra wide internal bus (2560
bit/s). That means there will be no performance hit at all and both the EE and GS can run in full
speed without spending the time for waiting for the RAM bus to get ready, where is XBOX will
have to wait for its ram to free-up, but if developers program their games to wait for the ram
then you cant get as detaile textures, sound, etc. but if they program the game to not wait for the
ram you will see a huge performance and processor hit and the game will run at a much lower
frame rate then programmed for. So, the RAM performance should be really really close
between X-box and PS2 in overall, PS2 mite have the upper hand. So the 64MB's of unified ram
is not a problem at all to the PS2.
3) X-box will use a special version of window 2000 as the OS for X-box. That will be a very big
mistake. X-box gave a huge 64 MB's of RAM, but this is base on the standard of a game console,
not for the unit to run windows. It can easily eat up 16 Megs for memory just for the core
system.
So, will X-box doom PS2? NO!!!
OTHER WEAKNESSES IN "XBOX"
IGNPS2 INTERVIEW WITH NAUGHTY DOG - THIS IS JUST SOME OF THE QUESTIONS
-CHECK OUT THE WHOLE INTERVIEW HERE
Question: Do you think there will be any major advantages to the Dynamic hardware setup of the
PS2 as opposed to traditional PC-esque hardware, it seems to making mainly problems for
developers?
Jason Rubin: Yes, I do think that there are advantages to the hardware setup of the PS2. The
multiprocessor setup of the system allows the PS2 to do so many things at the same time (if
programmed well) that the hardware is actually far more powerful than its spec. They say the X-
Box is going to be either 600mhz or 1000mhz. The PS2 is only 300mhz. But if you can have
your background engine running in 10% CPU and 50% VU1 time, then you have far more
power.
That is how we are doing our programming for the next game. I think that the problem with the
current games (the ones you have seen) is that they have opted to take the easy way out and they
are only doing CPU code.
So think of it like this. They are actually leaving a significant part of the power of the system on
the table. We plan to bring that power into the game.
And yes, of course there will be more powerful systems in the future. That is how technology
works. But the X-Box may not be that system.
Question: In what ways do you think the PlayStation 2 will change gaming as we know it today?
Do you think that it'll be revolutionary?
Jason Rubin: The PS2 will first be evolutionary... and then later, revolutionary. First, you will
see games that are similar to the past, but a whole lot better looking and smoother. Later, you
will see games that do things that have never been done before.
Broadband will have a lot to do with the Revolutionary part of the cycle. We will go beyond
deathmatch, and beyond the glorified chat room of EverQuest and get true multiplayer online
gaming with interaction. That is something we haven't seen yet.
The Xbox is actually slower than the PS2, if you think about it.
They conveniently forgot to mention the vector units on the PS2 when they compared the stats.
You know, that's like comparing cars to each other, but not mentioning one of them has a
Corvette engine and the other an old piece of junk engine.
Anyone who looks at Xbox 700mhz and PS2 300 mhz and immediately thinks "Oh, xbox is
faster!" is not thinking clearly. Xbox doesn't have vector units like PS2, and it's slow, pokey 6.4
GB/sec unified RAM used as VRAM can't compete with PS2's lightning fast 48 GB/sec RDRAM
(yes, PS2's VRAM is 7.5 times as fast as the Xbox's). Okay, yes, it has 4 MB, but I read in an
interview with Hideo Kojima that said the RDRAM is so fast, it really doesn't matter.
Know how many gigaflops PS2 can do? 6.2 gflops. A 700 mhz Pentium III can't get close to that
(what was it, like 3 gflops on a lucky day?). Don't forget the Xbox uses a **PC** processor, not
a console's! Just look at the N64 (100 mhz) and PSX (33 mhz) and compare them to computers
of the same speed. If you take the same mhz::speed ratio between PC CPU's and console CPU's,
the PS2 mops the floor with the Xbox. As you can clearly see, when it comes to consoles, MHZ
is NOT everything. There are plenty of other factors in there. Unfortunately Microsoft thought
mhz was everything, and made a terrible mistake.
Sure the Xbox has all those pretty specs, but are they too good to be true? Of course, remember
that it uses unified memory, so that memory has to be made to work with everything: sound,
system, video. It is un-specialized for specific purposes. On the Xbox, there is no DEDICATED
video RAM sitting right next to the graphics processor. There is no DEDICATED system RAM.
No DEDICATED sound RAM. It's all kind of in the middle, and it's too slow for the video.
That's the same problem the GeForce 2 has: Fast graphics chip but the slow memory bottlenecks
the whole thing and you don't nearly reach those specs nVidia was hyping up. Don't try to argue
that Xbox will be leaps and bounds faster than any normal 700mhz PC just because Windows
isn't there. Windows better be pretty freaking unbelievably sluggish if it's going to do that. If it
really is, why don't developers still make DOS games these days? Windows is a little sluggish,
but removing it is not going to make a PC 700 mhz CPU suddenly behave just like a console's.
Xbox is just a low price gaming PC, and thus it will have PC bottlenecks. The PS2 is custom
designed for the specific purpose of games.
PS2 will be a year ahead of Xbox, giving it plenty of time to build up a large library of games
and user base. Don't forget it can run hundreds of legacy PSX games in enhanced mode as well,
and functions as a DVD player. It's incredibly useful right out of the box. When the Xbox comes
out, it will only have a few launch titles, and that's it. DVD? Okay, I've heard that a lot of people
are buying the PS2 just for the DVD player, so why would they want to get an Xbox for
something they already have? There's no doubt the PS2 will reach $200 or $250 by the time
Xbox is out, so if they wanted a 2nd DVD player, getting another PS2 would be more logical.
After all, it'll have way more games and peripherals. Why would you want to spend $300 on a
new console with few games after you bought a highly useful one for $300 just one year ago? Or,
if you're choosing between two, why buy one that's ::ahem:: LESS powerful, has a tiny library of
games, and is more expensive? Exclusives, what exclusives? Ported PC games? Sheesh.
And let's not forget the target markets. PS2 is going for the general market, a wide range. Xbox
is going for the "hardcore" audience (who are the minority and very likely would have a PS2
anyways). Nintendo will pick up the kids market easily with the Gamecube. It's not like Sony
won't have their share in kids, though. Even a lot of young kids these days have Playstations and
know and respect the name.
Brand names are another thing. Ask any person on the street if they've heard about "Playstation"
or "Nintendo". Very likely they would have. Then try Xbox and watch the confused look on their
faces. MS is launching the Xbox right when Playstation will be more popular than ever, and
Nintendo's bombshell Gamecube will steal fire from Xbox, since they're coming out very close
to each other. Really, when you think of a game console, do you think of Microsoft first? Well,
in MS's case you might think of the DOJ trials and a Windows computer crashing.
$500 million marketing blitz is not going to work. True, after being bombarded with
commercials for Xbox, it will generate some interest. But that interest will be quickly wasted
away when the gamers (or parents) go to the store to check it out and sees a pitiful rack of Xbox
games compared to the huge PSX/PS2 library. And just because it's Microsoft doesn't mean
they're going to win. Remember the battle between PalmPilot and WinCE? Who won that battle?
How about Quicktime vs. AVI, who won there? Realmedia vs Windows Media? ActiveX vs Java
and Macromedia? AIM and ICQ vs MSN Messenger? Oh yeah, where are the EXCLUSIVE
GAMES, Microsoft? Where are your Zeldas, Crash Bandicoots, and Marios? Halo certainly
won't be, in fact it's coming out on Mac and PC. It might be old news by the time Xbox comes
out, unless MS ::cough:: "recommends" ::cough:: to Bungie that they delay it for computer
platforms until then. MS almost had a reason for computer gamers to buy an Xbox when they
bought Bungie and put Halo's availability into question, but they lost it when legions of angry
Bungie fans poured thousands of signatures into a petition saying "I'm NOT going to buy an
Xbox!" Speaking of games, where the heck are they? Have you seen the list on xbox.ign?
Pathetic. Even the developer list is meager.
In Japan, it's as good as dead. I was talking to a Japanese friend of mine over ICQ recently, he
said that nobody in Japan really cares about Xbox or Microsoft, they're all about Playstation.
I've seen a lot of Nintendo supporters bashing at PS2 recently. However, I think PS2 and
GameCube will go nicely together, just like PSX and N64 did. Stop fighting, and concentrate on
the true enemy: MicroBorg. They've assimilated your computers, and now they want to
assimilate your living room.
One last thing: How many of you, and how many people you know, held off on buying a
Dreamcast because the PS2 was on the horizon? Quite a bit I imagine. Now how many are doing
the same to the PS2 because of the Xbox? "Uh, X-what?" There ya go. Enjoy.
Facebook
Twitter