J Allard talks more on Xbox 2 hardware (still vauge)

some of this I know has been online and we've read it, but maybe not the entire thing.

http://news.teamxbox.com/xbox/5962/Xbox-2-Development-Secrets-

Xbox 2 Development Secrets Revealed
By: César A. Berardini - "Cesar"
May. 19th, 2004 05:42 pm


J Allard had the following words before introducing the latest XNA advancements at the Microsoft Pre-E3 press briefing held in Los Angeles last week: "Now, every time I talk, people ask me why I don't say a little bit more about hardware. It's always software, software, software, software. So tonight you guys win. Let's talk about hardware."

And to the disappointment of the crowd expecting Xbox 2 related information, the Saleen S7 featured in the second Crash demo made by Pseudo Interactive was unveiled. It is not that the XNA demos aren’t impressive, but it isn’t exactly the Xbox 2 hardware specs everyone was anticipating. In this first interview, J Allard confirmed to Computer and Videogames (CVG) that although the Xbox 2 hardware is “locked downâ€, Microsoft has decided not to talk about next generation hardware this year.

Luckily, CVG caught J Allard the morning after the Microsoft briefing in an effusive moment, in which the Corporate Vice President and Chief XNA Architect revealed some information about the Xbox 2 development.


"There's some stuff that's just knocked my socks off," stated Allard. "The thing we're looking at in the next generation is just an unbelievable amount of raw computing power - the architecture will be much more specialized."


Showing his excitement, Allard continued: "Right now you have your audio chip and you graphics chip and your CPU, and you're constantly trying to figure out the balances. In the next generation we're going to have so much silicon, so much raw computing horsepower - developers are going to be able to use this in interesting and exciting ways."

By now it was too late for Allard to stop talking and he continued to reveal additional details to CVG: "I've seen demos of terrain and worlds, with no textures in them whatsoever and no geometry - it's just a program that's creating a scene for you," and then Allard had to explain what procedural synthesis is all about:

"Art is the highest cost component of game development, and so much of the art is really repetitive and really intensive, and then doesn't come out to be very realistic. You know, bricks in a wall - very repeated textures."

"Let's go write the brick program and run the brick program to make a room full of bricks, lose the art expense and gain a more realistic looking room, because now we can focus on having the bricks there in a really realistic way. I get really excited about that kind of stuff."

"There are a lot of new techniques," Allard continued. "Like what shaders have done for 3D, there are a lot of new next-generation techniques for procedural synthesis that's really going to change how game construction is done, but also what the environment looks like so it feels a lot less 'cookie cutter'."

Sounds cool, eh? Coincidently, the latest issue of MIT’s magazine, Technology Review, has an article on Microsoft Research Asia’s Beijing lab. A division launched late last year that works on the Xbox and Longhorn (next version of Windows) is the focus of the article with some juicy quotes from experts around the world:


"Microsoft Research is by far the biggest contributor to graphics in the corporate world. It’s a powerhouse," says Paul Debevec, a graphics expert at the University of Southern California’s Institute for Creative Technologies. "The Beijing lab, in particular, has achieved “some amazing results,†he adds. “It’s not just, ‘How can we make a better Xbox?’"


Then the article continues: "But in fact, a better Xbox is ultimately part of the lab’s mission. Reminders that this is a business, not a researcher’s playground, are never far away. In an adjoining hallway, a large corner room has its windows plastered over with opaque sheets of paper. The sign on the locked door reads, “Xbox: Confidential.†Baining Guo, a former Intel researcher and now Microsoft Research Asia’s graphics research manager, isn’t allowed to talk about what’s going on inside. “Some of our best people work in there,†is all he’ll say."

Could this be the place where the latest developments in “procedural synthesis†are being conducted? We’ll have to wait a few months to find out the truth, but the excitement J Allard expresses while talking about next generation technology is easily contagious when you read information such as this.

For the entire Beijing lab article, click here. Stay tuned; we’ll have much more on the next wave of gaming technology as it becomes available.
 
Microsoft's Graphics Research labs are indeed doing some amazing stuff.

A lot of Siggrah papers come out of that group.
 
"Microsoft Research is by far the biggest contributor to graphics in the corporate world. It’s a powerhouse," "The Beijing lab, in particular, has achieved “some amazing results,â€￾
wow, coming from Paul Debevec, that statement is a huge complement
 
Allard continued: "Right now you have your audio chip and you graphics chip and your CPU, and you're constantly trying to figure out the balances. In the next generation we're going to have so much silicon, so much raw computing horsepower - developers are going to be able to use this in interesting and exciting ways."

AKA: "We'll dump specialised audio hardware, give you a shed load of CPU power that has some graphics capabilities and a unified shader model on the graphics and let the developer decide what he wants and where to process it."
 
I dunno, my shed is pretty small... One of those half-height things that sits against the house and barely fits a push-mower. :p
 
DaveBaumann said:
Allard continued: "Right now you have your audio chip and you graphics chip and your CPU, and you're constantly trying to figure out the balances. In the next generation we're going to have so much silicon, so much raw computing horsepower - developers are going to be able to use this in interesting and exciting ways."

AKA: "We'll dump specialised audio hardware, give you a shed load of CPU power that has some graphics capabilities and a unified shader model on the graphics and let the developer decide what he wants and where to process it."

AKA: "Let's see, shall we drop DD5.1, realistic AI and physics or that group of guys in the background that look and move great and have no relevance to the game whatsoever?"


Bu-bye neat-y features. You know that if devs want CPU performance, and devs always want more CPU power, they'll trade anything they can to make their "vision" come true.
 
Bu-bye neat-y features. You know that if devs want CPU performance, and devs always want more CPU power, they'll trade anything they can to make their "vision" come true.
Well, as noted many times before, multiprocessor stuff will be wrestling with existing codebases not aimed for it. Dedicating say, one core just for sound will simplify matters there, and improve hw utilization at the same time.
:p
 
Fafalada said:
Bu-bye neat-y features. You know that if devs want CPU performance, and devs always want more CPU power, they'll trade anything they can to make their "vision" come true.
Well, as noted many times before, multiprocessor stuff will be wrestling with existing codebases not aimed for it. Dedicating say, one core just for sound will simplify matters there, and improve hw utilization at the same time.
:p


But then you have the devs who thinks, "mmmm if i can get just that little bit of power to do this thing, it will only cost a sound channel after all!!"...

I'm not saying it's bad in itself, but lazyness will make this approach very shovel-ware attracting... This is all imho
 
Look at it from perspective of sound engineers - for decades they've been getting the shaft (literally) on game hardware.
Finally they're getting some real power to play with - even if they'll have to fight others for it :p
Anyway more general purpose power it's always double edged sword, but we've been pushing for this in graphics, so why not sound.
 
Fafalada said:
Look at it from perspective of sound engineers - for decades they've been getting the shaft (literally) on game hardware.
Finally they're getting some real power to play with - even if they'll have to fight others for it :p
Anyway more general purpose power it's always double edged sword, but we've been pushing for this in graphics, so why not sound.


Heh, that's right, now sound engineers will have to fight with the AI, physics people and also the Graphics people. They all want as much CPU power as they can get. And in the end you just know what will happen. Graphics first, then maybe AI and physics, then sound after everything else has been finalised.

In the end, they get the leftovers!! :LOL:
 
Fafalada said:
Look at it from perspective of sound engineers - for decades they've been getting the shaft (literally) on game hardware.
Finally they're getting some real power to play with - even if they'll have to fight others for it :p
Anyway more general purpose power it's always double edged sword, but we've been pushing for this in graphics, so why not sound.

Sound technology don't evolve at the same rate as graphics. I also disagree that sound engineers have been shafted for decades. The Genesis had a really good sound chip for that time period and everything since then have been equally balanced with graphics.
 
Whatever happens, anything less than DD and/or DTS in realtime on every single game worth buying will disappoint me. And it must be a good implementation of them. No point in having the "standard" if it's not used properly in game.
 
Its always interesting when technologies like procedural synthesis or nurbs are mentioned but what is the % of games that use this tech? I guess a tiny fraction. I remember Phil Harrison at the PS2 launch saying that procedural content was what PS2 "designed for".

If MS captures this advanced technology in XNA and made it easy for smaller developers to use, it could give XBox 2 an edge over Sony.
 
bleon said:
Its always interesting when technologies like procedural synthesis or nurbs are mentioned but what is the % of games that use this tech? I guess a tiny fraction. I remember Phil Harrison at the PS2 launch saying that procedural content was what PS2 "designed for".

If MS captures this advanced technology in XNA and made it easy for smaller developers to use, it could give XBox 2 an edge over Sony.

It's just that, like anything else, you have to see what's the trade-off. I guess devs will normally just do what's "easier" in that if they can get the same effect (or similar) with 2 different techniques, they will use the one that's more trouble-free to implement.

And so far, the polygon+texture way is just the easiest. OK, maybe not the "easiest" but by far the most "tried and tested". The most familiar to everyone. And saying "if Sony uses this technique, they will destroy the competition" is a bit dull. Destroying the competition relies as much on "graphics techniques" as it does on the dimensions of Ken Kutaragi's ass. (not even "graphics" by itself, let alone single "techniques"...?)

Also, it's not always clear what one means with "procedural", is it textures? Lighting? Geometry? There are quite a few of games with procedurally-generated content. I guess we should classify those contents.

Edited.
 
I can only think of 2 games that use procedural tricks, Silent Hill 3 (procedural textures) and Dropship. I might have played a lots of games where the effects are so subtle that I didn't notice them.
 
bleon said:
I can only think of 2 games that use procedural tricks, Silent Hill 3 (procedural textures) and Dropship. I might have played a lots of games where the effects are so subtle that I didn't notice them.

There's loads more. In some it's neat, in some it's just a marketing bullet-point feature.
 
If a procedural texture or model is good done, then you don't see that it's procedural.

XNA is not even something new. For example ".the .produkt" is done with a tool (developed by the same demo team that did that famous 64k FPS demo), where nearly everything is done proceduraly --> http://www.theproduct.de/tool1.html

T2B.JPG


I guess XNA will be something similar, I just wonder why it's such a big deal if this small demo team can come up with a procedural 3D game enviroment that does all those procedural things already, then every other big game company should be able to do the same if needed. ;)

Fredi
 
WSJ had a column about all the middleware companies exhibiting at E3. Some from farflung places like Italy.

Bigger companies like EA write their own tools and libraries but they might consider an outside middleware for instance to animate faces of football players in Madden.

What was telling was that Criterion has gone from 25 engineers to 160 in 4 years.

So smaller developers are more likely to depend on Criterion, since it's multiplatform, than something like XNA? BTW, some "bigger" developers use Renderware for their games too. Even GTA3 is Renderware, as are several of the Tony Hawk games in this generation.
 
Back
Top