Official Xbox2 Thread

[Sonic]

Same as PS3 thread applies.

1. What will it take for Microsoft to launch the Xbox2 as smoothly as possible? What are the best marketing moves and how should the system be marketed in general? Should it be seen only as a gaming console or should it also be seen as a center for home entertainment?

2. We haven't really heard anything much other than some great forms of speculation on a very powerful GPU and a decent CPU. What other areas of improvement do you think we will see in the next generation? In terms of sound, AI, physics, and of course online functionality.

3. Anything else anybody want to add?

 

[V3]

If MS want to waste money, waste it on free games or subscriptions instead. Consumers can see value in free games, but money waste on hardware, they don't always see.

Else, I think MS is doing alright as it is. As for Japanese market, well, forget that and concentrate else where.

 

[BoddoZerg]

If Microsoft is going to market the Xbox2 as a PC/console hybrid, they need to give it a lot more functionality. Despite being marketed as a major advantage for the Xbox, I've yet to see the hard drive truly be exploited to its potential. MS needs to allow some kind of modding community for Xbox Live, if not actual mods at least give us the ability to create and download maps and skins...

On the other hand, no matter how flexible Xbox2 becomes, I doubt that most of the hardcore PC gaming community would ever consider using a console to replace their PC. For that reason, I think the Xbox2 should not dream of being a PC, but would do better competing console-to-console. The Xbox has a rather different style of games compared to the other consoles... it's almost like having a console that plays PC games. This has served to its advantage, as a lot of previously pure PC gamers (like me) have moved partially to the Xbox.

Right now, the biggest advantage of Xbox over other consoles is the ability to play multiplayer games over System Link. Xbox LAN parties are great, and that's a functionality you cannot get with PS2/GC. Microsoft would do well to extend this functionality, perhaps with the option to make one Xbox a dedicated server so that you can have giant games with no framerate problems.

On the other hand, while having "PC-style" games is a big advantage of the Xbox, it is sometimes a disadvantage. There has been a relative lack of exclusive games on Xbox; a lot of the best games on Xbox are PC games or even cross-platform. Splinter Cell and Raven Shield are cross platform, and probably best played on PC. Unreal Championship was a mediocre port of UT2k3; the PC game is much better than the Xbox version. The Xbox needs to sell itself with exclusive games like Halo, MechAssault, and Brute Force, instead of relying on cross-platform stuff like Splinter Cell and Unreal Championship.

 

[zurich]

The 'downloadable content' is only as good as the infrastructure to support it. Look at the pathetic state of the MechAssault & Unreal Championship packs/patches/whatever for reference. I've been waiting for 5 months now for the promised maps/gameplay expansion... don't deadlines mean anything when it comes to freebies?

IMO, if Xbox Live is going to play a central part of Xbox2, then the sticker on the DVD case touting freebies and such should actually mean something.

*goes to trade in MA & UC tomorrow for Pro Race Driver*

 

[DeathKnight]

There's already been a handful of MechAssault downloads.

Downloads will only come as fast as the devs can put them out (and they may have to follow a set release schedule). If most of the team is also working on a new game then new content for previous games will be spotty at best.

Pro Race Driver? I could see possibly trading in UC toward that, but keep Mech Assault (unless the game isn't your cup of tea anymore).

 

[zurich]

There's already been a handful of MechAssault downloads.

Downloads will only come as fast as the devs can put them out (and they may have to follow a set release schedule). If most of the team is also working on a new game then new content for previous games will be spotty at best.

Pro Race Driver? I could see possibly trading in UC toward that, but keep Mech Assault (unless the game isn't your cup of tea anymore).

MechAssault DLC has been delayed for about 2 months now.. the came as shipped is half complete feature wise. The 2 maps released are basically worthless w/o new gameplay modes (CTF). The game was great, but only kept my attention for 2 weeks at most..

As for UC.. well, everyone online seems to suck Not much competition really (i'm a seasoned online FPS vet from the QW/NetQuake days!). I'd love to get Ghost Recon, but it seems the server search is still broken.

Pro Race Driver looks very unique and intriguing.. and I'm always one to reward narrative innovation (no matter how trite )

 

[BoddoZerg]

They need to allow modding, or at least mapping, to make DLC a viable model.

No console game company can churn out enough maps to match what is user-made for virtually any PC game. Right now, Downloadable Content consists of one or two maps every 2 months. That's not a viable plan if MS wants DLC to be a major attraction for Xbox.

Also, if DLC is to become the major feature that it should be, console game companies need to start adopting a more PC-game-company like attitude toward patches. There's no excuse for the steaming pile of crap that is the multiplayer user interface for MechAssault and Ghost Recon. Stuff like that is clearly half-done and needs to be patched. As long as Downloadable Content is printed in big blazing letters as a major feature of the game, people are not going to forgive games with glaring flaws - they will expect a patch to fix it.

 

[zurich]

Just out of absolute pure curiousity, what would IA-64/Flagpole/Itanium (whatever you want to call it) bring to the Xbox2? Now you don't need to tell me its horribly expensive/not designed for a console, but HYPOTHETICALLY speaking, would any of the devs here see a benefit/use for having a monster CPU like that in a console?

Would an Itanium in XB2 allow for graphics/AI/physics leaps and bounds beyond what we could get with X86, or would it largely go unused? I know its a pure server chip, but it does theoretically have some insane FP power.

Any programmers/devs care to hypothosize?

(this came to me while thinking what MS/Intel could cook up to counter CELL's TFLOP power)

 

[Deepak]

whatever CPU they put in XB2....it wont be like that they pick up a CPU and put it in...there will be some customisations...CPU will be tweaked up (or down)...

 

[Saem]

The Itanium would bring horrible x86 performance so you can throw out backwards compatibility. Otherwise, it'll bring some serious bandwidth, fp power, big caches, quirky integer performance and insane power consumption and heat dissipation. WEEE!

 

[Tagrineth]

The Itanium would bring horrible x86 performance so you can throw out backwards compatibility. Otherwise, it'll bring some serious bandwidth, fp power, big caches, quirky integer performance and insane power consumption and heat dissipation. WEEE!

Itanium2 has hardware x86.

But it would be a beautiful thing if MS use an I2 in their next console. Those things are seriously frickin' fast at native EPIC code.

 

[jvd]

The Itanium would bring horrible x86 performance so you can throw out backwards compatibility. Otherwise, it'll bring some serious bandwidth, fp power, big caches, quirky integer performance and insane power consumption and heat dissipation. WEEE!

Itanium2 has hardware x86.

But it would be a beautiful thing if MS use an I2 in their next console. Those things are seriously frickin' fast at native EPIC code.

I'm still saying that the hammer class chip would be the cheapest and most powerfull chip for ms to go with and in turn use more money on the gpu. I still say ms should have gone duron or even athlon instead of the celeron they ended up with.

 

[zurich]

The Itanium would bring horrible x86 performance so you can throw out backwards compatibility. Otherwise, it'll bring some serious bandwidth, fp power, big caches, quirky integer performance and insane power consumption and heat dissipation. WEEE!

Itanium2 has hardware x86.

But it would be a beautiful thing if MS use an I2 in their next console. Those things are seriously frickin' fast at native EPIC code.

I'm still saying that the hammer class chip would be the cheapest and most powerfull chip for ms to go with and in turn use more money on the gpu. I still say ms should have gone duron or even athlon instead of the celeron they ended up with.

Somehow I think Itanium > Hammer (if cost is not an issue)

 

[Tagrineth]

Itanium2 would brutally slaughter the Hammer.

In both performance AND price.

 

[Saem]

Itanium2 has hardware x86.

I guess you're not well versed with the Itanium 2's x86 performance. IT SUCKS HORRIBLY! Even with all the improvements that Itanium 2 brought to the table to x86 performance it didn't get it anywhere near current CPUs. As for native IA-64 code, it's fast yes, but a lot of that is coming from it's caches and expensive busses. Something which consoles wouldn't be able to stomach. It's a bit of an FP monster, see Spec fp.

Durons would be a bad idea. Everybody forgets that their heat to performance ratio blows compared to Intel. Mobile Duron parts have their max heat dissipations in AMD data sheets are crap, if you read the fine print you realise they're at best light work loads when put into the context of games. AMD doesn't have fab capacity. It's one natural disaster away from having it's production line wasted. Porting a processor to another fab isn't exactly worthwhile unless the returns are huge, which they probably wouldn't be. AMD didn't have a comparable processor anywhere near the time contract bidding happened, they had stuff in the pipeline, but it wasn't out yet -- you can't count on that.

Right now I'm thinking an modified banais would tear it up. It's FPU is a fair bit more powerful than a PIII with wicked power characteristics. If Intel did a bit of custom work a modified Banais would be pretty spiffy.

Hammer would be okay, the problem would be the weak connection between the CPU and GPU there is also still the big lack of fab capacity and lack of a reasonably priced low end Hammer. Search high and low and there is no big volume Hammer shipments.

 

[DeathKnight]

I still say ms should have gone duron or even athlon instead of the celeron they ended up with.

It's actually a P3...

They didn't go AMD for probably a couple of reasons, one of them being heat issues.

 

[AzBat]

On the other hand, no matter how flexible Xbox2 becomes, I doubt that most of the hardcore PC gaming community would ever consider using a console to replace their PC. For that reason, I think the Xbox2 should not dream of being a PC, but would do better competing console-to-console. The Xbox has a rather different style of games compared to the other consoles... it's almost like having a console that plays PC games. This has served to its advantage, as a lot of previously pure PC gamers (like me) have moved partially to the Xbox.

I too moved partially to the Xbox. I had previously been only a PC gamer, but due to new circumstances in my life I'm now mainly a Xbox gamer. Had I kept my job in the industry I'm not so sure that would have happen. Does that mean I'm not happy being a Xbox gamer who can't keep up with the PC industry? No, in fact I'm happier now since I actually get to be a gamer that plays instead of a worker that tested.

Anyway, my thoughts on Xbox Next is that they need to play up the strengths of what they have now. I'll try to list all that I consider to be strengths. You'll definitely see some of the ideas Boddo and others have mentioned:

1. PC roots. I believe this helped a lot of people think differently about game consoles. They need to continue embracing PC technologies and make it their own. Hard drive and Ethernet were the best ideas. They could embellish these and add even more PC technologies. Keyboard and mouse? No way. It's a game console not a PC. You can add PC technologies without making it a glorified PC. Think wireless and PC or Xbox connectivity(maybe similar to the media distribution idea that TiVo has just announced). That brings up another thing...

2. It's just a game console. I think Microsoft got this right. However, there are a few people that want the Xbox to do other stuff like TiVo, Internet access, etc. I think it's best to keep those ideas out of the Xbox. Sure it's more than capable of doing such things. Microsoft could make a higher-end/higher-cost Xbox that has those capabilities, but it shouldn't be included by default. Maybe they could provide expansion capabilities that would allow you to turn it into a TiVo or Internet appliance. The current Xbox didn't provide any expansion possibilities and I think some Sega/Nintendo/Sony console players expect that even if they're never used.

3. Live and multi-player gaming. I think Microsoft completely changed console gaming with this. They need to continue with it. Microsoft can't afford to let Sony outdo them on this front. Boddo's ideas on user content and dedicated servers are great and something that could even be done right now. Xbox Live should be the same service for Xbox Next and the 1st Xbox. This way you can keep your GamerTag. Also, if they provide backward compatibility, then old and new Xbox users could play against or with each other. I could see there being special Live services that's only available for Xbox Next users though. The Xbox Live service would come installed automatically, but you'd need to pay to enable it. What would be nice is if you weren't ready or able to subscribe to the Live service, then you could at least pay like $5 to reserve a GamerTag. That could give you a GamerTag@xbox.com email address and personal web space for placing game saves, custom maps, rosters etc. for trading online. As for improving multiplayer games, they could add the ability to use a second TV instead of using a split screen.

4. Exclusive games and exclusive features. Halo and Splinter Cell are probably the highest rated and highest selling Xbox games. They're also my favorite 2 games too. Microsoft was smart to keep these exclusive to the Xbox as long as they could. Microsoft needs to continue with this, but they need more of them. That means spending more money on the games. Could Microsoft just buy more companies? Possibly, but I'm starting to think it might be better to just put that money to use by buying exclusive rights. As for the exclusive features, they used technology that couldn't be done properly on other consoles. Like pixel and vertex shaders. Maybe Microsoft should be giving more money to those that take advantage of those features even when they are cross platform titles. Or go a step further and not allow cross platform titles on the Xbox unless they support some exclusive features.

5. Faster and more powerful. If it had not been for the fact the Xbox was faster or powerful than both the GameCube and PS2, I don't think the Xbox would be as successful as it is today. It definitely helped some people give it a chance considering Microsoft had never made a console before. This time around more people are going to expect the Xbox Next to do the same to the PS3. If it doesn't, then it could hurt sales. Even if it was slightly faster than the PS3 that would help. Hopefully they'll be able to take advantage of smaller and cooler processes to do this.

6. Little things. Xbox had some cool little things that stood it apart from the rest. Like break-away cables, custom music soundtracks and parental controls. Since Microsoft is coming from a totally different perspective, then they need to take advantage of this and add more of these cool little things. Below are a few ideas.


Other things that need added and improved upon for Xbox Next...

1. Backward compatibility. Say what you will about the PS2, but this was absolutely the killer feature for the PS2. Working with JPA I was able to send PS2 design suggestions before they released it and this was one my major ideas, though I think they already decided on this. And not just a perfect emulation of the old PS games. I also wanted them to make the old games look better. I hated seeing the point sampling/non-perspective correct textures. I wanted them to fix that with the PS2. And I think that's what Microsoft should do as well. Not only do they need to make Xbox Next backward compatible, but the games should be enhanced with faster frame rates or better image quality via FSAA if possible. If you make it backward compatible then the controller connectors need to be the same. Nice thing about using USB technology is that they could change the controllers without needing to change the ports. However, I don't see any reason why they would need to add or remove buttons or sticks. Anyway, do not underestimate the importance of backward compatibility.

2. DVD support needs to be standard. It needs to include support for progressive scan and the remote needs to support powering the unit and opening/closing the drive. Sony just announced an updated version of the PS2 and its remote does all of this. Except this time around include the DVD license in the price of unit. They wouldn't have to include the remote, but they could at least make the remote sensor built-in. That way it doesn't take up a controller port.

3. Take the content ratings for parental control a step further. Make it so that if you select Teen rating on the parental controls that the game will dial down the content to match the selected rating. This way the game can still be played, but without the objectionable material the parent doesn't want them to see. Another thing that would be nice is if you could leave the parental setting set for Teen and still be able to load a demo disk like those from Official Xbox Magazine. Though you wouldn't be able to run any demos that were rated above a Teen rating unless they supported changing the content to match the rating. As it is now you have to have the setting set for Rating Pending in order for the disc to load.

4. Bring back the Dreamcast's VMU. This was a very cool idea. I liked using this with football play selection. However, it should not replace the standard memory unit. Also, why not make it capable of playing a couple of songs from your soundtrack library? You could use this on a quick jog around the block.

5. HDTV compatibility. 1080i is absolutely cool. More games need to support it or at least 720p.

6. The music player needs to be as fully functional as Media Player 9. Like skin support or adding all kinds of visualizations. Xbox Live could even add a MusicMatch type service to automatically name all your tracks and albums.


What about the timing, marketing, pricing etc? It needs to be out in the USA at least 2-3 weeks earlier than the PS3. Other than that, I don't think there's much more they can do better than they already did with the original Xbox launch. As for launch titles, they need to have 5 or more 1st party and 15 more 3rd party titles. I think having at least a third of them be Xbox Live titles would greatly help add to the sales. Pricing at $300 with no pack-ins is pretty much a given.

As you can see, Microsoft already has a lot right and a lot that can be enhanced and built upon. I think Microsoft has learned a lot with the 1st Xbox and I expect them to surprise Sony and Nintendo when they release the Xbox Next.

Tommy McClain

 

[jvd]

The Itanium would bring horrible x86 performance so you can throw out backwards compatibility. Otherwise, it'll bring some serious bandwidth, fp power, big caches, quirky integer performance and insane power consumption and heat dissipation. WEEE!

Itanium2 has hardware x86.

But it would be a beautiful thing if MS use an I2 in their next console. Those things are seriously frickin' fast at native EPIC code.

I'm still saying that the hammer class chip would be the cheapest and most powerfull chip for ms to go with and in turn use more money on the gpu. I still say ms should have gone duron or even athlon instead of the celeron they ended up with.

Somehow I think Itanium > Hammer (if cost is not an issue)

i agree , but cost has to be a factor at some point and a 2000+ $ cpu vs a 100-200$ cpu is a big diffrence .

 

[archie4oz]

They didn't go AMD for probably a couple of reasons, one of them being heat issues.

Never mind that AMD wanted more money up front...

I guess you're not well versed with the Itanium 2's x86 performance. IT SUCKS HORRIBLY! Even with all the improvements that Itanium 2 brought to the table to x86 performance it didn't get it anywhere near current CPUs. As for native IA-64 code, it's fast yes, but a lot of that is coming from it's caches and expensive busses.

I gotta agree strongly with this... Hell I'm sitting in front of a dual Itanium2 right now doing some dev on it and I gotta agree, that it's x86 performance is abysmal (like the P4's x87 capability, it's basically there for compatability). Running scripts or interpreted data it's slower than snot even with native code. Even with Intel compilers, assuming you've ironed out all the various errors and funky results, it can be challenging to wring performance out of it (especially spaghetti code). And doing any assembly level optimizing makes PS2dev look like a walk in the park... Of course when you do get your code dialed in right, it does SCREAM, floating point code in particular...

Also it IS a rather expensive chip to impliment on a console... It'll be interesting to see how much low they will sell Deerfield as it's supposed to be a low-cost implimentation of the arch. Besides you'd still be neglecting the whole benefit of going Intel by using a different arch that so few devs have so little experience with and so few binary tools, APIs, libraries, etc... compared to x86...

I agree that Banias or a later derivative would a much better solution...

 

[Saem]

Itanium needs a multi-threading front-end which can repack instruction words and some sort of VM front end which can make it seem more normal. Oh wait, Intel != Transmeta.

Seriously, the Itanium2 definately has some really scarey execution units under the hood and it can easily have that firing on all cylinders (It is a two banger after all, . It's a problem getting the code and CPUs to love each other.

In anycase, I wonder if there is any sort of abstraction that can be done which would pull out more from the code and make the code friendlier to the MPUs that execute it. Right now some sort of multi-threading seems to be the only hope. It really does seem that there is a "gap" which needs to be bridged.

As for XBox 2, I expect a very aggressive CPU, NUMA and a GPU which is probably going to be significantly more powerful than R300 -- I do believe ATI will be providing the part.