Microsoft leaks details on Xbox Next

Status
Not open for further replies.
Anyone actually more interested in the speculation that it will launch ahead to PS3 in 2005? This could be the real decider in terms of increasin gtheir foothold next gen.
 
predictions:

-2 or 4 POWER / PowerPC core 'XCPU2' (not 3 cores)
-R500X or R550X 'XVPU'
-512 MB memory (DDR2 or GDDR3)
-20 GB HDD
-no back compatability
-1.5 billion verts peak (200-300M sustained with everything)
10 trillion ops/sec (not FLOPS!) like Xbox has 1 trillion ops/sec
 
It certainly could (as I've said, ANYONE could in some form or another), and in fact I think it might be more critical for them to do just that if they're going to be losing the HD and not offering backwards compatibility, as in side-by-side comparisons those points definitely hurt it (HD not that way, but in mentioned comparisons from Xbox previously, especially since it was a major point for them this gen and they wouldn't get to lean on it). Launching at least a few months ahead would get them some momentum that launching in the same area wouldn't provide.

However, since they still appear to be waiting on confirmation of some points until the PS3 announces theirs, it's extending the design process outwards while wanting to launch closer--squeezing their timetables. 3-6 months would be squeezing CPU and GPU design as well--which can always get better with more time--and game developers as well (who are able to proceed now, which is good, but are still waiting on some major details like RAM and what optical drive will be in place to work on streaming and the like). Xbox launched well enough for its small timeframe, but it's certainly one of the things I wanted to see Microsoft definitely NOT copying this round. I'm not sure how much marketing advantage they will get by beating to market (and by how much?), but there is certainly the possibility of it impacting the design process.
 
...

Why not? You don't really think they are putting multiple cpus in the system for nothing but generic housekeeping do you?
Out of 4 logical CPUs, one would run the kernel, one would run the user process(game in this case), and one would run the DX subsystem at any given moment. Developers would be free to utilize the fourth logical processor through multithreading, which may or may not be popular with developers. But MS has no use for Altivec. MS will do well without one.

Losing the hard drive is probably my biggest concern at the moment. Given the frequency of updates for Xbox Live, and the number of games that have benefitted from patches and updates (or that would have) relying on memory cards seems a bit risky. Even a gigabyte would fill up pretty fast, and I personally hate the saving/loading times you get with cards.
I am expecting MS to replace the HD with a 128 MB flash drive directly on the motherboard.(128 MB flash cards seem to be rather cheap nowadays).
 
...

Compatibility with the original Xbox, which is based on Intel and Nvidia chips, isn't guaranteed. Microsoft is concerned it would cost too much money in hardware or in licensing fees to enable the Xbox Next to play old Xbox games. This is risky in part because Sony's strategy has been to maintain compatibility with its old consoles.
We don't know what kind of licensing agreement that MS has with nVIDIA and how much of non-DX nVIDIA technology is in Xbox. Xbox's dependence on nVIDIA specific technology is real and MS does have to pay more get this into Xbox Next.
 
Since MS wants Xbox and Online to be synomonous, I can't understand how they could have an online-ready machine straight though till 2010 without some sort of hard disk... and if the HD is a seperate accessory, that will just splinter their userbase.

IMO, this X2 plan as outlined by Takahashi (no HD, fall 2005, 256 megs of ram, etc) just reeks of Dreamcast. IE: MS announces the specs of the machine, the public goes "woooo ahhhh", then Sony just laughs, mentions that their machine will be 100x as powerful, and the public goes "WOOOO AHHHHH". Insta-$99 Xbox2 by the time PS3 launches :?
 
Re: ...

Deadmeat said:
Developers like Sweeney say they are pleased it will be apparently easy to develop games for Microsoft's new box. That was one of the main advantages that Microsoft has had over its rivals.
"I can't imagine how you will actually program it" - Tim Sweeney on CELL. Tim has spoken...

I'm curious why Tim Sweeney keeps getting quoted about console software development, as if he's been some massive driving force in the console environment...
 
Speculating for a moment, assuming MS will ship Xmas 2005, and Sony has two options, an Xmas 2006 launch or an Xmas 2205 launch with reduced specifications.

What is in Sony's best interests to do?

Do you give MS a 12 month lead, and hope for the Dreamcast like effect, or Do you commit early and compete head to head?

And on a related note did the Dreamcast fail because of Sony's markrting or becasue developers didn't buy in?

Just something to think about. Personally I'm not sure, waiting untill 2006 is one hell of a risk if you can be competitive in 2005.
 
Re: ...

kaching said:
I'm curious why Tim Sweeney keeps getting quoted about console software development, as if he's been some massive driving force in the console environment...

All he ever did in that regard was a terribly hokey unreal (tournament?) port to the PS2 which ran at 30fps. It was as badly programmed as anything and everything else Tim's ever done.

Don't forget Tim was also the guy who said hardware T&L would be impossible with tile-based deferred renderers plus a load of other horsesh!te, and then we got Naomi2 arcade boards.

John Carmack he definitely ain't.
 
Re: ...

Deadmeat said:
Out of 4 logical CPUs, one would run the kernel, one would run the user process(game in this case), and one would run the DX subsystem at any given moment.

Unfortunately for you, all the CPUs - logical or not, as there is no difference visible on the software level - will have to run a copy of the kernel, or else they can't task switch, and thus be generally inefficient at what they do.

But MS has no use for Altivec. MS will do well without one.

You still didn't say WHY. Problem is, we all know you *can't* say why, because the reason simply isn't there!

It's like your circular argument regarding what CPU you think XBox2 will have, you got this bizarre opinion MS has no need for altivec (when it's actually *developers* that will use the CPU, not MS, and they most certainly will), and the Power-series CPU has no altivec, hence power will be the CPU, QED. Problem is, while your reasoning follows a logical pattern, it's based on an irrational assumption and is thus entirely invalid.

I am expecting MS to replace the HD with a 128 MB flash drive directly on the motherboard.

I'm expecting you to have to eat a very large, fat crow come 2005 at the very latest, but by then you will undoubtedly have slithered on towards slimier pastures... ;)

Flash memory, and as little as 128MB of it, will not replace a 6-8GB harddrive, cheap or not. You might just as well say, "let's get rid of all the PCs in a modern office landscape and replace them with typewriters, these are rather cheap these days!", it just can't be done.
 
hmmm, I doubt the hard drive info in this article. That doesn't add up. Also I'm pretty certain there hasn't been any xbox 2 info leaked to developers just yet, a good friend of mine working on a xbox 2 game would be one of the first people to know what these specs are since he's making the game for MS.
 
ERP said:
Speculating for a moment, assuming MS will ship Xmas 2005, and Sony has two options, an Xmas 2006 launch or an Xmas 2205 launch with reduced specifications.
What is in Sony's best interests to do?
Do you expect them to do simultaneous global launches? If not, you need to be a bit more specific about what those launch dates refer to :p

Just something to think about. Personally I'm not sure, waiting untill 2006 is one hell of a risk if you can be competitive in 2005.
A question right back at you. I see two possible scenarios for Sony:
1) Japan 2006March, holliday season 2006 US/Europe.
2) Global holliday season 2005.

CAN they be competitive with 2)?
 
...

Unfortunately for you, all the CPUs - logical or not, as there is no difference visible on the software level
Of course, there is none.

- will have to run a copy of the kernel, or else they can't task switch, and thus be generally inefficient at what they do.
SMP systems have single copy of kernel image shared by all threads. While any processor can technically execute any thread(be it kernel or user), the OS tend to reassign processors to a specific thread for a caching reason.(Thread/Processor affinity)

You still didn't say WHY. Problem is, we all know you *can't* say why, because the reason simply isn't there!
IBM's dual pipe FPU will be just as effective as a 4-way vector unit with a proper compiler optimization, this is why none of IBM's supercomputers are going vector. MS has no use for a vector processor in its code anyway, so why not skip it and cut the die cost? Hell, do you believe MS's code contains lots of MMX and SSE codes even though they have been in PCs forever???

In addition, doesn't MS have to pay additional license fee to Motorola since Altivec is a Motorola IP and not IBM's???

I'm expecting you to have to eat a very large, fat crow come 2005 at the very latest, but by then you will undoubtedly have slithered on towards slimier pastures...
Have you not learned not to insult others???

Flash memory, and as little as 128MB of it, will not replace a 6-8GB harddrive, cheap or not.
MS appears to be going in that direction anyway. Less moving devices = cheaper cost & greater reliability.
 
Very vague. The only thing of interest is 3 IBM processors and 2005.

Hey Deadmeat, since you're so good at this, why don't you give us an estimate of size and power consumption (maybe clock frequency estimate too) of this XCPU2 base on this info.
 
...

Hey Deadmeat, since you're so good at this, why don't you give us an estimate of size and power consumption (maybe clock frequency estimate too) of this XCPU2 base on this info.

CPU : Power5-
Fab Process : 65 nm
Logical Processors : 4
Altivec : no
L2 cache : 512 KB
Die Size : Sub 100 mm2
Clockspeed : 2 Ghz and up.
Power Consumption : 30 Watts???
 
I'd imagine that MS would launch in US first, but this is pure speculation. Xmas is less important in Japan and it's not exactly MS's strongest territory.

I'm sure Sony's preference is to launch in Japan first, but I have to wonder if they feel they can give MS significant lead time in the largest console market. And I find it difficult to believe either one will be in a position for a global launch.

IMO I think the worst possible scenario for Sony is a launch close to MS in a different territory with similar (or lower) quality software. Once they launch they can't play the "wait for our box it's much better game" they played against dreamcast, it'll just be software vs software. Having said that I don't think that even this scenario will give MS a win without other contributing factors.

At the end of the day it's Sony's battle to loose, I don't know of anyone in the industry who isn't expecting PS3 to be the lead development platform this time around.

Speculation is fun though ;)
 
I am guessing that there will not be a built-in Hard Disk. Instead I think we will see a Xbox Live + external hard disk (10 to 40GB) combo. The same Hard Disk could also be bundled with 'Microsoft Entertainment Center' stuff. Later on MS could then add the Hard Disk to each Xbox 2 bundle.
 
So all the strong points of the original XBOX design is being dropped because they lost the first battle?

No direct PC to XBOX2 conversion, knowledge on the PC won't help you on XBOX2?

No HD?

Not possible to play XBOX games, just think about it, the best(?) XBOX game is yet to come, imagine HALO2 being launched a year before the XBOX2

3 core cpu isn't gonna help the developers more than the VU0+VU1.

Seems they are going for a low low price and the plan is to pump Sony for money from the get go....

The weird thing is they may be on to something, how easy will it be to actually see the difference in the next gen. It's not like it's easy now so keeping the price low and the Specs "fair" may be a very smart move.
 
Re: ...

Deadmeat said:
SMP systems have single copy of kernel image shared by all threads.

It'll only be one copy IN MEMORY, but all CPUs in SMP (or whatever-MP), still run the kernel of course, or else the OS won't function. How can you do inter-process communication or taskswitch without it? You can't!

IBM's dual pipe FPU will be just as effective as a 4-way vector unit

So you've said more than once. What actual evidence do you have to back that up?

this is why none of IBM's supercomputers are going vector.

You don't know why they do or don't.

MS has no use for a vector processor in its code anyway, so why not skip it and cut the die cost? Hell, do you believe MS's code contains lots of MMX and SSE codes even though they have been in PCs forever???

What's the majority of code running on a CPU in general anyway, MS OS code, or user code? Certainly a larger percentage of MS code on a PC than a games console, and despite that, SIMD instructions are still so viable Intel's busy extending the SSE instruction set with each generation of its microprocessors! There's absolutely going to be an even bigger ratio of user (read: game) code on a console, and hence, more use for SIMD stuff too.

In addition, doesn't MS have to pay additional license fee to Motorola since Altivec is a Motorola IP and not IBM's???

You're so hung up on this license business, you think Sony licensed Cell from IBM, and now IBM licensed altivec from Motorola? If I were your daddy, I'd smack you for being silly.

Now go read Hannibal's articles on the G5 over at Ars Technica, mmkay? You'll see that while altivec is a Moto trademark, it is in no way a Moto IP, it was co-developed by Moto, Apple and...you guessed it, IBM. Big blue has their own name for altivec, I can't remember it right now, and it's not important either since everybody just calls it altivec anyway.

No need to pay any stupid license fee to anyone.

Have you not learned not to insult others???

Get real, it's not an insult predicting you having to eat crow. (Edit: removed what could possibly be construded as an insult since it was unneccessary to get my point across anyway.)

Flash memory, and as little as 128MB of it, will not replace a 6-8GB harddrive, cheap or not.
MS appears to be going in that direction anyway. Less moving devices = cheaper cost & greater reliability.
(Edit: fixed broken quote.)

Appears? How do you figure? XBoxes still have harddrives in them, don't they? So how could they "appear to be going that way anyway"? And greater flexibility? What's more flexible with a tiny flashrom memory compared to a comparatively huge 6GB harddrive from an XBox, much less what would be viable for a 2005 console?

There's no fact pointing towards MS eliminating the harddrive in XBox2, and the coder you like to quote so much says explicitly he hopes MS keeps the harddrive. Why do you dote so heavily on him in one regard, yet ignore him completely in this other? Can't be that his views on Cell happens to co-incide with your own, while it differs on harddrives now can it? ;)
 
Status
Not open for further replies.
Back
Top