What kind of specs would the 360 has if it was released in 2006

Status
Not open for further replies.
I think part of Sony's problem with PS3 is all the extra cost went into Blu-Ray, not higher tech, which people cant see on the screen. It's not so much the price, it's the perception of whether you are getting enough for that price, where Sony is struggling. If you put a monster tech console out there, hell lets say it had "only" one G80, and 1GB Ram, pretty reasonable, I think people would line up to pay 499, 599, for that kind of superiority.

I beg to differ. I'm doubtfull that even if you doubled the graphics potential could impress mainstream users by much or that they could even tell the difference.

Several Devs have commented along these lones back before the 360 and PS3 were released. They said that even if the PS3 was 2x more powerfull on paper, it would be hard to make that difference obvious to the mainstream users. You need something like a 10x jump to do that.

The Wii is a perfect example of this. That console is several more times less capable than the 360 or PS3. Mainstream users do not care. The value of functionality and features is greater to them than the tech specs on paper.

And I would argue that similarly built in BR is more important to mainstream users than a GPU that is 2x faster. Of course people in the HD-DVD camp will disagree, but I think the comparison is valid.

The only people who would line up for a system just because it had a G80, or 1GB or ram, or both, are geeks who are a very tiny minority in this world.
 
fun, a "what if" thread. :)

here is my entry:

XBox 360 @ end of 2006:

- 4-Core PWREfficient @ 3.2 GHz (=OOE-design) with a second Altivec-Unit per Core; 4MB L2-Cache
- 600MHz DX10 GPU with 16 TMUs, 64 Unified Shaders, second-Gen Unified Architecture
- 24MB eDRAM-Chip with 76,8GB/sec-Interface and additional functionality ( eDRAM = enhanced RAM / 3D-RAM )
. => with 24MB 1280x720 or 1365x768 with 64bit Backbuffer and 4xMSAA would be possible without tiling

- 1536MB =>
. 512MB Highspeed GDDR3 RAM @ 800MHz ( 256bit Interface )
. 1024MB Slow-RAM DDR2 RAM @ 533MHz with 64bit Interface ( using the second memory interface of the PWRefficient CPU )

- HD-DVD

- really good cooling solution with 0.2-0.3 Sone max.

- HDMI 1.3

- 20GB HD as standard


Manfred
 
Last edited by a moderator:
The Wii is a perfect example of this. That console is several more times less capable than the 360 or PS3. Mainstream users do not care. The value of functionality and features is greater to them than the tech specs on paper.

I'm not sure it's fair to make that comment. You make it sound like all the people who care about nice craphics are just tech geeks. I'm telling you craphics are very important for a huge portion of the mainstream masses, I'm sure that there is a large mass who feels the way you are saying, but definately not all of them.

Well you might be right about the numbers on paper, but the craphics on on screen definately matters and the jump from Wii to other consoles is big enough for the mainstream people to take notice, the difference between X360 and PS3 however will probably be almost impossible for mainstream consumers to notice, given the fact how closely Xbox and PS2 performed with much bigger gap in terms of power.

Wii is definately something else and despite it's great start, I wouldn't declare it as a winner just yet. Wii will just have to wait and see how far that strategy goes, right now it's a big question mark. It'll make tons of money for Nintendo, but outside of Japan I don't see it selling more than PS3 or even X360.
 
Well you might be right about the numbers on paper, but the craphics on on screen definately matters and the jump from Wii to other consoles is big enough for the mainstream people to take notice, the difference between X360 and PS3 however will probably be almost impossible for mainstream consumers to notice, given the fact how closely Xbox and PS2 performed with much bigger gap in terms of power.

IMHO, this is a reason why PLAYSTATION 3 developers should also worry about finding ways to use the SPE's that are not strictly graphics related: if you are close enough to Xbox 360 graphics (a little above or a little below what the Xbox 360 can do graphically), consumers will still be happy and we must remember that no matter how fast an engine can load, color and display 3d scenes you are just wasting money if you do not have enough content that really shows your users how amazing your product is.
Developers ought to find a way to make their PLAYSTATION 3 games easily discernible from Xbox 360 games with the greater CPU strength they have available (how can we improve interactivity ? how can we make playing PLAYSTATION 3 games a much difference experience ?): using the SPE's simply to move more polygons on screen than Xbox 360 or to have more light sources turned on or to soften a bit more shadows might not be a win unless you achieve such a dramatic difference that each user can pick up and be amazed by it ;).
 
And the biggest disadvantage is its GPU.
And PS3 games aren't going to outshine anything out on 360 in one/two generations and everyone and their momma know it. :) The Xbox360 is a powerful and balanced system as is.

The 360 is a powerful and balanced system, but you are also saying the PS3, with much better CPU performance (even after devoting SPEs to handle some of RSX's weak points), will remain behind all the way through its 5 year life cycle?

Each successive generation of software will improve upon the last, but diminishing returns means each round will probably not be as large an improvement as the previous.

The second to third round first party exclusive titles for the PS3 will have the huge budgets, the time, and the experienced coders to utilize much of the platform's peak capabilities.

By that point, the third or fourth round of first-part exclusive titles for the 360 will have the same, but they will be extracting from a more limited pool, unless you think optimized code can magically create peak performance potential.

We know there are workarounds for many of RSX's shortcomings, and they usually involve devoting something like two SPEs to the task. Even with 2 SPEs or even 3 or 4 devoted to something like that, Cell would still likely be better than the XBox 360's CPU.

RSX's peak numbers are closer to a wash with Xenos. There are some features not present in RSX, but it's not as signficant a shortfall as there is in terms of the CPU.

I'm not certain that under heavy load in future titles that the Xbox 360's CPU won't be holding the GPU back.

This only applies to the big titles that are focused on optimizing for a platform. Most multiplatform games and smaller projects won't have the time or money for all that.

In about 2-3 years, however, it is going to be obvious which big titles out there are for the PS3 and which ones are for the 360.
 
IMHO, this is a reason why PLAYSTATION 3 developers should also worry about finding ways to use the SPE's that are not strictly graphics related: if you are close enough to Xbox 360 graphics (a little above or a little below what the Xbox 360 can do graphically), consumers will still be happy and we must remember that no matter how fast an engine can load, color and display 3d scenes you are just wasting money if you do not have enough content that really shows your users how amazing your product is.
Developers ought to find a way to make their PLAYSTATION 3 games easily discernible from Xbox 360 games with the greater CPU strength they have available (how can we improve interactivity ? how can we make playing PLAYSTATION 3 games a much difference experience ?): using the SPE's simply to move more polygons on screen than Xbox 360 or to have more light sources turned on or to soften a bit more shadows might not be a win unless you achieve such a dramatic difference that each user can pick up and be amazed by it ;).

Therein lies the rub, by moving resources to CPU(interaction physics?) tasks, developers are taking away memory for graphics. The interactivity of the world may increase but at the expense of visual quality. Will PS3 still come close 'enough' to 360 graphics by then?
 
By that point, the third or fourth round of first-part exclusive titles for the 360 will have the same, but they will be extracting from a more limited pool.

I don't know much, but so far I haven't heard very credible evidence that the X360 "pool" is so much more limited that it will manifest into a large difference on screen. CPU is of course very crucial element in a game console, but it's still only a element out of many. Now it seems that X360 has more available memory, maybe has the edge on the gpu front eDram helping, easier tools etc. Sony has the edge on CPU, standard HDD, bigger disc storage capasity etc. All in all it's very hard for me to give either of these consoles a clear edge over the other, especially when the end result aka what's on the screen is the measured value. I again point out that X360 had twice the memory over PS2, GPU that was clearly superior and Standard HDD and yet PS2-games were within striking distance, with this in mind, how can you be so confident in saying that the difference between these new systems will be easily spotted in the future?

I for one will give the honours to the dev-teams instead of the hardware if I'll see some game to be techically superior to other games.
 
I don't know much, but so far I haven't heard very credible evidence that the X360 "pool" is so much more limited that it will manifest into a large difference on screen. CPU is of course very crucial element in a game console, but it's still only a element out of many. Now it seems that X360 has more available memory, maybe has the edge on the gpu front eDram helping, easier tools etc. Sony has the edge on CPU, standard HDD, bigger disc storage capasity etc. All in all it's very hard for me to give either of these consoles a clear edge over the other, especially when the end result aka what's on the screen is the measured value. I again point out that X360 had twice the memory over PS2, GPU that was clearly superior and Standard HDD and yet PS2-games were within striking distance, with this in mind, how can you be so confident in saying that the difference between these new systems will be easily spotted in the future?

I for one will give the honours to the dev-teams instead of the hardware if I'll see some game to be techically superior to other games.

When one game has three zombies in one tiny room, and another has fifteen in an arena, you might suspect something is different. Sure, they may look the same, but something was obviously changed.

When non-graphical features become more important, and eventually people will want more than the little spurt in increased graphics quality, it won't be the 360 that has more to offer.

The lowest common denominator games will not show much difference on either platform. Multiplatform releases will have to keep reasonably in-line between versions.

It's the small stable of big titles that will have a marked increase in features over the previous generation, and the PS3 will have more wow factor in that small selection of "brilliant games"--eventually.

Like I said, it may be too late to matter in the overall business picture of Microsoft, which just wants to gain ground.
 
Therein lies the rub, by moving resources to CPU(interaction physics?) tasks, developers are taking away memory for graphics. The interactivity of the world may increase but at the expense of visual quality. Will PS3 still come close 'enough' to 360 graphics by then?

I think so, RSX is not as awful as some people make it out to be compared to Xenos (especially in the Pixel Shaders department) ;) and right now some developers are achieving great results without using SPE's to amend RSX's "weaknesses". Also, even devoting 1-2 SPE to purely graphics related tasks leaves you 4-5 SPE's at full-time use that can work on non-graphics related stuff (1 SPE taken by the OS full-time and 1 SPE unused to improve yields through increased redundancy, 8 SPE's - 2 = 6 SPE's available to game developers.).

I wonder what developers will be able to achieve once they do more and more stuff in optimized SPE ASM code... SPE's might have a LOT of headroom left ;).
 
Last edited by a moderator:
I think so, RSX is not as awful as some people make it out to be compared to Xenos (especially in the Pixel Shaders department) ;) and right now some developers are achieving great results without using SPE's to amend RSX's "weaknesses". Also, even devoting 1-2 SPE to purely graphics related tasks leaves you 4-5 SPE's at full-time use that can work on non-graphics related stuff (1 SPE taken by the OS full-time and 1 SPE unused to improve yields through increased redundancy, 8 SPE's - 2 = 6 SPE's available to game developers.).
Yes, that's right :yep2:
 
Last edited by a moderator:
It's probably time we stuck to being more exact, that 6 SPe's are available to game developers. 1-2 used on graphics to make RSX perform as well as Xenos (as people think is required) would leave 4-5 for other stuff.

Ok, ok... darn it must mean alchool is still in my system... I'll edit my mistake and please edit your quote so this won't have ever happened...

...


Edit: quote edited.
 
It's probably time we stuck to being more exact, that 6 SPe's are available to game developers. 1-2 used on graphics to make RSX perform as well as Xenos (as people think is required) would leave 4-5 for other stuff.

I do not think 1-2 SPE's optimized to aid RSX in graphics work would just barely make RSX perform at Xenos level, given that RSX's weakness vertex attributes reading (and relatively slow triangle set-up performance) has no SPE-free fixes ever and given also the fact that Pixel Shading-wise RSX is very well endowed ;). RSX is not a weak GPU that needs 2 full-time devoted SPE's just to match Xenos, but we will let people with actual RSX coding experience in real world titles have the last word on this.
 
I don't think MS WOULD change much. Maybe 50 more Mhz on the GPU bus and a little more eDram. They're already competitive with PS3 graphics wise. They don't really need anything else. By the time PS3 is harnessed properly this generation will be nearly over IMO. The other side of the argument is that MS COULD have designed a more powerful console, but why would they when they're already competitive. They still have the option of putting HD-DVD/HDMI into X360 this year if they feel like it, just to remove Sony's last bullet points.
 
Yes, that's right :yep2:

Do we have conclusive proof yet that games cannot use the 7th SPU at all? Or is the real information still under NDA?

One version of that story I read said that the 7th reserved SPE can be used by developers but the OS is allowed completely hijack it as necessary. Although I don't know how they could manage that without also briefly hijacking the PPU to do some kind of context switch on the SPU.

Or many it's only fully reserved for the moment until better libraries that can gracefully handle a context switch on the SPU can be provided. I can even see a compromise like "you must use this particular SCE created SPU job queuing system if you want to have access to the 7th SPE". It would be better than nothing.

But if it really is fully reserved, I hope whatever they are doing with it is worth it! It would be such a sad waste of resources otherwise. I think am going to cry if I hear its mostly idle.
 
I think so, RSX is not as awful as some people make it out to be compared to Xenos (especially in the Pixel Shaders department) ;) and right now some developers are achieving great results without using SPE's to amend RSX's "weaknesses". Also, even devoting 1-2 SPE to purely graphics related tasks leaves you 4-5 SPE's at full-time use that can work on non-graphics related stuff (1 SPE taken by the OS full-time and 1 SPE unused to improve yields through increased redundancy, 8 SPE's - 2 = 6 SPE's available to game developers.).

I wonder what developers will be able to achieve once they do more and more stuff in optimized SPE ASM code... SPE's might have a LOT of headroom left ;).

How helpful do you think ASM codes can work around Cell memory limit, 192mb? That is a real question, no winks ;).

I think Sony and MS will know their respective hardware strengths. One will do physics and one will do graphics, so both will go different ways.

As for this topic, we have to remember that PS3 was also not a fall 2006 console. I dont think they built the PS3 for that date as it got delayed by the blueray specifications.

MS built the 360 with a more conservative dollars spent approach, maybe Xenon with 1.5mb L2 cache and a 64 ALU part Xenos on AMD 80nm and HD-DVD/HDMI, Wifi for the premium. 499/599 same prices as PS3.

PS3 if it was built for a fall 2006 release, Sony could use the time to add 128mb of DDR400 for the OS and a working scaler. Cell could get more refining because i heard Sony stop the Cell(ps3) works quite some time as they had 3mil chips tucked away waiting for blueray to finalize. DD4 Cell at 65nm with 8 SPE?
 
Do we have conclusive proof yet that games cannot use the 7th SPU at all? Or is the real information still under NDA?
It's public knowledge one SPE cannot be used by devs. That could potentially change in the future if Sony find they're hardly touching that SPE after adding in the features they want. That'd probably be something like multitasked OS with audio playback and video scaling, plus...um...video encoding for recording TV. Let's say they put that in and only use 30% of the SPE, perhaps they could free up some of the idle time. I wouldn't hold out for that though. Any reserved SPE time would probably be kept for even more features. 8 way video chat with people on PSPs around the world, while using EyeToy to control the XMB interface concurrently windowed with a game and PIP HD movie playback, streaming audio over the internet through the browser, with silly little men climbing all over the screen (thinking of an Amiga Workbench desktop toy here ;))
 
How helpful do you think ASM codes can work around Cell memory limit, 192mb? That is a real question, no winks ;).

You are asking me basically "how would have faster execution units solve the problem of having less available RAM (for now the Xbox 360 OS needs less RAM, both systems have a total of 512 MB of RAM so do the math ;)) ?"...

I am sure developers are already using quite good compression/decompression routines to load the data from the optical disc/HDD so improving on that area would not save us RAM space (assuming we leave data sitting uncompressed in main RAM)... perhaps more could be done if you stored vertex data, textures and other data compressed in main RAM and use some cycles on an SPE or multiple SPE's to feed data to RSX's VRAM or to other SPE's directly by sending the needed data through the EIB (CELL Broadband Engine's internal ring-bus) with a DMA operation thus saving on RAM space.
Perhaps if you are already keeping compressed data in main RAM you could afford to compress it even more (if it is possible).

There is nothing much else to be said unles we do recognize tht PLAYSTATION 3 developers do have to focus more on reducing RAM usage in their applications than Xbox 360 developers, that is a given (both have to fit their games inside the RAM they have available, just that one system for now as less total free RAM to start with).
 
MS built the 360 with a more conservative dollars spent approach, maybe Xenon with 1.5mb L2 cache and a 64 ALU part Xenos on AMD 80nm and HD-DVD/HDMI, Wifi for the premium. 499/599 same prices as PS3.

What AMD 80nm process?

AMD had nothing to do with the 360, and it only recently bought ATI. Even if the 360 were scheduled for the end of 2006, Xenos would still be on the same process it is now because nobody knew about this move several years ago.

If Xenos being targeted at 80 nm were enough to get enough extra units, and MS went for it, it would probably result in the machine being delayed into 2007.
 
I think likely improvements if the 360 had been designed to launch a year later...
Higher cpu clock speed. Possibly OOE. Had they implemented OOE, I'd expect Xenon to be showing higher real world performance then Cell in most situations, but then it's likely IBM could have used the same design expertise to give at least the main core of Cell OOE as well. Xenon and Cell lack OOE because it wasn't a priority for IBM, not because Microsoft wouldn't have liked it within their $100 million design budget. IF OOE had made it in, I think you would have been looking at a processor under 3ghz due to the additional heat concerns and design complexity. It may have also had to drop down to dual core to save die space. IMO, it'd still be superior design since it'd be far easier (possible) to get better results out of, though only if they could do it without ballooning the design budget. I'd expect 50%-100% more performance per core even at the lower clock which would more than offset the loss of a difficult to use additional core.

-Put it this way, Rogue Leader on Gamecube visually outperformed Halo, and I'd say Rebel Strike also outperformed nearly anything on Xbox. But the average xbox game still blew away the average Cube game. The most common result is the result people see and more important than a 1 time one-up over the competition.

Bigger edram.
The 64 unified shaders already in xenos wouldn't be downgraded to 48 for yield concerns.
Small chance of more system ram.
Harddrive would have been standard.

If a two sku system was still used, the higher sku may have included an hd-dvd drive and been priced at $500 to $600.

Oh, and as for the current Xbox 360 versus PS3, I'd say a fully optimized ps3 game will put out better screenshots, but if the xbox 360 can really do 4xAA at almost no performance hit, it will put out the better looking games in actual motion. Additionally, any advantage ps3 will have in graphics will be so small that it will be hard to tell for most people anyway, whereas the lack of jaggies on the 360 would be clearly noticeable. Ps3 will likely have more good looking exclusives though, microsoft draws too much on the PC game world to get exclusives, and that's really the only way they've been able to launch either one of their consoles with good game support.

PS2 did have a CPU and bandwitdh advantage over the xbox, but overall it suffered because of less ram and older GPU. However, this is not to imply that PS3 would suffer the same disadvantage.

PS2 had a flops advantage, I wouldn't say it necessarily had a cpu advantage. Additionally, the xbox made much more efficient use of bandwidth. The memory and cpu of both were used to match the rest of the hardware, so you can't just directly compare one spec and say it's better. Had PS2 had a modern gpu, it would not have had anywhere near the bandwidth. Likewise, the cpu may have been designed differently.

The Wii is a perfect example of this. That console is several more times less capable than the 360 or PS3. Mainstream users do not care. The value of functionality and features is greater to them than the tech specs on paper.

Even better example, the Wii is supposed to be 3x the power of gamecube, but it doesn't even match it in graphics!

What AMD 80nm process?

AMD = ATI, but anyhow ATI (and I think even nvidia) are already producing graphics chips on an 80nm process.
 
Status
Not open for further replies.
Back
Top