I think the logic of such an approach needs to questioned given the facts that we do know.
We do know that every generation of Xbox prior to this has had cutting edge graphics at the time of release.
We do know that every generation Xbox has been built on the cutting edge process available and encompassed a die budget around 500mm2 (as did Playstation).
That's the traditional console model.
Wii has presented a different approach:
Innovative and novel interface 1st priority and hardware 2nd priority.
This approach worked well to convince many casual gamers to buy the cheap Wii ($250 at launch don't forget), but their spend was significantly less in software for 3rd party and also involved peripheral sales.
Innovative new interface?
_______________
Now in MS' case, I'm not seeing the new innovative and novel interface to build the xb720 around. Kinect has already been launched, had exposure, and good sales. But this was on the existing xbox360. Essentially, MS has already blown their Wii load on the current gen. The can't pull kinect out again and say: "Ah look how exciting, new and fresh this is! YOU are the controller now!! Buy the new xb720 to get one today!". The "revolutionary new interface" mantra is long gone.
That train has passed.
New and improved is the only mantra they can stand on.
You can't package up old hardware and tout "new and improved" expecting to light the fanbase on fire. Especially given the fanbase that MS has built since 2001 which expects their platform to be the best (or near enough) graphics console available.
As for Kinect2, would casuals even be able to tell the difference between the two?
The Casual Demographic
__________________
Speaking of casuals and targeting that market, is it wise to build a console around such a demographic at this moment given the strong pull that tablets and smartphones are making to take casuals out of the market? If we look at the sales loss of Wii, do they directly correlate to the increase in Kinect, or are many of them dropping out of consoles and jumping on board a tablet or smartphone and being satisfied with the experience?
Overall, I'd say it's better to have casuals on board than not, but like any other business, it's important to evaluate revenue stream sources and figure out the best way to tap into them.
For casuals, an xbox360 with kinect is still pretty compelling. Tons more impressive than a Wii.
I think if you shrink the box again with 28nm chips, and put it in a Wii sized case and a nice fat IO for kinect2 then you've satisfied much of what the casuals want, for now. Just to make it all the more impressive, strap a new brand onto the box (Kinectbox) and remove the controller all together. Same guts as a xb360 mind you. Same compatibility can play all the kinect games and for those that choose to, even sync a xb360 controller to the box and play some COD.
Following such a plan lessens the need to target such a crowd with a low price, low wattage, low noise, small box mantra and instead lets xb720 target a clear demographic (much easier to do) and wait for the xb720 price/size/heat/noise to become perfect for casual adoption.
In the meantime Kinect2 is still compatible with both machines and isn't forced onto consumers which do not desire it (though the initial run of gamers will buy it anyway even with the extra baggage of kinect2 ... see Bluray+ PS3).
_______________________________________________________ Broken down for those that don't want to read the entire proposal:
Casuals: Kinect2Box xb360 28nmSoC with usb3 kinect2 in a shiny small new box (essentially MS' Wii) $250
Hardcore: XBox720 500mm2 28nm 200watt Big boy console. $400-500 (sold at break even Price)
Again, when you look at what a 6670 is going to look like in 3 years, 5 years, 8 years...for a console not even coming out for at least a year, probably two.
I bet that is not happening again. The extended generation hurt the industry. I seriously doubt Sony and MS aren't planning to intentionally design for a return to a 5 year replacement cycle again. That's why you are going to see, most likely, a console balanced towards much faster mainstream adoption then last gen while being more profitable early on at the same time.
Wouldn't you rather see what is possible in graphic and computing tech for a console in 5 years, and accept a little compromise now than be locked in with something slightly better for 8+? Or are you planning for cloud gaming to take over by then?
Reading your other posts, I would say that you represent the view of a growing minority, one that is less commercially viable for the console makers than 8 years ago. You are a hardcore graphics enthusiast. That is simply not where the mass market is anymore. Just as the old school hardcore gamers were marginalized by the dudebro fps gamers that followed the entrance of Sony and MS, so are you diehard console graphics enthusiasts getting marginalized now. If you simply must have the highest end gaming graphics then go be a PC gamer, but you want to have aspects of both camps- PC and console. Again, that is simply not where the console market is now.
You need to factor basic market realities like the fact that consoles are now "living room entertainment hubs" being purchased by a whole other category of consumer for whom the gaming function may be secondary, or even less than that, and are looking at these devices as streaming media hubs with some gaming functionality for the kids aside.
When I proposed this lower spec console solution on another forum you were vehemently against it trying to paint me as having some weird minority view. You espoused the the creed "power wins" in determining console success. Now you can see the view I was an early pioneer of is held by the majority.
Ahh, that explains why Wii is still selling gangbusters and xb360/ps3 are both in the bargain bin at $99 and still selling less yoy.
That also explains why Apple invested significantly less on the graphics end for ipad2 than ipad1 and then brought the same graphics light chip to their phones.
Would a 90watt Pitcairn Pro-based GPU be a realistic possibility for either the next Xbox or the PS4? Obviously both MS and Sony are going to be more conservative next-gen, but a mid-range 2012 GPU in a (more than likely) late 2013 console doesn't seem like that much of a stretch, but I'm sure there are factors I'm not considering...
Do you have any proof of this statement? Because half gen more than half the 360 owners playing GeoW 2 did on a SDTV. By what I read here and there the live is full of teenagers tards that don't know better.
It's possible that PS3 got more person who cares that much for graphics and IQ. They had the proper tv want brd etc.
In any case selling M rate games has nothing to do with "hardcore graphics heavy gamers". It's a point taken for granted that is not. Graphics heavy persons have played on PC for while, the most played games are available on pc with IQ that looks next gen already and it costs mostly a 150$ graphics card.
NOt too mention that MS manages to sell HDD less SKU. That doesn't scream hardcore/ technophiles to me.
If anything especially in US I see MS user base consisting more in spotty teenagers than anything else or at least it's every bit as likely as your assumption.
------------------------------------------------
Like the 6670 the 7750 show pretty good results with a limited peak theoretical FLOPS figure. I read both Anand and Tridam review (from hardware.fr). Whereas the 7750 does great I wonder if a manufacturer would deem worth it the increase in transistor numbers/die size versus the previous gen. It's a mix bag it goes from 0% with an average which seems around 10% to great in some compute test (like close to x2 in best case scenario). It comes at 50% increase in transistors number though it would be interesting to see how both generations (evergreen vs GCN) fares with optimized codes in compute operation.
Somehow that manufacturers could think that the sweet spot is lower than most think in regard to FLOPS. What I really don't like in previous "akin 6670 " rumors is the ROPs configuration way too weak. The difference in raw power is low the hd 6670 offers 93% of the 7750 peak FLOPS figures or the other way around the hd 7750 offers 106% of the hd 6670 peak FLOPS figures.
I wish that some would have tested both cards 7750 vs 6670 in a 720P test.
Anyway... we need more leaks about the hardware in test/prod now. I'm still thinking that a tiny SoC + weak GPU might offer a really good for cheap. For the people that want more well me too... that's not the problem is more the launch price. My dream system (still a bit under specced for most) is still a big SoC including a quad-core with strong performances (power& derivative) with a healthy amount of SIMD array (in the hd68xx range not matter the architecture chosen) and linked to main (GDDR5) ram by at least a 192 bits bus. But it's not all in the mood with all the early talk about the system, not only one leak rumors about "hte think killing everything".
So for those that considered a higher price range look elsewhere (and we may agree if I were considering a higher launch price).
I believe that such a system would offer incredible bang for bucks
CPU: 6 cores Xenon 2.
IO sorry for inconvenience devs
4 way SMT
bit shorter pipeline
same cache hierarchy
more cache
various improvement along the board.
Runs slower~2.5GHz
Main word is cheap and tiny
GPU
In two pieces
Akin to evergreen (VLIW 5 could make xenos 5 wide simd emulation easier no?)
16 ROPs are on SoC
128 bits bus on SoC
GPU 1 3 SIMD arrays on SoC
GPU 2 6 SIMD arrays no ROPs
SoC / Gpu 2 bandwidth >32GHz/s
running somewhere between 600MHz and 800MHz
Leitmotiv "be as cheap as possible", the design process could have been something like that:
1) Keep the CPU cores cheap
2) Keep the GPU core as cheap
3) hum really under whelming
4) add some GPU on the CPU as part of it power consumption will somehow blend into the GPU
5) A bit inelegant and asymmetric
6) esthetic in cheap engineering? Marketing guys tell me that it offers "plenty of options"
7) move to cheaper CPU cores
8) fit both chips on the same pad as in nowadays 360
9) as soon as possible (20nm) integrate both chips together.
Overall
Overall barely bigger footprint than the 360.
Tiny high yields SoC
Super tiny super high yields GPU.
Significant increase in CPU throughput at the cost of more pain for the devs.
Doing so MSFT somehow acknowledges the fact that the real cost driver in development is content creation...
Marketing FLOPS figure ~1TFLOPS (akin to the hd5750)
In real term better.
The real question about such a system would be RAM. 1GB is really tempting as far as costs are concerned I would hope they go with 2GBs. Problem is that they may still have "HDD less" SKU which implies some Flash storage... Taking this in account I sadly believe that the intensive to make the flash storage standard while lowering the amount of cache is high...
1 GB of GDD5 is 4 chips, with the SoC and the GPU being on the same pad it would be possible for a cheap and tiny cooling solution to keep everything cool (the 2 chips/1 pad and the 4 mem chips). It also implies a cheaper mobo. And as I don't believe that gamers know better now than some years ago (especially as youngsters keep entering the market) I could see marketing managing to sell Flash as "RAM somehow" say you have 4GB dedicated for caching standard it's a x8 the 360 ram. It doesn't make sense but doesn't that sound like the kind of things marketing guys do?
Overall I believe that 1GB + 4 GB of flash (or more I don't know about the flash chips size) could prove better than most expect as search times on flash are damned low. It would be interesting which turns out the most efficient 1GB of RAM + 4GB of Flash or 2 GB of RAM and some form of caching not being standard (and for me 2GB of RAM +flash would be wiser but I'm not the one either in charge or paying/subsidizing)
I've been there already by I can't help to find it more and likely as well as a system introduced at the same price as nowadays system and if MS subsidizes the system a bit some funky controller may be part of the picture pretty easily.
-------------
TheChefO along both there is a middle ground you have been refusing to see for pages long and on multiples threads.
Your arguments are less than solid, are you sure about who casual are? And who hardcore are? Or to which extend those are traits that can be affected alternately to quiet a healthy part of the overall gaming crown. You'll have to tough time making any point here. Stop presenting your assumptions as facts.
Criticizing the Wii will get you nowhere, both MS and Sony would signed now to get the money Nintendo did... As for Nintendo not doing what it takes earlier for the Wii it's a completely different matter. Trying to characterizing any kind of system that is underwhelming to the the technophiles here as Wii like is close to being dishonest, there is more to the picture, a lot more. A better perception is how a system relate to gaming PC technology. This gen related positively at launch (not for long though), next gen could conservative in this regard. On the other hand high PC are in the wall in power usage, the same was not true in 2005. The Wii is another matter all together, it never related to the tech of its time.
You're arbitrarily discarding some rumors and take some other for granted, arbitrarily. How do you know Kinect 2 will be there at launch? You assume which is not bad per self but what is bad is that you go further and present it as a fact. You did that for a lot of things and it's getting old to discuss in that conditions.
Speaking of Kinect how about keep it around for a longer period of time and try to through various SKUs strategy to offer cheaper upgrade path to users? It's a possibility MS could sell a replacement SKU, you keep you kinect, the HDD if you system had one and go for the bare system (+ a new kind of controller if MS decides it's a good move). A system as I described would have plenty of extra power and storage to throw at some of the problem Kinect faces now, Kinect could be perfected through greater use of plain image processing, in essence blending what they do now with something akin to what Sony does.
MS could go back to the more complex skeleton system they intend ed to use at first (or better) and discarded because of compute power concern and may storage/ram concern. I remember that the library for Kinect only take 50MB of the RAM. You will have more RAM, more bandwidth, more processing power.
It's only a possibility still for all we know Kinect 2 is only a possibility too. MS doing researches is not a proof of anything specific like "Kinect 2" will be there at launch fact and arguing from there.
And repeating a thousand time that 500mm2 is a de facto standard is bullshit, that's a good example of how bad one can make so called truth out of a empirical evidences. It's in no way a proof of anything especially taken in insulation.
Would a 90watt Pitcairn Pro-based GPU be a realistic possibility for either the next Xbox or the PS4? Obviously both MS and Sony are going to be more conservative next-gen, but a mid-range 2012 GPU in a (more than likely) late 2013 console doesn't seem like that much of a stretch, but I'm sure there are factors I'm not considering...
Would a 90watt Pitcairn Pro-based GPU be a realistic possibility for either the next Xbox or the PS4? Obviously both MS and Sony are going to be more conservative next-gen, but a mid-range 2012 GPU in a (more than likely) late 2013 console doesn't seem like that much of a stretch, but I'm sure there are factors I'm not considering...
I think its an order magnitude more likely than something based on 6670. In reality I expect something based more on products we will see in 2013. Not that I expect major architecture changes next year.
Thing is nothing is obvious. If things were to be obvious by looking at past new systems would have be connected to our tv for a while now.
MS would not be in the challenger situation on the OS front for what are now the most selling computing devices.
Sony would not be third, close to bankruptcy.
Nintendo would have done something clever again instead of letting the Wii lose its momentum.
Power consumption would still be cut in half along with transistor density double every new node.
P4 like design would lead the show at 10GHz clock speed.
Next AMD and Nvidia high end cards would burn 500 Watts
And so on
They have a console on the shelf right now that is perfectly suitable for those that think "graphics are good enough". There is no need to invest anything in R&D to appease this crowd.
On the other hand, the hardcore that DO care about graphics will want something to be excited about. Not a $100 graphics card.
Your arguments are less than solid, are you sure about who casual are? And who hardcore are?
No more so than anyone else that doesn't have access to focus group funding. All we have is sales data available for certain products over a period of time and extrapolating reasoning based on the products and sales in relation to their competition.
Criticizing the Wii will get you nowhere, both MS and Sony would signed now to get the money Nintendo did...
Both would like Nintendo's money that they made, but neither would have wanted to trade places with Nintendo at the outset, even knowing what their sales and profits would be.
Nintendo has garbage for an online offering, their 3rd party software sales are pitiful (important for future support), and most importantly, their technology pathway for sales in the future, is bleak.
In the longrun, Nintendo won't be in the livingroom.
You're arbitrarily discarding some rumors and take some other for granted, arbitrarily. How do you know Kinect 2 will be there at launch? You assume which is not bad per self but what is bad is that you go further and present it as a fact. You did that for a lot of things and it's getting old to discuss in that conditions.
I don't know that kinect2 will be there at launch, but I do know they are working on it and it will be available at some point. I also know that kinect has a strong push internally at MS HQ and improved Kinect technology isn't waiting on some external source to come to fruition.
BTW - this is a predict thread. Assumption being that one will predict, not state facts of things which will for certain come to be, despite our convictions of certainty.
Speaking of Kinect how about keep it around for a longer period of time and try to through various SKUs strategy to offer cheaper upgrade path to users?
Kinect or Kinect2 doesn't make much difference as far as software is concerned. It's not like kinect2 will be incompatible with kinect1 games.
Kinect was a great proof of concept and consumers ate it up. Now it's time to improve it.
MS could go back to the more complex skeleton system they intend ed to use at first (or better) and discarded because of compute power concern and may storage/ram concern. I remember that the library for Kinect only take 50MB of the RAM. You will have more RAM, more bandwidth, more processing power.
It's only a possibility still for all we know Kinect 2 is only a possibility too. MS doing researches is not a proof of anything specific like "Kinect 2" will be there at launch fact and arguing from there.
I concede there may be no kinect2 at launch, but there are severe limitations on kinect1 which can be improved with existing technology today.
When Kinect2 is launched will be up to MS, not bound by external factors.
And repeating a thousand time that 500mm2 is a de facto standard is bullshit, that's a good example of how bad one can make so called truth out of a empirical evidences. It's in no way a proof of anything especially taken in insulation.
As I've said multiple times, that is the standard die budget range in 3D consoles, aside from Wii which brought a novel interface to the table, in place of ~500mm2 die budget.
Is it 400mm2?
300mm2?
200mm2?
100mm2?
That's anyone's guess. Mine is based on past experience of what console makers have produced.
Yours is based on the success of one console which had a novel new interface to offset the shit graphics.
Think of it this way, how successful would Wii have been with a DS3 instead of a wiimote?
We've already had this discussion. I've pointed out before that 'every generation' is only two, and two is a statistically irrelevant sample size. I've already explained this isn't the thread for it. People can say whatever hardware prediction they expect for they expected strategy - discussion of the choice of strategy doesn't belong here.
The reason power consumption has gone up in GPU's over the years is because it can. The cards can run at higher clocks and higher performance, and so the gpu makers push for these higher clocks, higher performance, and higher power consumption.
Maybe "conservative" is too strong of a word, but "designed to turn a profit faster" seemed like a mouth full. Given the economy, and various market changes, it's obvious there aren't going to be any $599 consoles this time around. I think we're going to see smarter, better designed consoles, that balance power and price better than their current generation siblings.
(With that being said, I don't believe MS would be crazy enough to put all their eggs in the "Kinect 2.0" basic and release a 2013 console with 6670.)
And there are many ways to achieve this without gimping the die budget.
xb360 at launch was $299 for the core.
This means MS was comfortable* in selling the ~500mm2 die budget console at this price. Keep in mind XBL was nowhere near as successful at that point and their games sales were a very distant 2nd to Sony.
So designed to turn a profit sooner could be as simple as making sure the ~40million XBL members will transition to the new box.
It could mean increasing the base price from $299 core to $399 core.
It could mean ad supported experiences to offset the cost.
None of which require a reduced die budget which will impact the tech in the box.
Ahh, that explains why Wii is still selling gangbusters and xb360/ps3 are both in the bargain bin at $99 and still selling less yoy.
That also explains why Apple invested significantly less on the graphics end for ipad2 than ipad1 and then brought the same graphics light chip to their phones.
A company the size of Apple wouldn't invest in an arena they didn't feel would provide a return. Graphics are obviously important to Apple and I'm willing to bet ipad3 will provide yet another jump in visuals due to further investment in the GPU from Apple. There is a hard limit for improvement on this front though due to conflicting priorities in a mobile device for heat, batter life, and size. Having said that, Apple will still push for better GPU performance to the extent that they are a leader in mobile graphics for hand held devices.
And the reason is simple. Graphics still matter. A lot.
Which brings me back to the Wii, in the bargain bin, still struggling with sales yoy.
I'm not sure if it's the Kinect/Move which caused Wii sales to fade to obscurity, or if it was the mobile devices on the market which are currently providing graphic superiority to Wii in a mobile platform.
Either way, graphics still matter a significant portion as seen by the investments from Apple, and the drop in Wii sales.
The stats don't really back up your claims. Except perhaps in the case of the wii. Sony's sales have been fairly constant the last few years, MS just came off their best year. COD has managed to sell more almost every iteration (it must be irreparably damaged), bioware got a trilogy in on the 360, would not have made that in 5 year cycle.
I expect the people who would want a new box the most are the ones that have little interest in a lower powered box when the one they have is working just fine.
We've already had this discussion. I've pointed out before that 'every generation' is only two, and two is a statistically irrelevant sample size. I've already explained this isn't the thread for it. People can say whatever hardware prediction they expect for they expected strategy - discussion of the choice of strategy doesn't belong here.
Agreed anyway why not simply close this thread. I mean what people expects still have to be based on something.
There is no longer mystery about Nintendo and MS system at least. we won't see any exotic system dual graphic is the most exotic we can expect.
We know it's IBM and AMD.
We know the stuffs are in the oven/foundry.
AMD best up coming parts only includes Southern Island. I would be impressed if AMD would give GCN for somebody to integrate in their SoC and not integrating it themselves especially as APU are critical to their strategies.
AMD is also in better shape than it was last gen, it has implications in $ imho when it comes to licensing costs or willingness to license one last technology especially when previous IP still offer more bang for bucks than competition.
Etc.
If it's about stating whatever one disregarding or rejecting (which is imho in cases at hand close to trolling) any argument without motive and then looping on the same talk is not exactly a fair way to discuss.
Clearly what will be in the next MS consoles is completely business related what can be done with whatever silicon budget is not rocket science, there are products availble using 32 nm and 28 nm lithography. There is not longer that much to discuss on a technical POV alone.
This thread should be closed and a new one should be started for MS akin to the one in the main forum for Nintendo. Business and hardware prediction is linked, either way in this form this topic is dead.
I'd also add timeframe into this linkage as well. Predicting hardware for 2012 (28nm) will be different if MS/Sony are foolish enough to wait for 2014 (22nm).
