Probably because when it comes to Apple they may be able to get away with upping the margins. I really wonder how an Apple console would sell at $600. The thing could replace the weak AppleTV and have all those services integrated into it. The machine could have decent hardware and Apple could still make $200 per system. The market may be able to let Apple have decent margins and be profitable from day one. The only argument I really have towards this is that it's Apple. If they started selling houses I would imagine the housing market would see an uptick.
Predict: The Next Generation Console Tech
- Thread starter Acert93
- Start date
- Status
Probably because when it comes to Apple they may be able to get away with upping the margins. I really wonder how an Apple console would sell at $600. The thing could replace the weak AppleTV and have all those services integrated into it. The machine could have decent hardware and Apple could still make $200 per system. The market may be able to let Apple have decent margins and be profitable from day one. The only argument I really have towards this is that it's Apple. If they started selling houses I would imagine the housing market would see an uptick.
DoctorFouad
Newcomer
people tend to believe that history always repeats itself, which is obviously wrong for a very simple reason : people could learn from history and not repeat their own mistakes. it is a characteristic of the human nature, it is what makes humans special : free to choose, anything is possible with humans, they can change their mind anytime anywhere with the same contex, they can act differently....
I still believe that it is in the interest of each manufacturer to make the other believe that its next gen console is weak and far away from release, so they can suprise them and outdo them.
look at sony for example, they try to give the image of a 10 year plan for ps3, that they are concentrating on ps3 and next gen is still far away.....but it is very probable they have already advanced for their next gen project, they will reealse it fall 2013 and I bet God of war 4 and Uncharted 4 are already in development for ps4, and guerilla games are preparing something too....
Why Pitcairn class kind of GPU sounds doable for me. More precisely something in between Pitcarin and cap verde.
10< xx < 20 SIMD and a 192 bits bus interface.
Conservative clock speed 800Mhz at max. AMD tech allows to set the power consumption limit pretty well.
I would not that be that surprised if MS chose to pass on a 256bits bus (especially on a tiny chip) and would feel more comfortable with a tinier chip (below 200mm2).
Going with straight design forward means two pools of memory, expansive GDDR5, overal quiet some memory chips and buses on the mobo, etc. Shortly I can see them cutting corners.
For all intend and purpose there would make (for the intended target) no significant difference. Minor performances tweaks here and there.
-------------------------
Wrt to your comment Blackowitz, I do agree that Bobcat is definitely too weak for an hypothetical CPU powering a Sony SoC. Actually it's been discussed earlier and 3dilletante made the point pretty clear.
Thing I wonder how their successor will turned up. It would be better to almost discuss this in the CPU section but it's kind of related to our topic as AMD is in all the rumors and in some in more than one way.
I've been re-reading an article about Atom vs Bobcat (don't remember the site tho) and I wondered if AMD could use some tricks in use in BD and other in use in Atom.
I noticed that on the atom all the multiply/divide (integer and FP) are done by the SIMD units (if I got it right, that's it...).
I guess good SIMD costs silicon. I guess that why Intel avoid to have duplicate resources.
Intel has a process advantage on every body that includes console manufacturers no matter would they choose for the conception.
I wonder if it would be an option for AMD to share an up to date SIMD between two cores and so taking Intel's approach with Atom.
Overall I would hope that such a trade off would allow AMD (which is not only constrained by silicon budget vs Intel but also man power) to come with an overall more efficiently design chip (and hopefully more efficient chip).
So in my wiew the Bobcat successor would come as a cluster oft wo cores (not a module as in bulldozer) that would share last AMD SIMD unit (so one 128bit wide AVX/SSE "something").
The design goal would be to improve performances while pretty much retaining the silicon footprint.
Pretty I would look forward for overall +20% improvements in sustained perfs within a lesser increase in chip size
.
I do acknowledge that it's still a weak CPU but it's really tiny. Those brazos chip are around 75 sq.mm in size and the 2 cpu cores and their cache take less that a quarter of the die area.
It's doable to fit 8 of those CPU (Bobcat) on ~80 sq.mm using TSMC 40nm process. The ps4 chip is to be produced using 28nm TSMC process I would put the odds for that to happen around 95%.
Clearly even if the core inflate that let some quiet some room for the GPU in case of SoC.
Trinity is close to 250 sq.mm with only 6 SIMD. Not too mention that Trinity might consume way more as the recent review of the "A10 something M" (the 35Watts part) is the highest mobile SKU (highest bins parts). I would assume that the lot of the production perform way worse wrt to power consumption (and that's it the whole chip is functional).
In a lot of case dual core bobcat running at 1.6GHz provide between 1/3 and 1/2 the performance of an Athlon X2 255 running at 3.1 GHz. Not that bad.
So if we look at reworked bobcat running at let say ~2GHz (better process, hopefully better design / reworked pipeline, a bit higher power budget). the figure is not as bad as it looks.
Sony might not be in a situation to compete with MS on an equal footing this time around financial. So coming with a sleek system, easy to deal with (eight threads, X86, coherent memory space, great tools etc.) might prove a sane proposal.
I could see something like this.
Below 200 sq.mm chip
~65Watts TDP.
8 cores base clock speed ~1.8 GHz, turbo up to ~2.2GHz
10 SIMD/GCN base clock ~650Mhz, turbo up to ~800Mhz
the more difficult is the memory interface as trade off should be made. Bandwidth is much wanted but lack of RAM could prove a more dangerous pit hole than lack of muscle vs a stronger competitor.
With a die size below 200sq.mm a 192bit bus is doable (actually I believe even 256 has been made). But I wonder if 192bit bus would fit if the chip is to be shrunk (Nintendo usual doesn't care, and if Sony is comfortable with the cost from scratch they could pass too, especially as cost decrease as newer processes are launched).
I would discard a solution based on DDR3 on a two memory channels configuration, bandwidth would kill the GPU performances. Heavy bottleneck in a system Sony would want efficient in every way.
I'm really iffy about a solution relying of GDDR5 only as that would imply 2GB of ram at max. Consoles allows for net saving but 2GB could be prove too low. I believe that RAM can go a long in leveling huge difference in raw power.
On a 192bit bus I would the (fast 1600 min and 4GB) DDR3 only. Price ~45GB/s would do the trick. Nothing to wow at but Llano and trinity deal with less. ROPs does a way better job than before and more and effort is put in less bandwidth intensive form of AA.
Overall I see this as an option IF Sony doesn't position itself on the high end. That should be cheap to produce and should let them some margins in the pricing to adapt if needed.
I believe that system would keep up with a way more powerful system and I would favor the DDR3 set-up on a 192bit bus on the RAm price and the 4GB of RAM.
If they don't compete for the tech dominance low price is even more critical as there should be less "price elasticity" in that segment.
I would cheap the system without HDD. with some Flash for caching the OS and updates.
I wonder about passing on the HDD and offering Sd-card slots (more than one, four sounds right) instead wrt to the impact on the form factor. For all intend and purpose SD-Cards could act as an HDD for DLC.
10< xx < 20 SIMD and a 192 bits bus interface.
Conservative clock speed 800Mhz at max. AMD tech allows to set the power consumption limit pretty well.
I would not that be that surprised if MS chose to pass on a 256bits bus (especially on a tiny chip) and would feel more comfortable with a tinier chip (below 200mm2).
Going with straight design forward means two pools of memory, expansive GDDR5, overal quiet some memory chips and buses on the mobo, etc. Shortly I can see them cutting corners.
For all intend and purpose there would make (for the intended target) no significant difference. Minor performances tweaks here and there.
-------------------------
Wrt to your comment Blackowitz, I do agree that Bobcat is definitely too weak for an hypothetical CPU powering a Sony SoC. Actually it's been discussed earlier and 3dilletante made the point pretty clear.
Thing I wonder how their successor will turned up. It would be better to almost discuss this in the CPU section but it's kind of related to our topic as AMD is in all the rumors and in some in more than one way.
I've been re-reading an article about Atom vs Bobcat (don't remember the site tho) and I wondered if AMD could use some tricks in use in BD and other in use in Atom.
I noticed that on the atom all the multiply/divide (integer and FP) are done by the SIMD units (if I got it right, that's it...).
I guess good SIMD costs silicon. I guess that why Intel avoid to have duplicate resources.
Intel has a process advantage on every body that includes console manufacturers no matter would they choose for the conception.
I wonder if it would be an option for AMD to share an up to date SIMD between two cores and so taking Intel's approach with Atom.
Overall I would hope that such a trade off would allow AMD (which is not only constrained by silicon budget vs Intel but also man power) to come with an overall more efficiently design chip (and hopefully more efficient chip).
So in my wiew the Bobcat successor would come as a cluster oft wo cores (not a module as in bulldozer) that would share last AMD SIMD unit (so one 128bit wide AVX/SSE "something").
The design goal would be to improve performances while pretty much retaining the silicon footprint.
Pretty I would look forward for overall +20% improvements in sustained perfs within a lesser increase in chip size
I'm assuming the CPU not that vector friendly tasks and from comments here and there it sounds that a lot (most?) of the workloads that are handled in games by the CPU qualify.
I do acknowledge that it's still a weak CPU but it's really tiny. Those brazos chip are around 75 sq.mm in size and the 2 cpu cores and their cache take less that a quarter of the die area.
It's doable to fit 8 of those CPU (Bobcat) on ~80 sq.mm using TSMC 40nm process. The ps4 chip is to be produced using 28nm TSMC process I would put the odds for that to happen around 95%.
Clearly even if the core inflate that let some quiet some room for the GPU in case of SoC.
Trinity is close to 250 sq.mm with only 6 SIMD. Not too mention that Trinity might consume way more as the recent review of the "A10 something M" (the 35Watts part) is the highest mobile SKU (highest bins parts). I would assume that the lot of the production perform way worse wrt to power consumption (and that's it the whole chip is functional).
In a lot of case dual core bobcat running at 1.6GHz provide between 1/3 and 1/2 the performance of an Athlon X2 255 running at 3.1 GHz. Not that bad.
So if we look at reworked bobcat running at let say ~2GHz (better process, hopefully better design / reworked pipeline, a bit higher power budget). the figure is not as bad as it looks.
Sony might not be in a situation to compete with MS on an equal footing this time around financial. So coming with a sleek system, easy to deal with (eight threads, X86, coherent memory space, great tools etc.) might prove a sane proposal.
I could see something like this.
Below 200 sq.mm chip
~65Watts TDP.
8 cores base clock speed ~1.8 GHz, turbo up to ~2.2GHz
10 SIMD/GCN base clock ~650Mhz, turbo up to ~800Mhz
the more difficult is the memory interface as trade off should be made. Bandwidth is much wanted but lack of RAM could prove a more dangerous pit hole than lack of muscle vs a stronger competitor.
With a die size below 200sq.mm a 192bit bus is doable (actually I believe even 256 has been made). But I wonder if 192bit bus would fit if the chip is to be shrunk (Nintendo usual doesn't care, and if Sony is comfortable with the cost from scratch they could pass too, especially as cost decrease as newer processes are launched).
I would discard a solution based on DDR3 on a two memory channels configuration, bandwidth would kill the GPU performances. Heavy bottleneck in a system Sony would want efficient in every way.
I'm really iffy about a solution relying of GDDR5 only as that would imply 2GB of ram at max. Consoles allows for net saving but 2GB could be prove too low. I believe that RAM can go a long in leveling huge difference in raw power.
On a 192bit bus I would the (fast 1600 min and 4GB) DDR3 only. Price ~45GB/s would do the trick. Nothing to wow at but Llano and trinity deal with less. ROPs does a way better job than before and more and effort is put in less bandwidth intensive form of AA.
Overall I see this as an option IF Sony doesn't position itself on the high end. That should be cheap to produce and should let them some margins in the pricing to adapt if needed.
I believe that system would keep up with a way more powerful system and I would favor the DDR3 set-up on a 192bit bus on the RAm price and the 4GB of RAM.
If they don't compete for the tech dominance low price is even more critical as there should be less "price elasticity" in that segment.
I would cheap the system without HDD. with some Flash for caching the OS and updates.
I wonder about passing on the HDD and offering Sd-card slots (more than one, four sounds right) instead wrt to the impact on the form factor. For all intend and purpose SD-Cards could act as an HDD for DLC.
Last edited by a moderator:
DoctorFouad
Newcomer
interesting for a low end system, but I think GDDR5 for next gen is a given, any next gen GPU desperately needs it, even if the system is low end, neither sony nor mcirosoft ca bypass GDDR5 GPU requirements or they would kill the efficiency of the GPU...
look at the RSX in ps3 it is really bottlenecked with its 22.4 Gb/s low bandwidth and that was GDDR3 at 128 bit....
Why are you posting this on every single next-gen thread you can find?
bgassassin
Regular
That would be nice. But I've heard nothing indicating that.
And no it would not be more powerful than PS4 or Xbox 3.
That would be nice. But I've heard nothing indicating that.
And no it would not be more powerful than PS4 or Xbox 3.
Says who!
Until the CPU's are a part of thisc ard game v3.0:
http://shop.cpuwarsthegame.com/
It's just rumors and speculation.. and the best kind
DoctorFouad
Newcomer
DoctorFouad
Newcomer
according to developers that wont happen, they ALL compared the WiiU hardware to ps3/xbox360 level of performance albeit with more RAM and some modern GPU features.
Well bandwidth is definitely important there is no disputing it.
I based my choice on the amount of RAM, say MS really up the bar (whatever the reason) I believe that the extra RAM would come handy to keep up with the next box.
Extra RAM (like 2GB vs 4GB even only 3GB) may allow for plenty of dirty tricks (I'm thinking of effects that would rely on previously render targets), will ensure the same assets are used among both system etc.
If we look at trinity and llano they do way better with mostly the same bandwidth as the RSX. There would be 50% more bandwidth (possibly more if Sony goes with faster than 1600MHz DDR3 that is still significantly cheaper than GDDR5).
Definitely it's a significant trade off. But I think it would be pretty suicidal for Sony to try to compete head to head with MS if the latter decide to go with a loss leading hardware.
If they are to fail to reclaim the perfs crown they may better as well position from scratch the machine as the machine for the budget sensitive gamers. Pretty much moving from it does everything to it's that an affordable entertaining machine runs games for cheap.
I'm serious about them passing on HDD all together and letting costumers use SD cards instead.
Definitely a win for the form factor. With such a set-up I could see the thing behind in the dreamcast ballpark wrt the size.
Depends on many things, at the first spot is price, launch titles and marketing.Indeed.
And if they go that low on TDP, it'll likely match DC in sales too ...
Looking at Sony financial I would be really concerned if they were to use brute force to fight back an hypothetical MS power house. It's a loosing proposal, they are out matched.
They have things going for them in Japan and Europe, among other things free online should be preserved as it seems to not be that successful in that regions.
EDIT
And DC saleswere not that bad, DC main issue was Sega financial shape. If history is any lesson Sony should avoid going that path.
Last edited by a moderator:
If I were in SonyHQ's shoes, I'd have been planning to have a nextgen box designed for 28nm with a target $400 BOM and sell the machine as close to BOM as possible with a launch as soon as the silicon was ready (q4/2012).
This gets the drop on MS, ensures the tech isn't old, and ensures the company isn't losing an arm and a leg on hardware.
$400 BOM can pack quite a punch these days on 28nm ...
I'd have also been investing whatever it took to have feature parity with xblg while also looking to add new features on top of this and introduce their premium online gaming service for $50/yr just as MS has.
They need the additional revenue.
In the ps4 spec, I'd have:
Cell2.0 - 6 SPE's, 6 PPE's (well under 250mm2) @3.2GHz
Pitcairn 800MHz
Split mem pool again (4GB DDR3 + 2GB GDDR5)
BRD
No HDD
Last edited by a moderator:
Well even with a 400$ BOM they will be out matched by MS on all account.
Their network infra is better, they can subsidize more. Basically Sony will have a the lesser system and perceived as such from scratch, offering the same thing as their competitor, at the same price MS would make sure of that. What is the catchy proposal for the costumers?
Definitely taking in account their overall financial shape they should position them selves as the 'value' next gen system. So if they were force to subsidize (say a price war happens) and they comply to it at least they would ship early at a price every household can afford.
Depending on what MS ship I could see them testing the water with a system @299$ including a good exclusive (I could see three bundle GT5, GOW4 and LBP 3 at launch would be great as it would touch many targets), an SD Card with a hand few downloadable game and demos and obviously a controller.
Put an emphasis on the TCO for the buyer (no online fee).
Depending on the reaction if they have to cut price at least the system will be real cheap.
They can fail because as you MS as more powerful system,environment, etc. But this they can't change or compete with. They have to come with a different offering.
By the way on top of BOM you should consider marketing and that costs money. A lot of money.
I'm a bit pessimistic for them because of the overall company shape but I hope that if it comes together well for them to remain relevant. If the OS is sleek, some of their strongest IP at there at launch (or demo are available and they spread the launches to max the benefits), if the system is sleek (catchy shape, sexy form factor, noise is low, etc.), if they advertize free online properly. I can see them having quiet some success. Edit forgot quiet change the tone of the sentence quiet a bit lol.
If they enter the dick contest with MS, which they can't win on any front, most likely at the cost of other important expenses ( software: OS and games, proper advertizing campaigns, etc.), they are done for. For it's almost a given that MS won't pass and bleed them to death once for all.
EDIT Speaking of Sony failure have you ever seen or hear Sony advertize at E3 or elsewhere that the all relevant CoD and BF3 are free to play on their system? That should not happen ever.
The marketing is not doing its work should hammer the press online, paper and as much media as possible over this very fact.
On 7 years (360 life span) the 360 adds an expense of 7x~60$, let say 400$ minimum. It's tough for an occasional online gamer to justify this expense. Going with episodic subscription (more expansive) may result in the same expanse for the costumers. Are they tell? No or no where near loud enough.
Truth is Sony can't afford expansive ads campaign, I see almost no advertizement for them in US. In Eu they were still way behind MS as far as media presence is concerned.
They should not let that happen it's far more important than minor difference in resolution for quiet some gamers and their parents.
Last edited by a moderator:
While this round sees Sony as the underdog, mostly because they are bleeding money, i think there is a chance that they could do a real comeback. First of all, they do not need the fastest most bleeding edge hardware.. well they most likely still will have something that is close. They only need to be on par or close to par and then launch at the same time as the XBOX3 with some solid exclusive games. And they have a real chance of getting a good start with the current exclusive lineup.
And then a proper PR campaign please
One thing I still expect to be an issue is memory stacking. The difference between a console with stacked memory and one without would basically be a generation gap -- you can easily ship an order of magnitude more bw within the same cost and power envelope, and even with half of that the difference will be very visible. So given how much money Sony has poured into the tech, I'd assume they will ship their next gen exactly when they can manufacture a chip with wide stacked memory interfaces in volume.
- Status
Similar threads
- Locked
