bgassassin
Regular
Yep. It's all an act. I'm making up everything as I go.
Actually I didn't say all of that. All I said was the target specs Nintendo gave to devs.
Last edited by a moderator:
Predict: The Next Generation Console Tech
- Thread starter Acert93
- Start date
- Status
Yep. It's all an act. I'm making up everything as I go.
Actually I didn't say all of that. All I said was the target specs Nintendo gave to devs.
Which are these: http://www.neogaf.com/forum/showpost.php?p=34761975&postcount=4078
The Power PC part I meant from this post http://www.neogaf.com/forum/showpost.php?p=34759264&postcount=4062 and this http://www.neogaf.com/forum/showpost.php?p=34759696&postcount=4067
The Power PC part I meant from this post http://www.neogaf.com/forum/showpost.php?p=34759264&postcount=4062 and this http://www.neogaf.com/forum/showpost.php?p=34759696&postcount=4067
bgassassin
Regular
3 Core 2 way SMT PowerPC? That sounds familiar... I'm sure I've seen that somewhere before.
I wonder how much advantage it'll have over a 3 core 2 way SMT PowerPC XBox 360. I'm sure a fair amount, due to CPU and process improvements, but I don't think more than 2x or so I imagine.
I wonder how much advantage it'll have over a 3 core 2 way SMT PowerPC XBox 360. I'm sure a fair amount, due to CPU and process improvements, but I don't think more than 2x or so I imagine.
bgassassin
Regular
That's part of the reason why Wii U was getting the Xbox 360 comparisons early on. On paper it sounds just like a Xenon with more L2 cache. Considering it's Nintendo though, I expect it to be OoO.
http://arstechnica.com/business/new...s-for-2013-hints-at-a-possible-arm-future.ars
That could get interesting.
Tommy McClain
That could get interesting.
Tommy McClain
Yes. It is.
Allow me to speculate. Back to CES, Masaaki Tsuruta said that Sony is examining 28nm for the next console. There was also a rumor that AMD secured the CPU/GPU for PS4 back in the summer.
Global Foundries taped out a 2.5 GHz A9 cortex on 28nm back in December 2011.
AMD's roadmap from yesterday showed the 2013 successor to Trinity, Kaveri, on 28nm (which I found surprising) consisting of new Steamroller cores and GCN GPU. So the performance APU's will be there on 28nm next year.
Maybe I'm reaching, but it wouldn't be surprising to have the next PS4 consist of an AMD APU with 4-8 ARM cores and GCN shaders for the GPU.
By your own link, the P7 is faster in both integer and floating point performance. David Kanter uses specInt_rate / core as a measure for single thread performance because the specInt benchmark is cracked by auto parallelising compilers long ago.
If you look at the gcc subtest, something that is hard (impossible) to parallise and is actually used alot, Intel's 3.5GHz 2700K (Turbo to 3.9GHz) scores 12% higher than the fastest P7.
You're comparing a 32nm chip vs a 45nm one. You're also comparing a chip designed for desktops with limited TDP to one designed for heavy duty servers.
Also, Sandy Bridge CPUs feature integrated power control of the CPU and GPU parts. If you don't use the GPU, the CPU has more power available to it, - as is the case running a CPU benchmark, so you should be using the 95 W TDP.
It would be a Power 7 derived CPU. Caches, wide fetch/decode, SMT and the scheduling machinery could form the backbone of a console centric CPU. You don't need decimal floating point execution units, you don't need 4 floating point scalar issue ports, you don't need the RAS features of a high end IBM server chip.
Cheers
bgassassin
Regular
Revisiting this, I think that's rather inaccurate. Any bias I might have is irrelevant because first I've never said they would be very similar. I have said that there won't be a repeat of this gen. And I've also said cost and heat will keep Sony and MS in check. Don't let that strange need to "poopoo" on any notion that Nintendo may have actually gone back to making competent hardware cloud your view in other areas.
link?
Seriously where's the Kutaragi quote about MS targeting their old existing console and not the new one they were about to introduce.
It seems quite fitting.
Everything we've seen and heard up to this point WRT WiiU has it being far more similar to xbox360 than what most are expecting from ps4/xb720. The box is tiny. It has a gimmick built in (pad like controller). Their audience is mass-market/casual so it will have a low price. The demos and developer comments all point to a xb360+ ... not a true nextgen console.
This is following the same pattern Nintendo used in Wii with being a generation behind the current gen, but hoping the gimmick will be enough to sway buyers.
The only way WiiU will be in the same class as xb720/ps4 is if Both Sony and MS gimp their own offerings and follow Nintendo's extremely conservative design philosophy.
bgassassin
Regular
The problem with this notion is that to say "gimp" would depend on the person's view of the level of hardware to begin with. For example, what "gimp" to me may not be the same to you. Personally I think its ludicrous to expect a next gen console to have a 1400+ ALU GPU. And from there being under that to me would not make it "gimped". I get the feeling you'll be one of the main ones disappointed with the next gen consoles just from this post.
As for the rest of the post, you're only selecting certain things said like six or seven months ago to say it's similar to Xbox 360 (which have deeper reasons for that). After all didn't we just see a rumor saying Xbox 3 was 20% more powerful than Wii U? (Although I think IGN was missing something(s) in that article.)
Generations are typically defined by a performance metric above the previous generation.
Generally speaking, this is roughly 8-10x the transistor count to get roughly 8-10x performance.
Applying this to ps4 over ps3 we get the following:
Transistor Count:
2000 PS2 ~56m
2006 PS3 ~534m
2012 PS4 ~5,000m
The other measure is die size roughly being 500mm2 and fitting as many trans as possible in this budget on the latest process (at this moment 28nm).
FYI-
Tahiti (352mm2 4300million transistors) has 2048 alus.
This still leaves ~150mm2 or ~1,000m transistors for the CPU.
If I'm disappointed, it will be due to Sony/MS following Nintendo's gimped hardware routine with a gimmick accessory. Not because my expectations are out of line.
Last edited by a moderator:
steampoweredgod
Banned
I hear cell consumes less than 20W, iirc. Each spe can be made to consume 1~W, iirc. Floating point performance and intel's latest overclocked appears to provide...
Likely 100~W consumption for intel i7 part
Cell is above 200~Gflops.
Simple cpu multicore bus designs can scale up to 64 cores according to some papers. At 1~W and low die size spes can be added as needed up to that point.
EDIT
1~W per spe implies 25.6Gflop/Watt for spe component
Gpu comparison says
580 = 6.4 GFLOPS/Watt
So we get 4xGflop per Watt performance, on a more generally programmable unit.
Ageia claimed 200x physics performance increase at some tasks over traditional cpus with a design said to be similar, in their day years ago(a performance claim that seems to surpass that between gpus vs cpus on physics at the time). I also suspect that Gflop for Gflop a processor designed for physics(and ones similar to it) will yield superior performance at that task as compared to ones that are not primarily designed for such, with some exceptions.
I assume for psn software compatibility's sake(necessary to transfer purchases) cell's performance has to be matched or exceeded, which likely calls for more than 3 cores. In all likelyhood we can expect at least 6 spes, the easiest solution, as well as 3 general cores to allow easy code portability with the other consoles(I assume the 7th separated spe functions can be transfered to a regular core.).
If cell was powered by a substantially powerful gpu like it was back at e3 2005(dev kit), rather than hamstrung by a limited gpu as is. Realtime cloth as seen in FFVII demo would be possible. You have to also recall that many a programmer has commented that in many cases algorithm performance advances have matched or exceeded the performance obtained from hardware advances by moore's law. Given that realtime cloth simulation was in its infancy last decade, the added possible performance from algorithm advances should likely not have been exhausted.
Algorithm design advances are likely to continue post the introduction of new consoles too. If sufficient performance is there, an algorithm performance advance can bring things from the realm of the impractical nonrealtime to the practical realtime use.
Last edited by a moderator:
The vibe I get on this board these days is that every dev and person associated with the industry is expecting less potent hardware and better software platforms. The future isn't seen in hardware. Whereas many others are still expecting the console businses to carry on as it has done the last four generations without change and they want the console companies to lose massive amounts of money on hardware. The business side of picking next-gen tech doesn't not point to awesome hardware IMO.
steampoweredgod
Banned
Thing is we can expect there will be no blu-ray scarcity uber costs, thus the disc drive will not add significant costs to nextgen consoles. There is likely to be no uber investment in cpu R&D or gpu R&D, and hopefully no horrible yield, as occurred in this generation.
Microsoft released an uber console, losses obviously weren't that bad, if they were the red ring issue costs would have led to catastrophic losses. What we saw was aggressive pricing and profitability, they even managed to increase initial ram at the request of developers.
Another of the high costs is the Hdd, if flash memory prices continues to fall, a 16GB built in storage could possibly replace this costly and highly damage prone component.
We can also assume amd gpu licensing will match powervr(cents to a dollar vs nvidia's 5 dollars).
So the question is with most costs gone(heavy R&D, low yield issues, high disc drive prices, lower licensing prices e.g. vs nvidia), and using slightly customized parts what can they get within 300-400 dollars?
We know that cpu performance still exceeds state of the art pcs in some parameters, and the gap is likely to grow even bigger as moore's law slows. The power/heat constraints are the main issue, especially with regards to the gpu. Is there any design that can offer very good performance within such constraints?
The powervr thread suggests some gpu designs can likely offer a substantial nextgen(10~)jump within given constraints.
Either that's not true, or we can assume amd can match performance within the same constraints.
EDIT:
BTW,
IT is likely that console manufactures are privy to the differences between two chip solutions and SoC solutions, like happened with the 360(256 vs 512MB ram), dev. input will influence decision. Either the 2 chip solution will not offer a substantial jump in graphics, or it is likely to be taken and later scaled to a SoC. It would be ridiculous for a console manufacturer to do otherwise, as there would be a substantial gap from day one with a competitor that chose otherwise in the presence of the same decision.
Last edited by a moderator:
FWIW I don't think the console industry has ever been about "awesome hardware", it's been about a fixed platform that dev's can use to develop awesome software.
Go to any dev conference, and every dev is moaning about hardware, or dev software, and it's always been that way.
Until recently I was kind of down on next gen, not really having a positive impression on where games were evolving to, but more recently I've been thinking about the possibilities and I'm starting to get excited again.
About the only thing that concerns me going forwards is the cost of building these exciting new experiences, and what that means for small and mid sized devs.
I haven't seen anyone wanting this.
There are many ways to offset hardware costs.
1) higher entry MSRP
2) ad supplement
3) xbl gold-type membership fees
Looking at the last gen where substantial money was lost:
Sony are responsible for the financial mess they are in. They decided to bet the farm on Bluray which cascaded to other decisions which brought up the BOM. Not to mention the mistakes in other industries.
MS screwed up early with stupid engineering design decisions costing them $1b in RRoD.
Die size wasn't what put either of them in the hole.
Note: both Sony and MS die size budget was roughly the same, and MS introduced their box at $299 (core/arcade).
That didn't stop MS from turning in substantial profits ...
So getting back to reality here:
There is no reason to completely gimp the hardware. Even with cheaper hardware (wii bundles were going for $99 this past holiday) xb360 still sold very well which tells you people care about a premium experience and will pay for it.
That demographic isn't going away, it's growing.
You may be right and perhaps Sony/MS will Wii-up here, but that will be more for internal greed purposes of assumed consumer behavior rather than a business plan that no longer works and must be abandoned because they have to lose "massive amounts of money on hardware".
Mind adding power usage increases to each of those three PS'es as well?
PS1 at the time was head and shoulders above anything the PC could do.
PS2/xbox at the time was neck and neck with the best PCs
PS3/xb360 same thing
Now that's not to say the CPU/architecture was easy to get the most out of, or that on synthetic benchmarks, the hardware in consoles is head and shoulders above PC's, but the end result is a box that can play games which would require a top of the line PC to match.
That's what I'd consider "awesome hardware", especially for the price and the time of release.
The only time people did a double take on a new console's hardware is when Wii launched.
- Status
Similar threads
- Locked
