| Written by Mike Hartnett Published at 2009-01-29 13:51:12 Tagged: [Rumor] [Hardware] |
Usually we don't care to entertain bizarre rumors regarding the PlayStation 4, but this story/rumor seems to hold a bit more weight. It appears that some PS4 plans have apparently been leaked and we have them for your viewing pleasure.
Pardon the rough translation from Japanese.
"The PS4 will also continue to utilize the existing CELL architecture, rather than launch something brand new. Bravo, says me, because forcing developers to effectively pick up and move from their established country to Timbuktu every half a decade is a surefire way to cripple launch cycles and leave "un-anointed" third-party developers out in the cold. What's more, a twice-powerful PS3 would almost certainly offer enough "oomph" to keep the plaudits coming from the hardcore faithful, while allowing Sony to more vigorously reach out to the Nintendo "casual" market with a steady stream of cheap-to-develop titles. The idea by 2011, presumably, would be to have a pinched and tweaked CELL SDK that absolutely purrs." blogs.pcworld.com
The PS4 with the development of CPU-Cell B.E.
Sony Computer Entertainment (SCE) is the next game console "PLAYSTATION 4 (PS4)" a, Cell Broadband Engine (Cell BE) are starting to take full consideration based architecture. Currently, much of the reaction GEMUDIBEROPPA, said that the stage of exploring the concrete implementation. The SCE, PS4 generation PS3 production costs are lower than at the same time and generation, seems to be a way to curb investment to continue development architecture.
Cell BE with SCE PS4 are still kept away from the possibility that, for now, Cell BE & expansion seems to be moving in that direction based on the improved version.
SCE was originally a PS3 at the stage of concrete projects, Cell BE architecture PS3 was planning to continue in subsequent generations. Therefore, over the huge cost, to endure over the long term development as a CPU developed the Cell BE architecture. Cell BE-based PLAYSTATION original plan was put in a magnificent sight on the distributed computing network in the end.
The current plan is PS4, Cell BE-based from the original look of the PLAYSTATION + Cell BE as close to the expansion strategy, the reality seems to be no different. In order to reduce the investment in hardware and software, Cell BE-based PS4 is to develop a story. I have already invested a huge Cell BE, the investment in the software stack and tools, is that if you want to inherit would be the aim of SCE. SCE to now, PS3 and development of the same size, would not afford to perform in the next generation.
Moreover, PS4, also reduced the size of the chip itself, to be headed toward a lower cost. PLAYSTATION 2 of the past generation, the two main chip, the die (chip itself), starting at 200 square mm chip's high-performance, high cost and large size around to shrink the chip manufacturing costs and scaling process taking a strategy to reduce the price. However, PS4 generation of low-cost rival to the Wii, lower production costs in a small chip from the first car, and expected lower price.
SCE's direction PS4 or reduce the number of CPU cores
PS4 is, PS3 or the performance will be up for one. PS3 is the Cell BE, for control of the CPU core "PPU (Power Processor Unit)" the one to each CPU core operation for the "SPU (Synergistic Processor Unit)" the seven (originally eight pieces, for a yield disabled) has a total of 8 CPU cores. Then, Cell BE is a scalable CPU by increasing the number of cores, which can increase the performance architecture.
So, PS4 So, CPU core 32 or the numbers and reach. Cell BE is present, 90nm shrink because of the layout while maintaining its basic version, the smaller the degree of reduction of the die. 235 square mm for the 90nm version, 65nm version is 175 square mm, 45nm version is 115 square mm. However, PS4 again to fix the design of a significant number of 45nm chips on the CPU cores can be (SPU Toshiba has significantly reduced the area of core). SCE's original "IKEIKE" Under the plan, 45 ~ 32nm in the 24 core and 32-core Cell BE and had potential. As if Moore's Law, 32 reach the core. But SCE will probably do so.
PS4 is, PS3 for, is likely to remain大人SHII surprisingly improved performance. The SCE's plan, not a high-performance, high cost, it seems to be the direction to achieve a certain level of performance while reducing costs. To minimize production costs, reduce the die size of the chip block (CPU units if 100 mm square), Cell BE-based CPU, even increasing the number of cores is expected to hold. Perhaps, from the current 8-core will increase dramatically. 45nm process DATOSUREBA pay up to 100 mm square in one panel, is likely to remain in the core number of 10. Some frequency is likely to improve, but here is the dramatic and put a big hand to cut the circuit design and architecture. Then, CPU and only, live performance and twice the current 2.x, so will be dramatically cut.
However, CPU core number is still waiting for feedback from GEMUDEBEROPPA, has not been determined. Not decided on the specifications in the research stage. However, SCE PS4 say that we take a cautious route, you hear.
Incidentally, for memory, PS3 employs fewer cases of non-XDR DRAM memory instead, such as generic JEDEC (Electronic Industries Association of America in the infrastructure of the EIA, the semiconductor standardization organizations) plan to use a memory standard said to be emerging. This is because of cost savings. However, JEDEC DDR4 because in fact we have been canceled and仕切RI直SHI, PS4 in phases, in fact, have no choice DDR3 low-voltage version only. However, DRAM memory and CPU implementation of the package stacking or to die, SCE is considering.
SCE, the current generation of game consoles, is struggling big pinch. Although the situation is being improved, the Nintendo Wii is the purpose behind the current generation of machines that carve up the market. Put in a huge investment, high costs and high prices put the PS3 in a predicament because, SCE's PLAYSTATION strategy, is being forced to change. The traditional game, "winning equation" because it would have been destroyed.
So far, the war game had some theories. The competition for high performance on the high initial investment has been favorable to expanding the user下GERE selling prices and higher prices in the first layer with time and casual gamers to the hard-core gamer with a high-performance引KITSUKERE spread, and a graphic. However, Wii's success, he is now a question mark to such theories.
Wii because of the success of the performance that put the price by reducing the cost of investment and venture closer to his new proposals on how to play the content MANMASHININTAFEISU, from the first KAJUARUGEMA intuitive interface and low price and that the target layer. Wii or later, if that's turned the game if the rules of war, SCE must seek to lower development costs and production costs. The strategy will be aimed at立TENAKERE casual layer. The SCE, PS4 If you are trying to lower the cost of hardware, it is estimated that such a background. In addition, SCE itself, unlike the era Hisashi夛良Thursday, difficult to plan for high-risk, potentially low-flow and investment plans.
The Cell BE-based SCE, and other reasons, will release time. Generation now, incremental improvements are possible because of hardware and software, hardware generation change was considered a long span. However, the current SCE suggests that the movement, SCE that have similar plans in the span of the development of traditional and next-generation machine.
SCE is the one that preceded Microsoft's Xbox 360, was leading. Microsoft, when the original Xbox, SCE despite Did you make your best development environment, but failed in the delay. The Xbox was released a development kit is a major GEMUDEBEROPPA PS2 had already been made in the game engine. So, Xbox 360 offers a proposition that the PS3 ahead of the success to take the lead in many areas of the world market. Currently, Xbox 360 is being caught up with the PS3, I could make a significant market base.
If you apply the lessons of the Xbox 360, SCE will have to invest ahead of the next car. The game is to start shipping products from the actual chip development for at least three years, four years is usually required. For example, SCE is attempting to launch in the year 2011 if a plan is quite urgent. The best way to fast development is to improve the existing architecture.
Made a huge investment Cell B.E.
Cell BE PS4 you plan to expand the載SEYOU and the SCE. However, many GEMUDEBEROPPA the current Cell BE CPU, simple way to increase the number of cores do not want to be. Cell BE, but have received positive support from many GEMUDEBEROPPA still is not. Despite the Cell BE SCE give, because they have reason to be so.
Cell BE is for this game to a different CPU. SCE & Sony IBM, Toshiba and purpose of the Cell BE was developed during the game, not only can be used to create a general-purpose CPU architectures. CPU this game, unlike in an airplane rather than abandon the use of every generation, used in the broader market is highly scalable general purpose can花開KA non-game software assets is pursuing a CPU . Of course, for the game, not just one generation, he assumes that you use across multiple generations.
Therefore, SCE is the development of the Cell BE architecture, made a large investment ever. In other words, Cell BE development costs are amortized, PS3 by more than one generation, not a generation, PLAYSTATION diverted to other devices in addition to a great deal would be done from the beginning. For example, PS4 as the Cell BE can provide improved CPU architecture itself, rather than developing from scratch is a small burden. SCE is, Cell BE will be a preference for continued use.
The idea of future development, Cell BE, the architecture is easy to take to improve the performance and scalable. CPU simply by changing the configuration of the bus and ring connected to the core, it is possible to increase the number of cores. CPU without significantly changing the basic architecture, CPU scalable increase the number of cores, you can raise a theoretical peak performance.
Easy-to-deploy and scalable architecture Cell BE
In the standard multi-core CPU, CPU and the problem of increasing the number of cores. For example, the only traffic to keep the cache coherent, it is eaten, the bus bandwidth. Cell BE, which incorporate a mechanism to prevent from the start. The question many of the game software development, dedicated memory for each CPU core, "local store" It is the architecture.
Cell BE,設KEZU the cache to each CPU core to access the local store and explicit, CPU KOHIRENSHIMEKANIZUMU eliminates the need for increased drag on the number of core (Intel's Larrabee is reversed The point is that it has developed a simple extension KOHIRENSHIMEKANIZUMU). And move to the workload of the processing stream of the future, just because the data are not reused, the cache would mean eliminating the most (about prefetched buffer) that were read.
Additionally, Cell BE, each CPU having a separate memory core, and the portable memory, that can scale seamlessly to distributed computing on a network. SPU to access the main memory from the DMA must be disabled介SANAKERE is why there is such.
The Cell BE architecture is the future to facilitate the increase in the CPU core, was to facilitate the way to the distributed computing. Cell B.E.強味have to develop it. SCE as the PS4 in the next hope to develop this architecture, it is natural for a generation.
By the way, Cell BE development of SCE (Sony), IBM, Toshiba engineers in Austin, the company made in the STI Design Center, currently, SCE said AKITEKUTOCHIMU of a few. PS4 architecture for the research, and found that it returned to the headquarters. Therefore, PS4 in the Cell BE SCE is it possible to leave the architecture. However, SCE's, PS4 to have a Cell BE-based study, the results of the research PS4 After determining the direction of where you restart the development.
Reduce the burden of software development lie
Because of this situation, CPU to reduce the burden of development, Cell BE will continue to be a shortcut for SCE. However, SCE has more to worry about, it is estimated that the burden of software development.
SCE is in the PS3 software exceeded expectations and has been suffering the burden of development tools. Additional application development and maintenance of systems software and no end in sight. Compilers and tools can not meet the requirements of developers. BAKKUENDOSABA the system development and the accompanying expansion of network services endless. To reduce the burden on developers of the Cell BE development of software libraries to facilitate the optimization. Multi-core CPU with one of two ISA (instruction set architecture) with special education for the Cell BE programming. The huge burden of software, SCE loom. This is half the job done if the body releases a PlayStation 2 (PS2) and the generation of software development system is very different.
If, SCE to develop the PS4 and a completely new architecture must all start from zero again. Then, the software development costs, they will go up more mad. Than it is to inherit the architecture, and how best to promote the use of software assets and skills that are considered high potential SCE. As a result the PS3 hard to develop software, I finally appointed to a certain extent, SCE as not going to be the accumulation捨TETAKU.
The learning curve GEMUBENDA (proficiency) in terms of architecture, it is desirable to continue. Generation now, the gradual rise of the learning curve SOFUTOUEADEBEROPPA, particularly the PS3 platform is said to have a high degree of difficulty. PS3 titles in development is the only place I'm getting a spur. However, PS4 on the PS3 to have inherited it from, PS4 up the learning curve easier. However, in the GEMUDEBEROPPA, Cell BE architecture itself with a firm, say they can not be certain to increase the learning curve.
However, in principle, along with changes in the software side, the current problem could be alleviated to some extent. The original concept of the Cell BE, according to object-oriented software, 1 CPU was an object that can be assigned to the core architecture. Not by a large task granularity, each object to the small particle size CPU software assumes that you were assigned to the core. The smaller the grain size, Cell BE SPU's easy to fit the size of memory, the principle is to increase the number of CPU cores, will also help to reduce the grain size of the software.
PS4 is our ability to charm other than HADOSUPEKKU
Cell BE-based PS4 has emerged of the plan, SCE's PLAYSTATION strategy. The company has already expanded the concept to be stalling over its failure PS3 merit, PS4 seen signs have a plan. However, too much to keep HADOUEASUPEKKU, PS3 fails to differentiate themselves from the generation, PS4 may fail to start. PS4, it will be passed on to less version, Wii points will be added to one or the other HADOSUPEKKU like a charm."


From the pics and info above, it appears that Sony is currently working on a 32nm chip that will most likely be featured in the PS4, however, if developed early could these chips possible make an appearance in future PS3 and lead the way for a price cut as well as even cooler systems? We shall see. Since Sony is deciding to stick with the cell technology, developers should be much more comfortable with the hardware which means better games and easier development from the get-go.
These details are very interesting indeed and we'll be sure to keep you posted if/when they are substantiated. Until then, we're gonna leave this story tagged as a rumor.
[Source] [Share]
14 comments:
Dana Olson
|
The 10-year plan is rendered of none effect if a new system comes half-way through it. I'm probably one of the few who don't want to hear anything about new systems, let alone a date.
RevChris
|
I think the ten year plan is still intact , I mean the PS3 and PS2 have lived side by side for several years now.
Not 5 years like seems to be indicated above but several.
I've had such a good experience with the PS3 that if a PS4 launched at 5 or 7 years into the PS3 life I would save up the coin to buy one as long as it maintained some kind of backwords compatability in hardware (EYE, Controllers, Headsets, ect) because I wouldn't want to start investing in all new accessories.
Dana Olson
|
@RevChris: I think keeping the PS2 alive while the PS3 starts up is a much different situation than PS4 and PS3 side-by-side. The PS2 had a HUGE installed base before the PS3 launched, and the PS3 is nowhere near that, and so there will be much less incentive to continue developing for it after the PS4 launches - and I'm not into buying just EA sports titles.
Apotheosis
|
Just last night, I was musing that the PS3 was the Saturn of this generation, and my mind wandered to Sony's possible defeat next generation, in comparison to Sega. But reading this article has me realizing that Sony may actually have a serious plan. How reputable is the article? I don't want to be relieved by hearsay... still, this is absolutely fascinating. Just needs a better translation.
omodis420
|
Don't you guys think you should give GamingFront.net some credit for breaking the story. You used their exact sources. Maybe you shouldn't be sketchy and just give some credit. I'm talking to you Mike Hartnett.
RevChris
|
@Dana Olson: But who knows where the PS3 will be by the time the PS4 rolls out. Probably not near PS2 install base but as I see it there are only 2 options 1> PS3 has an install base respectable enough that the similar SDK for the PS4 makes it cost effective to dev for
or
2> the install base doesn't grow enough and Sony just pulls a mulligan with the PS4 and hopes devs are used to the similar SDK (which as I understand the above , the SDK will be similar) and migrate over
Either way I'm ready for the ps4 anytime after 2011 barring some drastic change in the landscape
RevChris
|
@omodis420: Click on the "source" link.
@Dana Below: Well then, perhaps I shouldn't be so smug in pointing out source links.
Something to work on. hah
Dana Olson
|
@RevChris: I fixed it after seeing his comment.
As for the "drastic change" you speak of, it'll happen. You'll see.
Apotheosis
|
@omodis420: Hey buddy, there is a little "Source" link to the top left of the comments... You are not going to believe where the link takes you.
Dana Olson
|
@Apotheosis: I fixed it after seeing his comment.
Aaron Gagnier
|
This stuff isn't exactly new news...if I can remember back, I submitted the tip when this site had less than 10 active users (by active, I mean, actively talking and commenting) and Dana scoffed at me, lol. It's basically all of the same information, just longer...the same pictures are even being used here.
Maybe I sent it to him in an email and not through the site, I can't remember for sure, but I don't see why this is suddenly getting any consideration. I don't see Sony releasing a new console for AT LEAST another year and a half or two years, so anything that's said now is more than likely going to be drastically different that far down the road. I honestly don't even really care about PS4 at this point...PS3 is where it's at, and I'm happy with it.
Dana Olson
|
@Aaron Gagnier: I said I didn't want to post bullshit PS4 news, and I still don't want to. Someone forgot, or something.
Aaron Gagnier
|
@Dana Olson: Tighten the leash on these staffers or you'll get all kinds of weird stuff in here, lol. Oh well though, I don't believe a word of it...nor did I back then...but it's kind of interesting, if that's even possible by taking into consideration it's complete and utter worthlessness :-)
4
43
199
933
If this is true, then the PS3 will have a "slim" version Fall of this year. A new console will come 2 years after the PS3 "slim" version is released. I don't think Sony can release a new console before MS, but the could release in the same time frame, it's a tough call on what to do, but that is why I'm not a business exec.