Talk:PlayStation 3/Archive 1

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Archive 1 Archive 2 Archive 3 Archive 5

PlayStation 3 versus Xbox 360 versus Nintendo Revolution

I thinking that there must be a section for:xbox360 is beter

  • PlayStation 3 versus Xbox 360 versus Nintendo Revolution

or at least:

  • PlayStation 3 versus Xbox 360 (wait for Nintendo Revolution spec)!!
No, this is not factual and belongs in a consumer review article, not a Wikipedia article. Anything you can come up with at this point is pure speculation anyway. -- uberpenguin 12:54, 2005 May 19 (UTC)
The competition between the systems is covered in the Console wars article. It's a series of factual comparisons of systems, their marketplace success, PR, marketing etc. Sockatume 20:29, 19 May 2005 (UTC)

Xbox 360 will be way better check out this link. It compares the two consoles facts [1]

Above link is in the Xbox section of IGN and was a Microsoft statement, lol.ShotokanTuning 08:30, 22 September 2005 (UTC)

A press document originating from Microsofts PR department probably isn't a good source to base such assesments on. Especially when you consider that full details on the PS3's specs haven't been disclosed yet.

Ati left a statement saying that in terms of power xbox 360 will have more, compared to the ps3, im not sure in what area or if they mean overall since im not that into the lingo for these new systems, i was hoping that someone could check it out and see which system in terms of power has advantages and disadvantages. please use NPOV. EDIT: this post was answered on the xbox 360 discussion board, please check there for the response

I was about to add this link *A side-by-side comparison between Xbox 360 and PS3to the links section of this article when I saw the "Please don't add anymore links without proposing it on the talk section first" note. This link is on the Xbox 360 article and I think it would be very useful here as well. I found myself wanting more information comparing the two systems when I first read the PS3 article and did not know where to go for it. To most of us the numbers in the specs don't mean anything unless we have a point of reference. So if there aren't any objections I will add this link.

As I pointed out earlier, this article is about the PlayStation 3, not a comparison of future competing consumer products. The link only repeats what has already been said in the various press releases and articles currently referenced; it doesn't even seek to add any meaningful interpretation to the technical specs, just copy them. That link might belong in an article that speculates and compares these products, but that is not the job of this article nor of Wikipedia. -- uberpenguin 16:18, 31 October 2005 (UTC)

Firmware

I think there should be some information on possible firmware upgrades that have been hinted, much like the PSP firmware upgrades we've seen.. obviously the PS3 is going to need to compete more as a media centre, also on the UK playstatin 3 website there are hints of OS X being able to work on the PS3, obviously Apple computer have not released a statement about this but they originally said they will not license OS X to anything they didn't make hardware wise.

I can't understand a word you just said... --Wulf 03:56, August 30, 2005 (UTC)

Name

it's PLAYSTATION 3(all caps) not PlayStation 3 (プレイステーション3)

Sony's press release and phrasing in the US always uses the "PlayStation" capitalization. I assume you are gathering the all caps from a direct romanization/translation of the Japanese kana, but Sony has themselves branded the PlayStation in the manner they wish it to be spelled in English. -- uberpenguin 12:56, 2005 May 19 (UTC)
I think you've taken this a little too seriously - the chosen font on the top of the case - the same font used for the Spider-Man 2 logo - has no lowercase letters, therefore they wouldn't be able to use them. Keep in mind that the Xbox is branded XBOX, the Super Nintendo was branded SUPER NINTENDO, and so on. The capitalisation on the box and on the case doesn't necessarily reflect any "official" spelling. It's just a logo. --Squirminator2k 14:38, August 26, 2005 (UTC)

PS3 specs?

You guys'll probably want to edit the parts of this article related to the system's specs, as they've been unveiled (according to PlayStation 3 announced for 2006, at least).


Released specs(source: outerspace.com.br):

Name: PlayStation 3

Release year: 2006

CPU

Cell Processor
PowerPC-base Core @3.2GHz
1 VMX vector unit per core
512KB L2 cache
7 x SPE @3.2GHz
7 x 128b 128 SIMD GPRs
7 x 256KB SRAM for SPE
  • 1 of 8 SPEs reserved for redundancy
total floating point performance: 218 GFLOPS

GPU

RSX @550MHz
1.8 TFLOPS floating point performance
Full HD (up to 1080p) x 2 channels
Multi-way programmable parallel floating point shader pipelines

Sound

Dolby 5.1ch, DTS, LPCM, etc. (Cell- base processing)

Memory

256MB XDR Main RAM @3.2GHz
256MB GDDR3 VRAM @700MHz

System Bandwidth

Main RAM 25.6GB/s
VRAM 22.4GB/s
RSX 20GB/s (write) + 15GB/s (read)
SB< 2.5GB/s (write) + 2.5GB/s (read)

System Floating Point Performance

2 TFLOPS

Storage

Detachable 2.5" HDD slot x 1

I/O

USB Front x 4, Rear x 2 (USB2.0)
Memory Stick standard/Duo, PRO x 1
SD standard/mini x 1
CompactFlash (Type I, II) x 1

Communication

Ethernet (10BASE-T, 100BASE-TX, 1000BASE-T) x 3 (input x 1 + output x 2)
Wi-Fi IEEE 802.11 b/g
Bluetooth 2.0 (EDR)

Controller

Bluetooth (up to 7)
USB 2.0 (wired)
Wi-Fi (PSP)
Network (over IP)

AV Output

Screen size: 480i, 480p, 720p, 1080i, 1080p
HDMI: HDMI out x 2
Analog: AV MULTI OUT x 1
Digital audio: DIGITAL OUT (OPTICAL) x 1
Disc Media
CD: PlayStation CD-ROM, PlayStation 2 CD-ROM, CD-DA, CD-DA (ROM), CD-R, CD-RW, SACD, SACD Hybrid (CD layer)
SACD HD, DualDisc, DualDisc (audio side), DualDisc (DVD side)
DVD: PlayStation 2 DVD-ROM, PlayStation 3 DVD-ROM, DVD-Video, DVD-ROM, DVD-R, DVD-RW, DVD+R, DVD+RW
Blu-ray Disc: PlayStation 3 BD-ROM, BD-Video, BD-ROM, BD-R, BD-RE

ren 01:58, 17 May 2005 (UTC)

Images

I uploaded this image, and placed it under fair use. Can someone advise me if this is ok? I won't add it to the article yet as I don't want to get Wikipedia into any legal trouble. What with companies like apple sueing the hell out of rumor sites. Can we use this under fair use? Should it be deleted? Jacoplane 14:25, 15 Mar 2005 (UTC)

What is the source of the picture? Wikipedia isn't a rumour site, so unless we have good reason to believe it is genuine we shouldn't use it. Thue | talk 15:01, 15 Mar 2005 (UTC)
Well this is most likely not the real design. I found it on the playstation 3 forum site. I guess we won't use it then. Jacoplane 15:29, 15 Mar 2005 (UTC)
Added the delete template Jacoplane 15:49, 15 Mar 2005 (UTC)
After careful examination, I've concluded that these are merely very elaberate photo-manipuations, probably put together in Photoshop. If you look closely, all three images have the same reflection super-imposed on top of them, mentally strip that reflection. I doubt that there would be a need for such things if these were photo's of a machine that actually existed. Plus, this image is missing things like, controller slots, a power butten and has the words, while it's theoretically possible that these are located on the back of the machine, I somehow doubt it, considering how much fans complained about the power butten on the back of the PS2 and all. WolvenOne
Well, for the sake of argument, it's reasonably likely that the controllers could be wireless. As for the power button, it could be a touch-sensitive thing like on Apple monitors (also, the actual power switch for the PS2 *is* on the back of the machine). That being said, I don't think it's a real photo either. Kairos 07:42, 14 Apr 2005 (UTC) EDIT: On a closer examination, I'd say the red light is indeed meant to be touch sensitive power button in this image. The blue light is labeled "EJECT", and I don't see much reason for there to be an indicator light for that.
Odd how the reflection is a trolly or subway train


$3Billion Cell?

I removed a statement that seemed to suggest the new processor for the PS3 will cost $3Billion to produce, but the $3Billion refers to the cost of building the new manufacturing plant in East Fishkill. At least this is what I deduce from reading the article on the BBC website [2]
-- Camster342

Anybody notice parallels between the Fifth generation computer systems project and the massively-parallel Cell chip used in the PS3? — ChrisErbach

I uploaded up-to-date PS3 rumor on Internet. --Mateusc 01:19, 24 Mar 2005 (UTC)


Ram Wars

Memory tapped to make PS3 chips 512MB XDR DRAM

Ram Wars -- Gamespot article above asserts that PS3 has 512 Megabytes of RAM; they claim this is from a Reuters article. I wasn't able to find a Reuters article that gives a figure for the total amount of RAM, although there are Reuters articles in English mentioning that Elpida will make XDR RAM for PS3 (specifying neither the size of the chips nor the total), and in Japanese mentioning the 512 megabit chip size but not the total. pc.watch.impress.co.jp quotes unnamed sources as expecting that the box will contain 4 512 Megabit chips for a total of 256 Megabytes, and goes on to say that Elpida, Samsung, and Toshiba all have limited capacity to produce 512Mb chips and Sony may be forced to use 256Mb chips for a total of 128 MB. I am more inclined to trust the PCWatch article since it is much more detailed, and since I can't find any Reuters article backing the gamespot interpretation. Willhsmit 21:14, 15 Feb 2005 (UTC)
It is INDEED, the 512 Megabit XDRAM memory, 64 Megabytes per unit, with up to 4 per Cell Chip, for a maximum of 256 MB. The schematics are readily available for that. Gamespot doesn't know what they're doing. Terrapin 05:37, 15 Feb 2005 (UTC)
Good clarify, I didn't get what the user was trying to state ("PS3 has 512 megabit RAM" parses as "PS3 has 64 megabytes of RAM" in my head), and that Gamespot article was dodgy. Sockatume 18:02, 15 Feb 2005 (UTC)
Strange, when I translate that page with Google they seem to suggest that the console will use 128MB of Ram, not 256. Of course I may just be reading it incorrectly. Anyhow, after poking around at the website for awhile, and translating a few different pages, I've noticed that PCwatch appears to speculate alot on future products. Now there's nothing wrong with that, but because they speculate alot, I think everyone should just assume that we won't know for certain how much RAM is on the thing until Sony makes an official announcment concerning the inner workings of the PlayStation 3. WolvenOne
Nothing is known about the configuration of PS3, not even the number of SPEs per Cell Chip, so PC Watch is off its rocker. Terrapin 05:40, 16 Feb 2005 (UTC)
My interpretation of the article is that PCWatch speculates that XDR RAM shortages will make Sony reduce the amount. Personally I'd bet that Sony would delay the console a few months rather than ship with less than 256...but my main point was that the Gamespot article has no confirmation, conflicts with what is known about XDR manufacturing, and could easily be a misprint. I actually got a live person there who promised to get back to me. IMHO, the main article could either present competing rumors or just skip over the issue for now. Willhsmit 21:14, 15 Feb 2005 (UTC)
Unfortunately, PC Watch is using TODAY'S shortages of XDR Ram as its source, and not the new processes coming out by Toshiba for 512MB, which they're testing now, and will be available two years hence. Listen up, guys, IGNORE anything from PCWatch or Spong, or Gamespot. Terrapin 05:40, 16 Feb 2005 (UTC)
I agree that Sony would probably delay the release a few months rather then go with less the 256MB's of RAM. Considering that both Microsoft and Nintendo are likely to throw at least 256 in thier machines, and considering that the PS3 is likely to be the last of the 3 next generation consoles to be released on the market, it'd simply be inconcievable to me that Sony would allow itself to be beat out on RAM. Of course, they may not delay it, they might just launch the system and allow there to be massive shortages, much like they were for the PlayStation 2 during it's initial release. Of course though, this is just more speculation and has no barring on the article. WolvenOne
It's concievable that the PS3 could have more than one Cell chip (that's what they're designed for, after all, shoving as many as you need in a consumer device instead of using different powers of chips in each one) with more RAM available to it. If it does have only one Cell, 256MB would be fairly slim. I'd like to know how much VRAM it would have before making judgements, though. Sockatume 23:32, 15 Feb 2005 (UTC)
Indeed, it has been widely speculated that as many as four Cell proccesers could power the PS3, though I feel that a full four would be unlikely considering the increased cost and the the fact four Cell's could be considered overkill. If there are multiple Cell processers, my logic tells me that it'd be limited to two, based on what I feel to be the likely cost of each processer. I should point out though, that this is merely more speculation on my part. Speaking of speculation, perhapes somebody should create an area in the article purely for speculation about PlayStation 3 components that do not yet have solidly confirmed specs. After all, there is quite alot about this console that we don't have solid information on yet, and much of what we do take as facts are merely rumors or came from unconfirmed sources. WolvenOne
Agreed -- the presence or absence of a hard drive, the number of Cell chips, the resemblance as far as speed and number of SPE's between the Cell demonstrated and the shipping Cell, and the density and number of RAM chips are all speculative. Is the width of the memory bus at 64 bits hard info? I believe the maximum of 4 RAM chips is based on a 64 bit bus, 16 per chip. Gamespot has corrected their article to read 512 Megabit, by the way. Willhsmit 00:18, 17 Feb 2005 (UTC)
Well, the specs for the X-Box 360 came out today, and they list the amount of RAM as 512MB's. Now of course, this isn't Sony, however I very much doubt that Sony is going to allow themselves to come out a year later and have less RAM then thier competitor. So I expect Sony to up the density of thier RAM from 64MB to 128MB per chip. WolvenOne
It seems this discussion is rather outdated... The PS3 will use "256 MiB Rambus XDR DRAM clocked at CPU die speed (3.2 GHz), 256 MiB GDDR-3 VRAM clocked at 700 MHz"... --Wulf 04:24, August 27, 2005 (UTC)

OpenGL, etc.

nVidia has stated recently that the GPU for the PS3 wouldn't use either Direct X or Open GL, and would instead use Sony's own API. With this in mind the section marked PS3 standards, at least needs to be edited and may need to be deleted altogether.

Source? Sockatume 14:32, 13 Feb 2005 (UTC)


Source: http://www.xbitlabs.com/news/multimedia/display/20041228125957.html

It should be noted though that Nvidia was trying to cover for one of it's employee's claiming that the PS3 would use Direct X and also have a few other features typically only seen in Windows machines. We really don't know what features Sony's API will utilize so there may indeed be some open standards within it.

Nvidia has said that the GPU will support basically the feature set of NV40 with some more advanced features. Supporting DirectX and OpenGL or not is basically an issue of the drivers -- I doubt that Sony could port DirectX to the Cell without Microsoft's support even if they wanted to. I'm sure the underlying GPU hardware will support most or all of the DirectX 9 features that NV40 does for the PC, but I doubt there will be software support for calling DirectX 9 APIs.
What Nvidia said exactly is that the GPU would feature technalogy from thier current generation of GPU's known as NV40, as well as technalogy found in the NV50 architecture. However because these are rather vague statements we do not know exactly what architecture will be used as the basis of this GPU. So stating that it'll be a NV40 with a few tweaks taken from the NV50 lineup is mere speculation. It's just as likely that the GPU architecture will be customized to take advantage of the Cell processer, though that'd be speculation as well.WolvenOne
PS3 will not use DirectX. While OpenGL is strictly an API, this is not true of DirectX. DirectX is also incorporates hardware abstraction and software implementation of various aspects of 3d rendering. It's also highly unlikely that DirectX could be used without a windows kernel. Even if it were techincally possible, it wouldn't make any sense for PS3 to license technology from MS when there are perfectly reasonable alternatives. Timbagas 07:41, 28 Mar 2005 (UTC)
The PS3 will use the OpenGL ES API.

I altered the page to accomodate some confirmations on the PS3 using OpenGL ES 2.0 as well as Nvidia's Cg shading language. Both of these were confirmed by PlayStation 3 to be easy on developers, Sony vows, and the headline at [3] "Sony announces support for OpenGL ES 2.0 in PlayStation 3".

--Mateusc 03:50, 17 May 2005 (UTC)

PlayStation 300

Someone moved this to "PlayStation 300". That doesn't appear to be the system's name, so I moved it back. Please check your facts carefully. From a marketing standpoint it would be pretty dumb for Sony to number their system 300 when Microsoft's is numbered 360.. people will buy the higher number :)

Vandalism.. the oficial name is PlayStation 3. --Mateusc 01:12, 17 May 2005 (UTC)
Isn't that why it's X-box 360 and not just X-box 2? ShotokanTuning 06:28, 13 October 2005 (UTC)

Microsoft did not want the xbox sound inferior to ps3 in any way, so instead of having xbox 2 and PlayStation in the 3rd generation, xbox went with a new play on words, hence xbox 360. Tik 18:57, 4 December 2005 (UTC)

Console & Controller

Sony's reveiled the design and build of the PlayStation3 controller and console.

Got these from Gamespot. Pretty sure everyone knows of them and therefore can trust their information. Here are some links to the pics from the forums.

Console and Controller http://img191.echo.cx/img191/3848/e32005ps3official2005051605324.jpg Guy holding the PS3 http://us.news3.yimg.com/us.i2.yimg.com/p/ap/20050517/capt.ksd10105170114.video_game_expo_ksd101.jpg The controller. http://image.com.com/gamespot/images/2005/news/05/16/ps3_screen008.jpg

On a side note; look at how big that thing is!! The controller looks like it wants to come alive and muder me. It also looks like it has handlebars.

Looks like a banana to me. A big grey banana. I think Sony's crazy for changing a good thing. Rhobite 02:59, May 17, 2005 (UTC)
looks like a boomerang and the console seem a barbecue grill. I sincerelly wait for DualShock 2 support. --Mateusc 03:50, 17 May 2005 (UTC)
I'm more concerned about the design of the Console itself. It looks like a Trouser Press. --Squirminator2k 14:41, August 26, 2005 (UTC)
Hmmm... The September GamePro (or was it GameInformer...) shows a picture of the controller actually being held by a hand... It seems to be not much larger than the DualShock. Also, I beleive Ken Kutaragi has said it is merely a design concept. --Wulf 04:51, August 28, 2005 (UTC)


How does this controller work?? do you plug it into your PS3 when you need to recharge it, or do you replace two AA batteries?? I admit it looks kinda ugly but at least there are no sharp corners to get hurt on or to get yucky stuff stuck in (like the old PS1/PS2 controller), also I think that when you begin to use it you will feel more comfortable and apreciate it's design. maby they were trying to incorperate the aboriginal boomerang into it!

But it has a 24 hour charge according to arstecha

Can you back that up with a link to a reputable source? (i.e. one that isn't just extrapolating or speculating) Otherwise it has no business in the article. "Sources" is dangerously close to a weasel word. -- uberpenguin 19:34, 2005 May 26 (UTC)
I suspect that the final word there is mean to be his source. Could he mean Ars Technica? Sockatume 17:16, 3 Jun 2005 (UTC)
I figured as much and did a quick search of Ars Technica, which turned up nothing. Consulting Google, all the articles that mention the number "24 hours" say "nearly 24 hours on a charge" (emphasis mine). Furthermore, none of them cite a verifiable source, but rather state it as if it is assumed fact. All this seems pretty spurious to me, and unless someone can turn up a source I'll remove the text. At best this is probably Sony marketing and the phase should read "Sony has claimed that the PlayStation 3 controller will function up to 24 hours on a charge" (if it can be verified that Sony did, indeed, claim this), rather than the weaselly "Sources claim [...]" I know I'm being a little anal, but if we weren't picky then this article would have been reduced to a pile of speculative trash by now (a quick perusal of the edit history backs this up). -- uberpenguin 17:31, 2005 Jun 3 (UTC)
Just to clarify, even if the 24-hour run is a popular rumor, we shouldn't include it here unless it can be positively attributed to Sony. -- uberpenguin 17:34, 2005 Jun 3 (UTC)
Agreed on all points. Sockatume 22:57, 3 Jun 2005 (UTC)
Why doesn't the console have ventilation holes?
It does have holes MS started a rumor that it didnt. See for yourself. http://en.wikipedia.org/wiki/Image:Sce07.jpg Taladar 22:48, 15 November 2005 (UTC)

I just notice that Sony was too lazy to come up a new font for their PlayStation 3 logo that they just borrowed the font used in their Spider-man movie logos. Just want to point out that because I found it kinda interesting. --Chill Pill Bill 03:01, 17 May 2005 (UTC)

Watch the conference in gamespot, Spiderman appears many times to demostrate PS3 capabilities --Mateusc 06:58, 17 May 2005 (UTC)
Heh... we were watching the conference thinking the same thing ^_^ It's not even changed to appear unique. --131.7.251.200 19:24, 17 May 2005 (UTC)
And font forget that you can get Spiderman2 on the PSP if your one of the lucky ones who pre order. Ablaze 08:58, 24 May 2005 (UTC)
I like the font anyway and couldn't care less if it got changed. If they do make a new font I hope it will be one where people think that is the PlayStation 3 font. • Thorpe • 15:08, 23 July 2005 (UTC)
I was told by several people at E3 that the font is a placeholder. And since they own the Spidey font, they used that to tie in with the spidey demo. Terrapin 17:09, 24 May 2005 (UTC)

How odd.

well since the controller looks like a bataarang i think they may create a superhero system that when not in use fights crime and takes out the trash.....very admirable indeed sony.

Ethernet

What does this mean?

Ethernet (10BASE-T, 100BASE-TX, 1000BASE-T) x 3 (input x 1 + output x 2)

One-way Ethernet (either in, or out) sounds rather futile, by the very nature of Ethernet: is this a way of saying that the PS3 can act as a firewall/router/switch, with one "outside" and two "inside" ports? -- Karada 12:22, 17 May 2005 (UTC)

According to E3 conference the PS3 is going to be some kind of switch maybe to connect 2 together at a future date? 1 imput means thats where the ethernet cable goes in and the other two ports are for where you would connect the device that you would use, maybe a external hard drive, or another PS3, or some other goofy ethernet device.-- mattrb
I updated the phrasing to reflect this. Saying "input" and "output" ports is fairly misleading, since each 1000BASE-T port is bi-directional anyway. -- uberpenguin 20:42, 2005 May 17 (UTC)
No idea whether this sheds any light on things, but Phil Harrison, when asked about why the PS3 has three network ports says:
Because it can be a hub, rather than just being a terminal at the end of a network. Also, we want to be able to have a Gigabit port for an IP camera. So one of the ports is an 'in', and two of them are 'through'. It can be a server as well as a terminal. Sockatume 20:35, 19 May 2005 (UTC)

Macintosh computers have had ethernet ports for years, why is Sony deciding to use them now??

So have PCs. It's what's called an "Industry standard". There's an Ethernet port on the PS2's aptly named Ethernet Adapter, so it's not a "Just Now" thing. Ethernet is all to do with networking and online play which, until the current generation, hadn't really kicked off before. --Squirminator2k 14:44, August 26, 2005 (UTC)
Agreed about being an industry standard. But the PS2 "Ethernet Adapter" you mention, is actually called the Network Adapter (although it does have an ethernet port). The Nintendo GameCube has the Ethernet Adapter. --Wulf 22:38, August 30, 2005 (UTC)
Any additional feature raises manufacturing costs and lowers profit margins. Therefore you wouldn't expect for something like, say, broadband-readiness to be a stock feature in consoles until broadband internet has reached a reasonable level of target market saturation. Like, oh, say... now. -- uberpenguin 01:11, August 31, 2005 (UTC)

uberpenguin edits - wrong !

"12:32, 17 May 2005 Uberpenguin m (→System Bandwidth - cleanup, I have no idea what "SB" is, and the press release doesn't make that clear)"

If I didn't understand something, I'd be a bit cautious in just removing references to it. Also, changing 7*SPE 3.2Ghz to 7*SPE 1.2Ghz seems wrong
Suggest someone reverts these changes, and the other ones by Uberpenguin at 12:32. --Sgkay 12:47, 17 May 2005 (UTC)
I'm referring to Sony's own press release [4] for this information. If you have a better source, please post it. Also, I did NOT change the SPE clock speed to 1.2 GHz or PowerPC to PC, User:204.9.144.52 did that. Please check the edit history before you go pointing fingers at people. Furthermore, I suspect "SB" refers to System Bandwidth, which is a marketing term more than any useful metric. Without knowing exactly what this means, it isn't useful to include it in the article. -- uberpenguin 12:51, 2005 May 17 (UTC)
Additionally, please forgive any typos I make while cleaning up, I will triple-check my work and correct any mistakes that I make. Someone just copied in the entire hardware section of the press release without formatting or properly clarifying it at all, and I'm trying to fix it. Please bear with me for a few minutes... -- uberpenguin 12:54, 2005 May 17 (UTC)
Read the Binary prefixes article first. When someone lists a specification like "256 MB" they assuredly mean Mebibyte, not Megabyte. There is a lot of confusion on this issue, which is why I changed the measurements to the correct binary prefixes. 256 MB is supposed to mean 256*2^20 bytes, not 256*10^6 bytes. Using the binary prefixes removes ambiguity and is starting to become the standard with other storage and computer articles (see the hard drive article). If you seriously object to using the correct binary prefixes, please explain why. -- uberpenguin 13:02, 2005 May 17 (UTC)
Oh, I'm sorry... I see that you were reverting User:204.9.144.52's edits, not mine. Well, it's still useful to explain why I changed to binary prefixes. -- uberpenguin 13:04, 2005 May 17 (UTC)
Incidentally, I would like to rewrite parts of this article to make it slightly more than a big list. Few people want to read a list of specifications that they may or may not understand, and it is certainly against Wikipedia stylistic guidelines. I realize that there isn't terribly much to say about the PlayStation 3 as of yet, but the format and flow of the article still can stand for some improvement. -- uberpenguin 13:08, 2005 May 17 (UTC)
Okay, let me explain why I removed the "Overall System Floating-Point Performance" section. There is no way of telling how this number was arrived at, and likely it is a creative concatenation of the theoretical floating point performance of the CPU and the graphics processor. This sort of thing is nothing more than a marketing figure that is designed to make people drool over without knowing what it is they are looking at. Furthermore, the text "twice that of the XBox" is very misleading in that it suggests that FP performance is the only metric by which computers (of any kind) can be measured against one another. Do integer and DSP performance account for nothing? For the time being I'll leave it, but unless someone provides compelling reason to leave it, I'll remove it again. Let's keep in mind that it's not the purpose of Wikipedia to mirror press releases that contain a healthy dose of marketing. -- uberpenguin 13:29, 2005 May 17 (UTC)
Fair comments. At the moment, Sony's details of the server are pretty much all we've got to go on. Lot of interest in what the specs are though (people will be comparing specs with the Xbox 360). Hadn't realised the MB/MiB thing, point taken. --Sgkay 15:20, 17 May 2005 (UTC)
Thanks for the understanding, I'll do my best to tread carefully and justify anything I do with this article. -- uberpenguin 15:49, 2005 May 17 (UTC)

2.18 Teraflops

Based on slideshow Sony E3 Conference, with comparisson Xbox 360, A64... I will back with information, and uberpenguim, watch the conference in Gamespot It's free before doubt.. It's oficial slideshow from Sony presented by Ken Kutaragi. --Mateusc 16:38, 17 May 2005 (UTC)

I'm not questioning the fact that it's an official Sony claim, I'm saying its a worthless measure of system performance. Knowing that Cell can perform some constant times the number of FLOPS of the XBox 360's processors doesn't indicate whatsoever how one console will end up stacking up against each other. Furthermore, as I already said, this number seems largely a marketing fabrication since we have no indication of how they came up with "overall system floating point performance." Being that we can't verify how that number was arrived at, and it is theoretical at best, I don't believe it should be included. I certainly believe that drawing the comparison between two single-dimensional and questionable metrics to compare two yet-unreleased game consoles is inappropriate. So for now, I'm going to remove only the XBox 360-related phrasing, and we can continue to discuss whether or not it's appropriate to leave the TeraFLOPS metric. My argument against it follows in nice, fun bullet point format:
  • We don't know how that number was derived, and it is theoretical at best. I already have some qualms listing theoretical maximum memory rates since they will NEVER be attained, but I figure that's a small nit to pick.
  • The metric is no kind of indication of system performance, it only is impressive if you are running something like LINPACK on your game console. This is the same reason you can't definitively say that Blue Gene/L or Earth Simulator, for example, are X times as fast as your x86 desktop PC because said computers are ranked on only one metric: their peak and average floating point capability.
  • Having it there is confusing and misleading to the average reader, and is realized as worthless by the more advanced reader. While the advanced reader will give it little regard, the average reader might take it to be an important benchmark and then we are back at my previous point.
Your input is much appreciated! -- uberpenguin 17:11, 2005 May 17 (UTC)
See the section directly below this one for an example of how this measurement can confuse the lay man. -- uberpenguin 17:48, 2005 May 17 (UTC)

Regardless of how the number was arrived at, it is part of the specification and, because Sony felt it important enough to spend their precious time putting a screen up for it, it should be mentioned here. It should be treated like any other public announcement: as fact until the company officially changes it's mind.

Overall System Floating Point Performance

I noticed that this has been temporarily changed, but it doesn't seem feasible that the PlayStation 3 would have more TFLOP performance than the Xbox 360. The Cell Processor is only a third of Xbox 360's custom IBM, which runs at 9.6GHz (3.2GHz per core, at three symmetrical cores), while the Cell runs at 3.2GHz.

Once again, I noticed that this was fixed, but unless I missed something, Xbox 360's page doesn't have an overall performance section, so it is unfair to compair when no specs are given.

I've already written a good bit about this, read above. Suffice to say that your thinking is quite a bit flawed. Three symmetrical cores running at 3.2 GHz are not equivalent to a single 9.6 GHz processor. There is a very large software factor here; read the SMP article. Furthermore, the number Sony provides is probably not solely due to the vector processing capabilities of Cell (to which one MIGHT be able to compare to the XBox 360 AltiVec-bearing CPU cores), but also to the highly SIMD graphics processor as well. You are correct that these sort of comparisons are unfair, because estimated floating point measurement is largely a creative guess, and because FP doesn't indicate all that much about the final performance in any given application. Really, who buys a game console based on how many shader pipelines the graphics processor has or what the theoretical peak memory bandwidth is? A lot of these specs are irrelevant to the average reader, and should be condensed some. Right now I'm trying to clean this article up a bit, but there are so many edits and vandalism on it currently that it's hard to get much done.... Bit by bit... -- uberpenguin 17:40, 2005 May 17 (UTC)
Also realize that Cell is a very different design from a triple-core PowerPC 970. The model used in the PlayStation 3 has seven used vector processing units that are very agressively timed. In other words, Cell is designed to be a very fast vector processor, which partially accounts for its expected high floating point performance. -- uberpenguin 17:45, 2005 May 17 (UTC)
On the Xbox 360 page, it says that the IBM custom processor has three symmetrical cores, with each core running at 3.2 gigahertz. 3.2 x 3 = 9.6. Therefore, it seems logical that 9.6GHz is the final processing speed. I could be wrong, but that's my two cents. Anyway, I'm sticking to helping out with the Xbox 360 page.
You're wrong, sorry. The three processors are independent of each other, they are just manufactured on the same IC. Threading apps rarely if ever creates a perfectly additive performance boost, and since all three processors are in contention for system resources (mostly memory access), real-world performance is degraded even further. -- uberpenguin 18:03, 2005 May 17 (UTC)
Okay. Honestly, I don't know much about PCs and technology, I only know the basics. So, if anything, I'll thank you for basically teaching me something new. -- Apadilla80
Hey, no problem, I work with this sort of stuff so I'm expected to know it. Not everyone in the world is an electrical engineer! -- uberpenguin 01:13, 2005 May 18 (UTC)
Ok, it's good now. --Mateusc 11:42, 18 May 2005 (UTC)
Great, I'm glad we got that resolved. On to bigger and better things! -- uberpenguin 13:12, 2005 May 18 (UTC)
Ok, but is important after all keep this information here, because Microsoft and Sony had divulged in its fact-sheet's and conferences (respectively). --Mateusc 00:11, 19 May 2005 (UTC)
Hey Uberpenguin obviously your electrical sockets from your job have got to you because your addition to the floating point capability is wayyyy to wordy. Lighten it up buzz.Mattrb
Not my choice; my first preference would be to remove it alltogether. Read down a bit for a very long discussion with Mateusc that explains why I felt the need to add all that. Mateusc insists on keeping the marketing info, so the compromise is to add an unnecessarily wordy blurb justifying its existence. Somewhat silly, but I don't mind all that much since there is very little to be said about this console at this point in time. -- uberpenguin 03:34, 2005 May 20 (UTC)

There are too many excuses in the "Overall floating-point capability" section. Obviously the specs are "theoretical" under "100% efficient operation". All specs are. It is redundant to note this not ounce but four times. And it reads like "fanboy" drivel in an attempt to discredit the hardware. I am removing the redundancy.

Also someone is incorrectly interpreting IBM's Linpack benchmarks and stating that "In real-world applications, IBM expects that the Cell will achieve about 75.9% of its maximum performance (or roughly 165 GFLOPS)". The benchmarks yield an efficiency range of 1.02% to 91.8% depending on the conditions (e.g. matrix size, numbers of SPUs tested, double/single precision etc). It is misleading to arbitrarily choose one metric (75.9). I am deleting this also. Xkxdxmx 16:11, 15 January 2006 (UTC)

You need to re-read those benchmarks. The 1% figure was for completely unoptimized code. Once you SIMDify it, efficiency jumps up to about 50%. Various other tricks yield the 75.9% figure, which is fairly consistent. I am reverting your sole edit until you can prove how the 75.9% figure is demonstrably wrong, or that you are not a sock puppet.the1physicist 18:26, 15

January 2006 (UTC)

Only *one* test condition yields the 75.9% efficiency metric: 4kx4k matrix parallelized across 8 SPUs. That hardly qualifies as "fairly consistent". And this is only an *efficiency metric*, it is very liberal (to be kind) to apply this metric to the single-precision GLOPS metric which is very clearly stated in Table 2 to be ~25 GFLOPS (SPEsim/Hardware) per SPU. I have removed this inaccuracy ounce again. Xkxdxmx 02:07, 16 January 2006 (UTC)