• Hey Guest. Check out your NeoGAF Wrapped 2025 results here!

Cell Processor to Run at 4.6 Ghz

When Sony reveals the technical details of the highly-anticipated advanced microprocessor code-named Cell, they will announce that the new processor will run at 4.6 gigahertz. IBM, Sony and Toshiba will present four technical papers at the International Solid State Circuits Conference to be held next month in San Francisco.

The Cell is a multicore chip comprising a 64-bit Power processor core and multiple synergistic processor cores capable of massive floating point processing. Cell is optimized for compute-intensive workloads and broadband rich media applications, including computer entertainment, movies and other forms of digital content.

Sony plans to use the Cell to power its next generation PlayStation as well as home servers for broadband content and high-definition televisions.


Other highlights of the Cell processor design include:
Multi-thread, multicore architecture.
Supports multiple operating systems at the same time.
Substantial bus bandwidth to/from main memory, as well as companion chips.
Flexible on-chip I/O (input/output) interface.
Real-time resource management system for real-time applications.
On-chip hardware in support of security system for intellectual property protection.
Implemented in 90 nanometer (nm) silicon-on-insulator (SOI) technology.

http://news.teamxbox.com/xbox/7511/Cell-Processor-to-Run-at-46-Ghz/

Have at it, techies.
 
I find that hard to believe. Neither IBM, nor Intel are able to hit evn 4.0 gigahertz, nevermind more than 4.5.

:lol
 
Culex said:
I find that hard to believe. Neither IBM, nor Intel are able to hit evn 4.0 gigahertz, nevermind more than 4.5.

:lol
uhh...wasn't that one of the the reasons for designing it?




anyway...does this mean it's been downgraded from the previous 4.8 estimate?

edit: also, it seems like IBM has found a way :b
 
The 4.6Ghz figure came from an ISSCC preview slide months ago - there was never a 4.8 number.

So this is kind of old news. I'm guessing it represents the max speed they've got a Cell chip running stabily at - probably the speed of the chip they'll be discussing and/or demonstrating. It doesn't say much about the speed of the chip that'll be in PS3, except what might be possible (although precedent suggests that it should be close to what the final PS3 chip speed will be the PS2 CPU turned out to be the same chip Sony showed at the ISSCC in 99 or 2000, except with a speed bump).

I'm not sure why TeamXbox is running with this story just now..?

Also :lol :lol @ some of the reader comments for that story "4.6Ghz just sounds weak... " :lol
 
"Sony plans to use the Cell to power its next generation PlayStation as well as home servers for broadband content and high-definition televisions.
"

This home media server ap is the future for computers. I really feel bad for MS since they're crippled by legacy technology. They're gonna be in deep sh** If this cell is all that it's cracked up to be.
 
Keyser Soze said:
But they could by the time of release. Infact it is very likely they could break the 4Ghz barrier in a few months time.

Check Intel's 2005 roadmap. There aren't even plans yet to break 4 gigahertz. 2nd and 3rd quarter are dual-core, ~3 gigahertz chips.
 
ToxicAdam said:
This has me pumped to play Catwoman 2 now!!!


WOOOOOOOOOOOOOOOOOOOOOOOO!!!
2003%20-%20april13%20-%204th%20grade%20v.%20rockville%20centre%20-%20Adam%20and%20Peter%20celebrate.jpg
 
So Cell is going to break Moore's Law? Until I see something solid, this is the 'Emotion Engine' all over again.
 
lockii said:
So Cell is going to break Moore's Law? Until I see something solid, this is the 'Emotion Engine' all over again.

We shouldn't jump to conclusions, of course. Wait until the ISSCC at least.

I still am puzzled why TeamXbox chose to run this now. It's months old news. Perhaps they're trying to inflate expectations so people can crack the whip at Sony if and when the PS3 doesn't emerge as powerful as this kind of figure suggests :P
 
Why would it break Moore's law? When's Cell coming out? 2006? When was the first 2.4Ghz cpu out? I think it follows the rule of thumb.

Side note, this is about 1000 times faster than an NES.
 
4.6 Ghz is impressive, but in this day and age, haven't we learned by now that clock speed is not the end all that we once thought it was. The other components are just as important these days, and bandwidth between the processor, memory, video card, etc. are just as important.
 
The days of MHZ/GHZ defining a processor speed are long over. It's all about the efficency of the chip. Even Intel have said they are giving up the race for endless GHZ improvements.
 
gofreak said:
We shouldn't jump to conclusions, of course. Wait until the ISSCC at least.

I still am puzzled why TeamXbox chose to run this now. It's months old news. Perhaps they're trying to inflate expectations so people can crack the whip at Sony if and when the PS3 doesn't emerge as powerful as this kind of figure suggests :P
They be all about the unrelevant and months old news? :lol j/k

Seriously, it's an Xbox site, so only thing I can think of is to feed the fanboys?

Still cell is shaping up to be a monster of sorts. Maybe just it'll end up being a tiny one, but a monster nonetheless.
 
MattKeil said:
Sounds like Sony Math to me.

Care to explain? You might be able to magic up theoretical performance numbers, but speed ratings are readily measurable and verifiable. The ISSCC wouldn't be very receptive of the kind of trickery you are referring to.

And Cockle's is right - clock speed isn't as important as it used to be. It's all about what you do per clock cycle - and funnily enough, a cell chip should be doing an awful lot per cycle (which is where those massive FLOP numbers that have been speculated on come from). Efficiency etc. is also an issue..I guess we'll have to wait and see re. real world performance, but on paper Cell seems to be shaping up nicely.

Also, on one detail, re. memory bandwidth - it should be huge.
 
I think it was speculated that only the fastest parts of the Cell would run at 4.6 Ghz, and the cores would be only half or 1/4 of that.
 
SpoonyBard said:
I think it was speculated that only the fastest parts of the Cell would run at 4.6 Ghz, and the cores would be only half or 1/4 of that.

Yeah, that's a possibility, the APUs may be running at that speed, and the Power core at half that or less. But it doesn't really matter, as long as the APUs are kept firing on all cylinders...all estimates and calculations re. a Cell chips performance are based on the APUs alone, and not the Power core. The Power core's main job is to keep feeding the APUs (although I believe general code can also be run on it).
 
Sheeit if sony/IBM starts making these processors for desktops AMD/INTEL better release the floodgates. Then again it wont be out for a while so it wont be all that spectacular by the time the PS3 drops. 4.6ghz is nuts though i wonder what NM process the chip is fabbed on .65nm maybe.
 
DonasaurusRex said:
4.6ghz is nuts though i wonder what NM process the chip is fabbed on .65nm maybe.

The chip being talking about at the ISSCC was manufactured on a 90nm process. They might get down to 65nm for PS3, we'll have to wait and see.

Also, these chips are kind of biased towards certain types of app that require a lot of floating point performance (games, multimedia etc.), and benefit from paralellism. They might not be equally suited toward other applications, so their suitability in desktop machines might be questionable. But someone else could discuss that far more knowledgeably than me..
 
Duckhuntdog said:
IF this is true, the PS3 is clocked higher than the Xbox 2....

Well this isn't necessarily the speed of the PS3. Although it gives an indication of what might be possible.

Clockspeed also isn't everything. You can't necessarily say that if PS3 is clocked 1.5 times higher than Xbox2 (for argument's sake), that it is 1.5x as powerful. It depends on what you can do in each clock cycle (although Sony seem to have that base covered well too).

On a side note, I find it kind of funny that a lot of people seem to be perenially surprised at the possibility of PS3 being more powerful than Xbox2. All else being equal, if it's coming out after Xbox2 that simply makes sense.
 
Doom_Bringer said:
Why is Teamxbox hyping PS3? I don't think this number is real, lets wait for the official word.

It's a real number - it was on the ISSCC preview slides for Cell (these were released months ago, which is why I'm wondering why Teamxbox are just reporting it now. slow news day?). That doesn't mean it's the speed of the chip in PS3, though. It's just the speed they've got Cell chips running at, and presumably/possibly the speed of the chip they'll be discussing and/or demonstrating at the ISSCC.

Also consider that Cell speed may be variable across the chip. Some think the APUs are running at this speed and that the core is running at a slower rate.

edit:

Here's the original slide for reference:

011bl.jpg


Also, the 4.8Ghz figure was in there somewhere, actually...IIRC, that is the SRAM speed.
 
How much RAM it's got matters more I believe. Hopefully 512MB at least .... it's the main thing developers complain about late in its life cycle - there just isn't enough RAM to do things.
 
Hollywood said:
How much RAM it's got matters more I believe. Hopefully 512MB at least .... it's the main thing developers complain about late in its life cycle - there just isn't enough RAM to do things.


I bet its 256 MB for both Xenon and PS3!
 
gofreak said:
Well this isn't necessarily the speed of the PS3. Although it gives an indication of what might be possible.

Clockspeed also isn't everything. You can't necessarily say that if PS3 is clocked 1.5 times higher than Xbox2 (for argument's sake), that it is 1.5x as powerful. It depends on what you can do in each clock cycle (although Sony seem to have that base covered well too).

On a side note, I find it kind of funny that a lot of people seem to be perenially surprised at the possibility of PS3 being more powerful than Xbox2. All else being equal, if it's coming out after Xbox2 that simply makes sense.


You are assuming that the prospect of the PS3 being more powerful than an Xbox 2 surprises me. While it is granted that the Xbox 2 will have 1 PPC with 3 cores at 3.5 Ghz, which isn't too shabby, as time goes by I am more than inclined to believe the Xbox 2 will end up towards the bottom of the power rankings.

The xbox 2 isn't about power, being the most powerful, or any such concept. The xbox 2 is about one thing: finally making some freaking money off this hardware. If we don't, chunk that bitch, we got our exit plan ready to go anyway: XNA.
 
Duckhuntdog said:
You are assuming that the prospect of the PS3 being more powerful than an Xbox 2 surprises me. While it is granted that the Xbox 2 will have 1 PPC with 3 cores at 3.5 Ghz, which isn't too shabby, as time goes by I am more than inclined to believe the Xbox 2 will end up towards the bottom of the power rankings.

The xbox 2 isn't about power, being the most powerful, or any such concept. The xbox 2 is about one thing: finally making some freaking money off this hardware. If we don't, chunk that bitch, we got our exit plan ready to go anyway: XNA.


Sorry, I wasn't referring directly to you when I said that, although your post was reminscint of the type of post that express such surprise. I didn't mean you were, it was a slightly tangental comment.

Also, don't get me started on XNA :P
 
gofreak said:
It's a real number - it was on the ISSCC preview slides for Cell (these were released months ago, which is why I'm wondering why Teamxbox are just reporting it now. slow news day?). That doesn't mean it's the speed of the chip in PS3, though. It's just the speed they've got Cell chips running at, and presumably/possibly the speed of the chip they'll be discussing and/or demonstrating at the ISSCC.

Also consider that Cell speed may be variable across the chip. Some think the APUs are running at this speed and that the core is running at a slower rate.

edit:

Here's the original slide for reference:

011bl.jpg


Also, the 4.8Ghz figure was in there somewhere, actually...IIRC, that is the SRAM speed.

ahh...
 
Doom_Bringer said:
they will probably try to increase the speed. I wonder who will present their next gen plans at E3 first


should be very interesting :lol

Indeed, the potential for drama is probably higher than its ever been at E3, if all 3 manufacturers keep their cards close to their chests till then. I can only imagine what GAF will be like...it'd probably be the single biggest news day ever for the forum, so much controversy, so much debate and argument. It'll be insane.
 
Top Bottom