By using this site, you agree to our Privacy Policy and our Terms of Use. Close

Forums - Sony - Frame Rate Issues and Newer Versions of the PS3

Viper1 said:
6 Ghz? Oh no, are they pulling an IGN and adding clock speeds together or something?

No, actually IBM had the 90nm Cell clocking 6.7GHz for all 8 SPES and the PPE. Main problem was it gave off a temperature of over 190 degrees. At 65nm that heat should be reduced greatly, maybe at that speed 150 degrees. But I think Sony does have the ability to increase the speed of the cell a bit just they did with the PSP. Maybe they will increase the speed to 3.4GHz for all the SPEs and the PPE but they would never increase it to it's max potential because the console doesnt have a flash freeze system to keep it from melting. Even if the Cell is shrunk to 11nm (theoretically the smallest dyes will get to) the cell would give over 90+ degrees. Which an 11nm cell wont happen for another at least 8 years.

Cell dye size shrink (theoretical):
90nm -> 65nm -> 45nm -> 32nm -> 22nm -> 16nm -> 11nm
2007 -> 2008 -> 2009 -> 2011 -> Cell v2 -> Cell v2 -> Cell v2

I dont think that the cell will shrink to below 32nm but the next generation cell (Cell v2) should. Cell v2 = PS4.

 

like you said, past dye shrinking they wont change anything but maybe up the clock speed through the bios. A 200 MHz increase will only increase heat by 5 degrees maybe 8 but nothing that would hurt the system new or old, 90nm or 65nm. 



PC gaming is better than console gaming. Always.     We are Anonymous, We are Legion    Kick-ass interview   Great Flash Series Here    Anime Ratings     Make and Play Please
Amazing discussion about being wrong
Official VGChartz Folding@Home Team #109453
 
Around the Network
xbebop said:
smallflyingtaco said:
xbebop said:
 

In other news Sony is now shipping a liquid nitrogen cooling solution with the PS3 to accomodate the 6ghz clock speed.

Please, learn something about PC hardware before you talk.


I will ignore the attitude and respond politely. That number, assuming it is from IBM, is probably a theoretical maximum of improved frequency from lower voltage and shorter distances. If that is the maximum, yes initial improvements will be more moderate. Yeah, if you assume that the future will never exceed a 2004 Prescott your going to need liquid NO2 to hit 6 Ghz but that assumption does not seem to be wise as it has already been exceed with liquid NO2, I think the limit is a tad above 8 Ghz currently. The more interesting question, which one opinion has been kindly provided, is if this will be available in future PS3 or if it will be caped either in software or hardware so that new versions do not take advantage of it. If your an EE great, keep it civil.


The thing is, the 45nm Penryn C2D chips won't even get anywhere near 6ghz. At least not without melting shortly after.

Different architectures are not going to necessarily have the same limits.  Again, if this is a theorectical limit it may not matter as 6 Ghz may be only achievable in a lab circumstance but regardless they will get faster processors as time goes on and it will present the same circumstances.

 

 

 



Proud member of the Sonic Support Squad

El Duderino said:
^^ Thats ok, I´m no tech expert, just seemed very funny, I was always under the impression that frame rate issues where a problem cause by a game and the way it was programmed, not the console itself....

 Not sure, I have never written a game or even anything that is more graphically intense than making sure my numbers end up in the right file.  I was hoping someone who knew something might have an idea, as I can easily see this being more related to some other part of the system. 



Proud member of the Sonic Support Squad

It All has to do with the devs, F5 screwed up while Ea just plain old sucks.



Hus said:
It All has to do with the devs, F5 screwed up while Ea just plain old sucks.

Pretty much. Factor 5 tried to push it too hard. 1080p isn't needed to make Lair gorgeous. If they had done 720p, they might've been able to get a stable 60fps. 720p is the sweet spot this gen, due to processor speeds and memory bottlenecks.

I sincerely doubt they will change the speed of the cell processor. It doesn't make any sense. Tell the devs to learn how to take advantage of the SPUs, then the frame rate issues will disappear. 



Around the Network
ssj12 said:
Viper1 said:
6 Ghz? Oh no, are they pulling an IGN and adding clock speeds together or something?

No, actually IBM had the 90nm Cell clocking 6.7GHz for all 8 SPES and the PPE. Main problem was it gave off a temperature of over 190 degrees. At 65nm that heat should be reduced greatly, maybe at that speed 150 degrees. But I think Sony does have the ability to increase the speed of the cell a bit just they did with the PSP. Maybe they will increase the speed to 3.4GHz for all the SPEs and the PPE but they would never increase it to it's max potential because the console doesnt have a flash freeze system to keep it from melting. Even if the Cell is shrunk to 11nm (theoretically the smallest dyes will get to) the cell would give over 90+ degrees. Which an 11nm cell wont happen for another at least 8 years.

Cell dye size shrink (theoretical):
90nm -> 65nm -> 45nm -> 32nm -> 22nm -> 16nm -> 11nm
2007 -> 2008 -> 2009 -> 2011 -> Cell v2 -> Cell v2 -> Cell v2

I dont think that the cell will shrink to below 32nm but the next generation cell (Cell v2) should. Cell v2 = PS4.

 

like you said, past dye shrinking they wont change anything but maybe up the clock speed through the bios. A 200 MHz increase will only increase heat by 5 degrees maybe 8 but nothing that would hurt the system new or old, 90nm or 65nm.


Sso they were just tossing theoretics around?  As if we don't get those numbers enough already from Glfops and polygons, etc...

 

The PSP chip was already capable of 333mhz it was just locked from development due to battery drain issues mostly. 



The rEVOLution is not being televised

Actually the very first iterations of Cell processor were designed to run at 4.6 GHz. The PS3 version has been down clocked for reliability, power consumption, size of the enclosure, cooling and other factors. It very much was designed to go faster, and smaller processes should allow transistors and therefore cycles to operate faster. Remember, Cell wasn't designed specifically for PS3. The colaboration was intended to produce a new kind of chip that combined the advantages of parallel massive floating point calculation throughput of a GPU with the logic flexibility of a CPU. Soon enough useful applications will make use of the type of power which only this kind of architecture can supply. It's just that Folding@Home is currently the only one.

PS. The Pentium 4 processor line was originally designed to scale up to 10 GHz before Intel screwed it up.



Just wanted to point out that you can technically clock any chip to 6 Ghz ....its just a matter of if the thing will explode violently before it gets there or not =P.

But there are currently cooling techniques in development that in theory should let us get to 10Ghz safely....but thats *in theory*. It would mean a lot more if someone could do it in practice =P



To Each Man, Responsibility

you could try to dip the ps3 into motor oil, like that case mod that used cooking oil for the PC on tomshardware article.



Phase change all the way :)