Was Cell any good? *spawn

Status
Not open for further replies.
I guess the argument to that is whether Cell+Xenos would have been worth the cost. If PS3 had the same GPU as XB360, would the investment in Cell have delivered worthwhile benefits, or would it have been a waste of money where a simpler CPU would have done the job nicely?

We were seven years into the generation before I saw a game where PS3 outran Xbox on the CPU, and even then PS3 is still severely crippled on the GPU end and there's always room for CPU optimization. Cell was interesting tech and thus was fun to work with, but in no way was it ever a good investment.
 
We were seven years into the generation before I saw a game where PS3 outran Xbox on the CPU, and even then PS3 is still severely crippled on the GPU end and there's always room for CPU optimization. Cell was interesting tech and thus was fun to work with, but in no way was it ever a good investment.

Nah, that's nonsense. For instance ever since Havok 4.5, Havok runs quite a bit better on PS3. You could tell occasionally that the PS3 performed better in there in multi-platform titles, but generally mp devs have kept physics relatively simple in order to cater for the weakest platform (as with everything, obviously also in PS3's favor plenty of times). A game like Motorstorm went pretty wild with the Havok stuff though, and at the time of that games release, Havok ran better on Cell than almost any other CPU other there (that would run games that have Havok).

It may not have been a great investment, but if you would replace it with something like the 360's CPU and leave everything else the same, I think the PS3 would have been much worse off.
 
Yes but imagine if Microsoft would'be been out of luck with their GPU (i.e. ATI not being ready with their unified shader architecture or whatever) and then both, PS3 and 360, having a very similar GPU? Then Sonys advantage in terms of the CPU would've been the factor that could've completely changed the playing field.

Or the other way around, Nvidia having a faster GPU in the back of the store... it could've easily gone either way for Sony or MS.

The sooner you realize that sony didn't have an advantage in cpu, the more clarity you will have.
 
The main problem that Sony had with the CPU, is programmability. In the beginning of the gen, many teams stuck with just the 1 core to avoid the headache of multi-threading. As the generation moved on and programmers got more comfortable with multiple cores, first code like Audio would move to the second core and then slowly more and more stuff got properly multi-threaded. On the 360 however, this was easier to do - while multi-threading still isn't easy, big advantage was just being able to use the exact same code over the three cores, while the powerful VMX unit on the 360 CPU still allowed for some decent performance optimisation.

Comparing this to Cell development, it is much harder to use the additional cores there. You couldn't reuse your code nearly as easily, and to really make use of them well, you'd need different algorithms and even data-flow. As was discussed many times, these changes are usually changes for the better even on 360, but that doesn't hide the fact that it is much harder.

And I think when everything is said and done, that remains the Cell's biggest problem: very powerful, but too difficult.

EDIT: see this post just made by sebbbi:

http://forum.beyond3d.com/showpost.php?p=1658660&postcount=13730
 
The sooner you realize that sony didn't have an advantage in cpu, the more clarity you will have.

It will take a few people a long time to realize that given the reality of the situation. It's understandable you detest the architecture of CELL and all the shortcomings that go with it. But without CELL the PS3 would not have kept parity with the 360. Even you should be able to acknowledge that. Now there are workloads where 3 PPE cores are better than one PPE and 6 SPE's and the reverse is also true. But CELL has allowed PS3 to keep up with 360 where it otherwise would be left to twiddle its thumbs with a not as capable CPU.

Now as for CELL + Xenos being good combination. I do believe it would be best of both worlds if the machine had UMA.
 
The sooner you realize that sony didn't have an advantage in cpu, the more clarity you will have.
The old Cell does not get much respect, but it does have its purposes.

I am guessing here but when developers have to commit to program..., they try to find ways to use its advantages.

However being difficult to program means that many developers simply don't have a budget for third-party engines built around Cell and have to make do with what they've got. :rolleyes:

I don't think Cell is a good CPU for "poor" developers at all, which kinda sucks.

Necessity is the mother of invention they say; with a little imagination you can do far more with the Cell than you might think possible. I guess....

I've often thought it'd be fun to hold a developers contest to see just how far creative developers can go with a CPU such as the Cell.
 
So if XB360 had an RSX instead of Xenos, it'd still perform as well as PS3 with Cell and RSX?

Yes. The cell is the programming nightmare because its whole entire programming model is broken which means that for any given amount of optimization work, xenos is going to run faster.
 
Of course they did..

What are you talking about?

no they didn't. The vast majority of PS3 games had 1 single CPU core. To get any use out of the SPUs required a vast amount of resources and optimization, that if spent on Xenos would result in similar speedups, but devs didn't need to spend those resources with xenos because even a simple implementation of a mutli-threaded program ran well enough.

So Xenos did everything the devs needed without much if any additional work. Cell required large amounts of additional work and optimization just to reach an acceptable level of performance.
 
It will take a few people a long time to realize that given the reality of the situation. It's understandable you detest the architecture of CELL and all the shortcomings that go with it. But without CELL the PS3 would not have kept parity with the 360. Even you should be able to acknowledge that. Now there are workloads where 3 PPE cores are better than one PPE and 6 SPE's and the reverse is also true. But CELL has allowed PS3 to keep up with 360 where it otherwise would be left to twiddle its thumbs with a not as capable CPU.

Without CELL both consoles would of had better games because every game would require significantly less dev resources. CELL merely sucked up additional developer resources while given no real advantage.

Now as for CELL + Xenos being good combination. I do believe it would be best of both worlds if the machine had UMA.

Both dropping SPUs and having just PPUs would have been better. Sony could of gone 4x PPUs with roughly the same cost.
 
T
I don't think Cell is a good CPU for "poor" developers at all, which kinda sucks.

Cell isn't a good CPU for good or excellent developers. They would be better off with 3-4 PPUs than 1 PPU+6 SPUs.

I've often thought it'd be fun to hold a developers contest to see just how far creative developers can go with a CPU such as the Cell.

It would be pointless.
 
Without CELL both consoles would of had better games because every game would require significantly less dev resources. CELL merely sucked up additional developer resources while given no real advantage.



Both dropping SPUs and having just PPUs would have been better. Sony could of gone 4x PPUs with roughly the same cost.



Did CELL single handedly make development this generation as expensive as it has become? Is that the cause of it? Did it suck up so much resources that nobody is making use of the SPE's because they're just so darn expensive to program for?

And here we have this same debate. Would a 4 core PPU PS3 with RSX be able to keep up with the 360? Or is Xenos just in a league ofits own compared to RSX where one additional PPU makes little difference?
 
Did CELL single handedly make development this generation as expensive as it has become? Is that the cause of it? Did it suck up so much resources that nobody is making use of the SPE's because they're just so darn expensive to program for?

And here we have this same debate. Would a 4 core PPU PS3 with RSX be able to keep up with the 360? Or is Xenos just in a league ofits own compared to RSX where one additional PPU makes little difference?
Sorry to interfere in your discusion but I feel like it's not a fair question. Cell was the main focus of the design without the cell the ps3 would have been completely different.
The cell came with requirement of its own like high bandwidth xdram. For example (keeping mostly the same design) for the same price as the xdram I suspect that Sony could have used 512MB of ddr2(may be more?).

NB I wrote a page ago that I believe that Nintendo should have come with a cell-like as it seems there were constrained both in power consumption and silicon budget especially as thanks to the ps3 it's a known quantity now.
 
Last edited by a moderator:
Yes. The cell is the programming nightmare because its whole entire programming model is broken which means that for any given amount of optimization work, xenos is going to run faster.
:???: Sorry, I'm still a bit confused by this, because I wasn't asking about Xenos.

Putting it another way, you are saying that Cell doesn't outperform Xenon, XB360's CPU, in any way, and Xenon vs. Cell will be an even match in most gaming scenarios?
 
no they didn't. The vast majority of PS3 games had 1 single CPU core. To get any use out of the SPUs required a vast amount of resources and optimization, that if spent on Xenos would result in similar speedups, but devs didn't need to spend those resources with xenos because even a simple implementation of a mutli-threaded program ran well enough.

So Xenos did everything the devs needed without much if any additional work. Cell required large amounts of additional work and optimization just to reach an acceptable level of performance.

For some reason I feel that you are still stuck in 2007
 
no they didn't. The vast majority of PS3 games had 1 single CPU core.

No they didn't... They "had" 1 dual-issue PPU & 6 SPUs...

To get any use out of the SPUs required a vast amount of resources and optimization, that if spent on Xenos would result in similar speedups, but devs didn't need to spend those resources with xenos because even a simple implementation of a mutli-threaded program ran well enough.

Rubbish...

There's no way in hell you'd get similar speed ups with Xenos if you spent "vast" amounts of resources on optimizing CPU code...

The problem was that most multi-platform developers didn't "have" vast amounts of resources to burn on PS3 dev but even without that you could still do ALOT with even basic use of the SPUs. I know because our studio were in this exact same position.

So Xenos did everything the devs needed without much if any additional work. Cell required large amounts of additional work and optimization just to reach an acceptable level of performance.

Xenos did what developers "needed" but nothing substantially more. If you wanted to use Xenos to make God of War III or Uncharted II you couldn't because it just doesn't have the raw vector processing power nor the available parallelism to be up to the job...

Sony provides lots of libraries to allow developers to make the most of the Cell without breaking too much of a sweat but it took large amounts of work for most developers coming from the PC-space/PS2 era with lots of legacy code to get the PS3 paradigm to "fit" in with what they were doing.

If you were like us however & came at PS3 development from a clean slate then it didn't require massive amounts of work to get the SPUs working for you, anymore than it would have doing it in compute shaders for example. It's just about thinking about the problem & approaching it in a different way, that clearly paid off for most devs who invested in PS3 friendly tech (hence the reason why everyone & their dog is using job-based data-parallel execution models in their engines these days, PS3 or no..)
 
Ain't that costed Sony a lot of work? (all the library?)
Sorry guys but Aaron (and not only him by the way)has won the point. Plenty of not only software but hardware people has vouched the Cell a failure or more a dead end at launch even before release.
They've been proved right, it was a wrong solution for the problem they wanted to tackle. It's dead now along with this approach.

And looking only at the picture through the prism of the ps3, say if for example the ps3 would have been a 4 cores Xenos (mostly the same silicon budget as the cell) and the RSX, Sony could have afford more memory. I believe that twice the main ram would have made more of a difference (visible than more than a handful of games) than whatever the SPU brang on the table.
 
Last edited by a moderator:
Status
Not open for further replies.
Back
Top