PPE = Xenon CPU core confirmed?

[JF_Aidan_Pryde]

Has it been more or less settled that the PPE in Cell is the same as the cores used in the Xenon CPU?

 

[The GameMaster]

No... there is very little details about the 3 custom G5 CPUs that the XBox360 is using. But I will tell you that they are NOT the PPE that was being using in the Cell CPU design. I would suspect Microsoft will unviel that eventually...

The GameMaster...

 

[hugo]

Has it been more or less settled that the PPE in Cell is the same as the cores used in the Xenon CPU?

It definitely isn't the same.The full Cell 9-core processor can be overclocked above 4Ghz while the XCPU has already reached quite its peak.

It would be a blow if Sony would up the clock frequency of the current Cell processor spec that was presented at E3 recently.I am sure it can be done at the last minute.The 3.2Ghz clock rate that they stated could be part of their plans not to reveal everything all at once.

 

[Acert93]

It would be a blow if Sony would up the clock frequency of the current Cell processor spec that was presented at E3 recently.I am sure it can be done at the last minute.The 3.2Ghz clock rate that they stated could be part of their plans not to reveal everything all at once.

Or like the 7 SPE design it could be a testiment to the fact they want good yields. Look, they already outperform their competitor in FP performance by almost 2:1, I would think at this point limiting heat, controlling power consumption, and improving yields would be more important.

 

[ERP]

Has it been more or less settled that the PPE in Cell is the same as the cores used in the Xenon CPU?

I will say this, I used to believe this was true(minus the VMX changes of course), but these days I'm not so sure.

 

[hugo]

Or like the 7 SPE design it could be a testiment to the fact they want good yields. Look, they already outperform their competitor in FP performance by almost 2:1, I would think at this point limiting heat, controlling power consumption, and improving yields would be more important.

As you've mentioned they've probably reduced the number of SPEs from 8 to 7 was because of better yield.The possibility of increasing the final frequency of the PS3 Cell is even more likely with this.There is still space for 200-300mhz for the Cell to go IMO.

As a comparison the prototype STI Cell could hit 4.6GHz without problems.I would think the current 3.2Ghz clock frequency is just a little low compared to that.

 

[Farid]

I will say this, I used to believe this was true(minus the VMX changes of course), but these days I'm not so sure.

Here's the answer of the thread.

 

[nAo]

ERP: those issues you was referring to in another thread..could be due to different compilers technology?

 

[ERP]

I don't believe it accounts for the size of the difference.


But you can never tell, I recently witnessed a piece of code running 10x faster in debug mode than in release, because a compiler tried to reorder some writes to write combined memory.

 

[V3]

Has it been more or less settled that the PPE in Cell is the same as the cores used in the Xenon CPU?

From what I heard from one IBM guy, it isn't the same. Not even close according to him.

 

[blakjedi]

I am sure it can be done at the last minute.The 3.2Ghz clock rate that they stated could be part of their plans not to reveal everything all at once.

I wouldn't be so sure about that. I think that the fabrication process is the same for both chips... for now at least whatever limitations affect the PPC core in 360 also affect (and even worse) in Cell.

After touting 4.6 Ghz theoretically for Cell publicly clocking it at the SAME frequency as X360 core is telling.

 

[passby]

This just reminded me that PPE details have not been disclosed at all. As a reminder by realworldtech:

Neither microarchitectural details nor the performance characteristics of the POWERPC Processing Element were disclosed by IBM during ISSCC 2005. However, what is known is that the PPE processor core is a new core that is fully compliant with the POWERPC instruction set, the VMX instruction set extension inclusive. Additionally, the PPE core is described as a two issue, in-order, 64 bit processor that supports 2 way SMT.

Funny that we actually know more about a SPE than a PPE. Both PPE and X-CPU are still work in progress?

 

[3dcgi]

I'm surprised Sony even announced clock speeds at this point, but I consider them to be targets that could go up or down. Until they make a significant number of chips it's difficult to know what speeds will result in the best bang for the buck regarding yields and performance.

 

[Shifty Geezer]

I am sure it can be done at the last minute.The 3.2Ghz clock rate that they stated could be part of their plans not to reveal everything all at once.

I wouldn't be so sure about that. I think that the fabrication process is the same for both chips... for now at least whatever limitations affect the PPC core in 360 also affect (and even worse) in Cell.

After touting 4.6 Ghz theoretically for Cell publicly clocking it at the SAME frequency as X360 core is telling.

Wasn't there rumour a while back that XB360's cores were being clocked lower than expected because of yield problems? That rumour seems bunk (the rest of it) but fabbing issues makes sense, as to why both XB360 and PS3 are clokced at c. 3 GHz, plus IBM isn't reknowned for fast CPUs.

 

[PC-Engine]

I am sure it can be done at the last minute.The 3.2Ghz clock rate that they stated could be part of their plans not to reveal everything all at once.

I wouldn't be so sure about that. I think that the fabrication process is the same for both chips... for now at least whatever limitations affect the PPC core in 360 also affect (and even worse) in Cell.

After touting 4.6 Ghz theoretically for Cell publicly clocking it at the SAME frequency as X360 core is telling.

I also found it funny that they were claiming 4.6GHz speeds in the lab to one up Intel but then had to packpedal with a mediocre 3.2GHz chip.

 

[London Geezer]

I know!! The PPE is actually a..... PPU!

 

[Mordecaii]

It's hardly mediocre, the fact is that it gives them plenty of performance at 3.2 GHz, which will give them better yields and less heat, which will save them money. Do you have anything constructive to say at all PC-Engine, or do you just enjoy trolling?

 

[PC-Engine]

It's hardly mediocre, the fact is that it gives them plenty of performance at 3.2 GHz, which will give them better yields and less heat, which will save them money. Do you have anything constructive to say at all PC-Engine, or do you just enjoy trolling?

Actually I'm kinda wonderin why you've suddenly stopped your KZ is realtime crusade.

 

[rabidrabbit]

I also found it funny that they were claiming 4.6GHz speeds in the lab to one up Intel but then had to packpedal with a mediocre 3.2GHz chip.

In the first unveiling of the "Cell" it wasn't said that "Cell" is the one that's going to be on PS3, which of course has many other things in it that affect the "Cell" clockspeed.
Who knows, didn't they up the PS2 clockspeeds, maybe the 3.2MHz will be upped too.
The PS3 shown at E3 was a prototype after all.

 

[London Geezer]

I also found it funny that they were claiming 4.6GHz speeds in the lab to one up Intel but then had to packpedal with a mediocre 3.2GHz chip.

So now 3.2GHz for a home console is mediocre? Because that's the number Sony gave it? Wow... You never cease to amaze me... I guess anything else clocked at around 3.2GHz is mediocre for you then, including the X360 and most PC CPUs out now.