When are the 65nm PS3's and XB360's shipping?

Status
Not open for further replies.
But things are very very...very complicated.
Heh. Thanks for the exposition. :smile: Not too hard to follow, even for one such as myself (the last time I actively tried to learn about these things proper, Intel had just launched the Pentium).

I guess I was thinking along the lines of: "Given what we know about the ICs current power profiles, node reductions in general (assuming that any redesign will be limited to fixing unexpected problems with static power), and the process they're using in particular; they're likely to reduce P by n to n'."

But I guess it's just too many variables to consider these days for such 'blanket estimations' to make much sense.
 
For what its worth, there was a report from IBM IIRC about a 65nm Cell, and it's thermal output had dropped hardly any amount. There was discussion in the tech forum. Try a search.
 
Heh. Thanks for the exposition. :smile: Not too hard to follow, even for one such as myself (the last time I actively tried to learn about these things proper, Intel had just launched the Pentium).

I guess I was thinking along the lines of: "Given what we know about the ICs current power profiles, node reductions in general (assuming that any redesign will be limited to fixing unexpected problems with static power), and the process they're using in particular; they're likely to reduce P by n to n'."

But I guess it's just too many variables to consider these days for such 'blanket estimations' to make much sense.

I had fun trying to remember that, and that was from a course I took only just a couple years back. :)

I guess a good example is that history slide on EE+ GS node reductions that Carl posted in another thread; some revisions didn't immediately reduce the chip area by any significant degree when it was expected, but then they later caught up appropriately. The overall theme being that process node reductions don't guarantee anything, and it gets more difficult at sub-90nm with the leakage.

With that voltage and frequency dependence, it only starts to get tricky in doing a full node reduction versus a half-node with fixed specs. :(

On the PC side, we might look at Conroe to Wolfdale, but they made a bunch of changes like FSB and cache size amongst other things that mitigated the absolute power reduction. And being OOE, it's probably not a great example to begin with. :|

For what its worth, there was a report from IBM IIRC about a 65nm Cell, and it's thermal output had dropped hardly any amount. There was discussion in the tech forum. Try a search.


Oh... I must have missed that. I'll try to find that. Thanks Shifty. :)

mmm...Google. Is this it?
http://researchweb.watson.ibm.com/journal/rd/515/riley.html


edit 2:

Oh... neat. A bunch of papers recently published online regarding Cell (meant for Sept. issue). http://researchweb.watson.ibm.com/journal/rd/515/tocpdf.html


The SPE design has roughly 20.9 million transistors, and the chip area including the SMF is 14.8 mm2 (2.54 mm x 35.81 mm) fabricated with a 90-nm silicon-on-insulator (SOI) technology. The 65-nm version of the design is 10.5 mm2.
10.5/14.8 = ~71%



edit 3:
The 65-nm design contained roughly 239M devices.

...versus 241M in 90nm.
 
I think some are more sensitive to the sound of their 360 or PS3 than others. If you have a theater room with a quiet sound floor, than the noise coming from the PS3 and the 360 may be quite noticeable and is substantially higher than dedicated dvd players.
 
Someone over at xbox-scene opened up the Halo 3 Edition console, but didn't take off the heatsinks. It could have been Zephyr, but there was no sign of the name on the box as it is normally shown on Pro/Elite boxes. Curious.

The board shots do look identical to the Zephyr board though. The console was manufactured 2007-08-07. I assume that means August 7...

http://forums.xbox-scene.com/index.php?showtopic=621541
 
Someone over at xbox-scene opened up the Halo 3 Edition console, but didn't take off the heatsinks. It could have been Zephyr, but there was no sign of the name on the box as it is normally shown on Pro/Elite boxes. Curious.

The board shots do look identical to the Zephyr board though. The console was manufactured 2007-08-07. I assume that means August 7...

http://forums.xbox-scene.com/index.php?showtopic=621541

yep, doesn't look like anything new. What are they doing with these 65nm chips that have been in prod for months now?
 
Someone over at xbox-scene opened up the Halo 3 Edition console, but didn't take off the heatsinks. It could have been Zephyr, but there was no sign of the name on the box as it is normally shown on Pro/Elite boxes. Curious.

The board shots do look identical to the Zephyr board though. The console was manufactured 2007-08-07. I assume that means August 7...

http://forums.xbox-scene.com/index.php?showtopic=621541
And the 65nm PS3, anything new?
What are they doing!
 
i'm not buying another 360 (my first one had a LOUD dvd-rom and scratched my ONLY halo 2 disc to shit (the ONLY game i played) until ALL the important chips (cpu, gpu, edram) go 65nm. right now, we're STILL looking at 90nm all the way. pretty soon (with falcon), we'll be looking at ONLY a 65nm cpu (with the edram manufacturer switching from nec to tsmc (with the same 90nm fab).

(i don't know if this has been posted already, but just in case: http://blogs.mercurynews.com/aei/20...mean_for_when_you_should_buy_an_xbox_360.html)
 
Status
Not open for further replies.
Back
Top