Predict: The Next Generation Console Tech

Status
Not open for further replies.
The tiniest part I saw supporting a 256 bit bus was ~230 sq.mm so it implies quiet a chip. for the reference it was the geforce gts 250 produced @55nm including 750millions transistors.

Not that it's super important, but AMD 38xx chips were 192mm2, had a devilish 666 transistors @55nm.
 
AlphaWolf said:
That's pretty far from confirmation of anything. In fact it's almost exactly speculation on his part.

But look at the reasoning.

That's nothing more than an assumption. And things have been pointing towards AMD having all three GPUs.

EDIT: Beaten

You do know that Jen-Hsun Huang is the CEO of Nvidia and as such is accountable to the stockholders for the information he is sharing?

If he says they are building one of the next gen consoles, I am inclined to believe there is some truth behind it.

Who are the sources behind the idea that AMD got the GPU contracts of all the next gen consoles?
 
You do know that Jen-Hsun Huang is the CEO of Nvidia and as such is accountable to the stockholders for the information he is sharing?

If he says they are building one of the next gen consoles, I am inclined to believe there is some truth behind it.

Who are the sources behind the idea that AMD got the GPU contracts of all the next gen consoles?

Rumors from some E3 or CES or something were the originals claiming AMD holds contracts for all 3.

The way Huang said what he said pretty much confirms AMD is making all of them if nV had the deal in their pocket even for one, he would have made it clear, and not just state that "one company isn't capable of building all 3" as the only reasoning why nV would make gfx for one of them.
 
I have spent considerable time googling information about 3D and 2.5D IC stacking, I could spam you with links on the subject.
2.5 D stacking is happening right now and TSMC is a driving force, they are preparing manufacturing and assembly plants to start offering such services on a larger scale in 2013. They already have it in production for some customers (Xilinx), but it is still a very exclusive technology.
2.5D is what they're doing right now with memory chips, right? They said it works well only if all chips are identical, so it could be perfect for stacking the Cell processors, connecting the EIB between identical chips. Adding as many as the target TDP allows.
 
Not that it's super important, but AMD 38xx chips were 192mm2, had a devilish 666 transistors @55nm.
Indeed I double confirm :LOL:

It's somehow important at least for my talk... :LOL:
~200 sq.mm 256 bit bus
~150 sq.mm 192 bit bus
~100 sq.mm 128 bit bus.

Then depending on the price per lithography, 200 sq.mm with high yield may be cheaper than a tinier part on bleeding edge process.
Big IF 90nm wafer are on third of the price of 32/28nm then 200sq.mm is cheap. Cooling will be easier cheaper (bigger area).

Whatever is stuck to the Northbridge you remove quiet an headache looking forward, like Edram but in a more flexible way and with more benefits.
 
chip stacking may refer to the memory only. there was talk on that forum about putting stacked memory dies on the same package as a CPU/GPU chip, both on top of an intermediary layer with a huge lot of connections, and that would give truly ginormous amount of bandwith on about a gigabyte of memory.

well there was talk of it on this very topic :), the use of "silicon interposer".

that would be seriously great on the PS4, and is an integrated design you won't find for a while on PCs.
it's also easier/understood technology, manufacturing DRAM and stacking it is easier while stacking power hungry CPU/GPU slices is a much harder problem (including cooling) and is far away, perhaps the end of the current decade.
 
Rumors from some E3 or CES or something were the originals claiming AMD holds contracts for all 3.

The way Huang said what he said pretty much confirms AMD is making all of them if nV had the deal in their pocket even for one, he would have made it clear, and not just state that "one company isn't capable of building all 3" as the only reasoning why nV would make gfx for one of them.

Sigh.
 
You do know that Jen-Hsun Huang is the CEO of Nvidia and as such is accountable to the stockholders for the information he is sharing?

If he says they are building one of the next gen consoles, I am inclined to believe there is some truth behind it.

I know who he is, he didn't actually say it, he sort of implied they would have to do it. So he is not responsible in any way to anyone for that tidbit of non-information.

I really don't know if they will or won't, but I don't see that quote as any confirmation one way or the other.
 
2.5D is what they're doing right now with memory chips, right? They said it works well only if all chips are identical, so it could be perfect for stacking the Cell processors, connecting the EIB between identical chips. Adding as many as the target TDP allows.
Not really. 2.5D means that several chips are mounted on a common silicon substrate (interposer). The main advantages are that it allows a compact package and high bandwith at low signal voltage, read very power efficient.

This is a quite informative read on the subject

http://www.monolithic3d.com/2/post/2011/12/is-the-buzz-around-xilinxs-25d-fpga-justified.html
 
it's also easier/understood technology, manufacturing DRAM and stacking it is easier while stacking power hungry CPU/GPU slices is a much harder problem (including cooling) and is far away, perhaps the end of the current decade.

My impression was that stacking was to be more useful for ULP mobile setups, not a couple hundred TDP behemoths, let alone >50W. :p
 
My impression was that stacking was to be more useful for ULP mobile setups, not a couple hundred TDP behemoths, let alone >50W. :p

That is the beauty of 2.5D. You dont stack power hungry chips on top of each other, but next to each other on a common silicon interposer, which still gives most of the advantages with regard to fast and power efficient communication that 3D stacking offers.

Sony seems pretty keen on this technology.

http://eandt.theiet.org/magazine/2011/12/maasaki-tsu-interview.cfm

This vision will need to leverage an emerging chip manufacturing technology: through silicon via (TSV). It stacks multiple pieces of silicon in 3D structures interconnected by pathways that run through the chips themselves. The technique promises to hugely reduce latency and boost performance by greatly reducing the wires signals must traverse. It offers a high integration of traditional and graphics processing alongside analogue.Given that today’s advanced fabs are operating in 28nm nanometer process geometries and advancing on 20nm, it becomes clear how incredibly delicate and complex a task this is. After all, it’s hard enough right now to lay out a ‘flat’ 28nm chip and get it to yield in profitable quantities with such minute features. As a result, the leading chip foundries are working hard on 3D but offer typical customers an interim/2.5D alternative, often called a silicon interposer. This looks to integrate silicon more closely side-by-side than stacked.

and....

http://www.electroiq.com/blogs/insi...l-summit-part-2-3dic-technology-and-test.html

Recall in IFTLE 62 I discussed the nomenclature confusion part of which was "stereoscopic 3D" being confused with 3D IC. [ see IFTLE 62, "3D and interposers: Nomenclature confusion"] Well, I never thought I would see a presentation about 3D IC being used for stereoscopic 3D but that's just what happened when Taiji Utaka, SVP of technology platforms at*Sonydiscussed the incorporation of 3D IC chips into the stereoscopic 3D Sony PlayStation. Sony is looking at the potential of improving 3D image quality by using 3D IC memory to increase performance (pixel fill rate improved by higher bandwidth) and improve latency. Sony sees the major impediment to using 3D IC as current cost, but also includes test protocol, thermal performance, proven reliability, standardization, and the availability of multiple suppliers as issues that need to be improved. Utaka interestingly noted that "game machines are required to have longer lifetime than PCs."
 
Not sure how applicable it is, but a true 3D chip, with processors and ram, taped out in october from Sony-Ericsson:
http://eda360insider.wordpress.com/...the-advanced-notes-can-you-say-tour-de-force/
"The finalized Wide I/O spec operates at 266 Mtransfers/sec using an SDR (Single Data Rate) protocol, which delivers 17 Gbytes/sec of data throughput via four parallel, independent, 128-bit memory channels (4.26 Gbytes/sec/channel)."

This is at minuscule cell-phone power level. So even if they end up doing "only" 2.5D, could we get something like a terabyte or two per second?
 
OK I may have been wrong :p

In all seriousness though I dont tend to believe random rumors, but then again I suppose the Forbes name gives it a bit more cred.

It's funny to see how both companies "AMD & Nvidia" are contradicting to each other in their statements. Technically speaking one should put more credibility in Jen sung Huang's words since he is still NV's CEO over an ex AMD employee. I would be very surprised if Sony switched company this soon. But in the end I think it's better for Sony to side with NV mainly due to backwards compatibility reason.
 
It's funny to see how both companies "AMD & Nvidia" are contradicting to each other in their statements. Technically speaking one should put more credibility in Jen sung Huang's words since he is still NV's CEO over an ex AMD employee. I would be very surprised if Sony switched company this soon. But in the end I think it's better for Sony to side with NV mainly due to backwards compatibility reason.

But the only thing I've seem to have heard from nVidia was the assumption that they would make a GPU for a console only because no one could handle all three. I had heard people talk about nVidia saying they were handling a console in the past, but that was my first time seeing the actual context. The few rumors I have seen about the PS4 have all pointed to an AMD GPU.
 
Status
Not open for further replies.
Back
Top