http://www.guru3d.com/news_story/in...have_pci_express_4_ddr4_and_sata_express.html
I just read some latest developments and intel would be releasing the 14nm for consumers by 2014 only. As far as 10nm is concerned then looking at the way intel had managed these things in the past. I really doubt that they would release 10nm before 2016 for consumers. It would require change in material for that. And they would really want to squeeze out the last bit of profit before jumping to that one. Untill or unless AMD would do something about that.
Please read the news from IDF, okay? You posted an old Xeon roadmap, not a current desktop roadmap. Also, Skylake is the second generation 14nm product.
At IDF Intel stated 10nm in 2015 and 7nm in 2017.
Please understand that Broadwell is not shipping for revenue this year. Krzanich was referring to ESs shipping to OEMs to develop their products.
already 50 functioning Broadwell products are on display.
Wanna bet on that?
OEM's already have samples. How do you think they are showing products? IDF is more than a couple of speeches, already 50 functioning Broadwell products are on display.
TR said:IDF — Today, during his opening keynote at the Intel Developer Forum, Intel CEO Brian Krzanich demonstrated a working laptop based on a 14-nm chip. The processor, code-named Broadwell, is a die-shrunk version of the 22-nm Haswell parts underpinning today's Core i3/i5/i7 CPUs.
Krzanich didn't offer too many details, but he revealed that Intel's 14-nm process is now viable. The company expects 14-nm Broadwell processors to begin shipping by the end of this year, with products reaching the hands of end users in 2014.
Kirk Skaugen said:During a keynote, Kirk Skaugen, the Intel senior vice president in charge of PC clients, showed off a new $200 Senz3D camera from Creative. It supports apps that use Intel’s perceptual computing software development kit.
Intel’s 14-nanometer Broadwell chip, coming for Ultrabooks in late 2014, will support integration of a 3D camera directly into the bezel of a laptop, Skaugen said. That’s a lot of progress since 2012, when the company announced the perceptual computing initiative to move beyond the mouse and keyboard.
As indicated previously,the first Haswell silicon was available in Q4 2012,and CPUs launched over six months later. Ivy Bridge silicon was available months before retail availability. Trinity and Kaveri silicon was available months before retail availability(shown in laptops),even the BD samples which indicated its probably performance.The first Broadwell silicon is supposed to ship to customers by the end of this year, and in systems next year.
Excavator is 28nm and will be coming in 2015.
Do you have any reliable source of such information or it's just pure speculation?
Do you have any reliable source of such information or it's just pure speculation?
It is pretty normal for prototype chips and low volume ES chips for validation to be around months before any sort of retail availability. People underestimate how long it takes to validate and test things especially the infrastructure surrounding such CPUs. Prototype motherboards,validation of motherboard designs,final CPU clockspeeds,new instruction sets,etc. This is not new.
You're absolutely right. The likely explanation is that Apple's chip is mostly GPU, which can be made much more densely than CPU.
I simply believe TSMC's denser than Samsung at the same node designation. There's no other way to explain why Cortex-A9 is of comparable size on TSMC's 45nm (OMAP4430) vs Samsung's 32nm (A5r2, A5X, Exynos 44xx.) It's not like TI achieved better density by sacrificing electrical performance, OMAP44xx played at the same clock speeds and power consumption as their competitors. Freescale's i.MX6 was also very dense - quad core A9 and pretty decent GPU at only 52mm^2 (http://cache.freescale.com/files/32bit/doc/app_note/AN4579.pdf search for die size) on TSMC 40nm process (http://www.newelectronics.co.uk/ele...iconductor-expands-apps-processor-range/39377). Compare with Apple A5 on Samsung's 45nm, dual Cortex-A9s and GPU that's better but not earth-shatteringly so, which weighs in at a whopping 122mm^2. Look how much people were slagging Apple's die sizes, comparing them to much more sophisticated Intel chips. And monsters like A5X were GPU-heavy compared with everything else, so if that argument didn't help then.. Tegra 2-4 also look very favorably when compared in die size vs their nearest Apple equivalents.Some people are saying that A7 is a lot denser because it's proportionately more GPU and GPU is denser, but I doubt this because higher end Series 6 is supposed to be more area efficient than Series5XT - which is what you'd expect because they don't eat the area inefficiency cost of stacking together separate cores. Since they "only" get 2x the GPU performance and since I suspect part of that would be due to a clock boost I don't expect there'd be substantially more than 2x the GPU transistors used, even accounting for the extra space needed for new features.
But for what it's worth, I don't agree with that explanation. This is what I posted about it on beyond3d:
I simply believe TSMC's denser than Samsung at the same node designation. There's no other way to explain why Cortex-A9 is of comparable size on TSMC's 45nm (OMAP4430) vs Samsung's 32nm (A5r2, A5X, Exynos 44xx.) It's not like TI achieved better density by sacrificing electrical performance, OMAP44xx played at the same clock speeds and power consumption as their competitors. Freescale's i.MX6 was also very dense - quad core A9 and pretty decent GPU at only 52mm^2 (http://cache.freescale.com/files/32bit/doc/app_note/AN4579.pdf search for die size) on TSMC 40nm process (http://www.newelectronics.co.uk/ele...iconductor-expands-apps-processor-range/39377). Compare with Apple A5 on Samsung's 45nm, dual Cortex-A9s and GPU that's better but not earth-shatteringly so, which weighs in at a whopping 122mm^2. Look how much people were slagging Apple's die sizes, comparing them to much more sophisticated Intel chips. And monsters like A5X were GPU-heavy compared with everything else, so if that argument didn't help then.. Tegra 2-4 also look very favorably when compared in die size vs their nearest Apple equivalents.
I know people are always saying Samsung is supposed to have better density because of gate first vs gate last but I'm just not seeing it. Is my line of thought really so off here?
And really, Apple never boasted about transistor counts or density before. They didn't talk about it when they made outright shrinks in particular (the pure shrinks like A5r2 weren't in any way even acknowledged at all, just silently slipped into existing products). All of a sudden they're talking about making a chip that packs double the transistors into the same area to an audience that mostly knows nothing about transistors, while meanwhile giving the same old generic "2x better" crap for CPU and GPU. Sounds to me like they're happy with TSMC giving them better density than Samsung and are boasting about it as a way of publicly flipping Samsung off.
Now, I know the GlobalFoundries supporters aren't going to like this, so as a disclaimer I'll just say that, while to the best of my knowledge Samsung and GF's process are supposed to be "similar" I don't directly intend for this stuff to apply to GF w/o more information. I pushed that before and didn't really have good information to back it up so I take it back
There isn't any truth to it is the bottom line. Someone grabbed onto the fact Intel didn't scale their M1 pitch at 22nm as aggressively as they did for 32nm and decided to try to make a mountain out of a molehill before everyone else published their 22nm pitch design rules...once the other's published theirs the picture got quite a bit different because Intel actually has tighter pitch than the 22nm competition (IBM).
But the real reason why there is nothing to the original argument is the simple fact that nodes have no definition. There is no rule or law or industry standard for defining what constitutes a 22nm node, a 20nm node, etc. It is a label that is used solely at the discretion of whoever wants to use it.
You could, right now, declare to the world you have invented a 7nm node if you wanted. What are the design rules? Could be a breadbox layout for all it matters, there is no definition of what a 7nm is such that your claims would become technically invalid.
It is just silliness really, marketing silliness.
did it not indicate gate length at one time?
At IDF Intel stated 10nm in 2015 and 7nm in 2017.
Sounds to me like they're happy with TSMC giving them better density than Samsung and are boasting about it as a way of publicly flipping Samsung off.
Now, I know the GlobalFoundries supporters aren't going to like this, so as a disclaimer I'll just say that, while to the best of my knowledge Samsung and GF's process are supposed to be "similar" I don't directly intend for this stuff to apply to GF w/o more information.
Theses expectations are for marketing purposes , there will
be no 10nm in 2015 ; let alone 7nm in 2017.
Think about it , if 14nm is released in 2014 , at best ,
this would imply one year use before replacement
by 10nm , this doesnt make sense at all.
14nm is shipped to OEM's in 2013, available to regular people in 2014.
10nm to be shipped to OEM's in 2015, available to regular people in 2016.
7nm to be shipped to OEM's in 2017, available to regular people in 2018.
Nothing unnatural here. The Kabini/Temash SoCs die size is 110 mm2. And they are made at TSMCs 28 nm process node, with more then 1 billion transistors inside.Just thinking about the Apple event yesterday. They said 1 billion transistors on the same die size (100mm2), which means they must have shrunk it to 20nm.
You can forget the two years dev cycle for both 10nm
and 7nm , theses are no more trivial nodes , current tools
are to be replaced completely as they wont be extendable
past 14nm.
Who knows? Tick-tock is still alive and well despite all the complaining about actual performance. 10 and 7 might also prove to be available as planned.
Right now, the question that I'm most interested in is whether Broadwell will be available in non-BGA so that I can buy one for my s1150 motherboard.