Slashdot Mirror
Moore's Law Fails At NAND Flash Node
An anonymous reader writes "SanDisk sampling its 1Y-based NAND flash memory products and has revealed they are manufactured at same minimum geometry as the 1X generation: 19 nm. The author speculates that this is one of the first instances of a Moore's Law 'fail' since the self-fulfilling prophecy was made in 1965 — but that it won't be the last."
There's a granularity to advancement as it is made of discrete units of advancement and invention.
Also, I wouldn't pooh pooh the use of other techniques to keep things moving. In the terms economists use to analyze advancement, this is called "substitution", and is the source of the counter-intuitive but powerfully predictive observation that, in a free economy, people can invent ahead of the curve faster than things become problems, like shortages.
(-1: Post disagrees with my already-settled worldview) is not a valid mod option.
Moore's Law applies to the number of transistors in a chip. Just because you have found an increase in performance that did follow Moore's Law for a while does not mean that Moore's Law is somehow about flash memory. Therefore, when the increase no longer follows Moore's Law, it does NOT mean that Moore's Law has failed. The only thing that has failed is your own prediction that things other than the number of transistors would follow that curve.
How does this have anything to do with Moore's law?
Moores law doesn't refer to density in any way. Especially not that of storage.
Moores law was talking about CPU's and their complexity.
But if you insist on claiming storage should be measured this way, then
Take a look at average flash disk size (and actual cost to manufacturer since you must apply some sort of baseline ) and I'd say you're easily still getting a doubling every 2 years.
I don't think the summary writer knows what that means.
From the article: Some might argue that the die area saving achieved is equivalent to a process node move, and that as Moore's talked about the number of transistors per IC his law is not dependent on a reducing minimum geometries. I think that most will see that this runs against the "spirit" of Moore's Law.
From Wikipedia: Moore's law is the observation that, over the history of computing hardware, the number of transistors on integrated circuits doubles approximately every two years.
From article linked off the main article: SanDisk has now revealed that 1Y – now described as a generation rather than a node - is the company's second generation at 19-nm. What the company does claim to have achieved is a reduction in the memory cell size from 19-nm by 26-nm to 19-nm by 19.5-nm, delivering a 25 percent reduction of the memory cell area.
So, if you can fit more cells using the same size process, it doesn't go against the spirit or the letter of Moore's law. Moore's law is about computing power. If you get more computing power without reducing size to do it, that still counts.
Big apple, new Yorik, undig it, something's unrotting in Edenmark.
Even if Moore's "law" finally runs out, we'll still find ways to advance. Just as the multi-core shift has prolonged it kinda sorta in the CPU space, 3D chip design will continue to move us forward for the time being, until quantum computing or something novel based on memristors becomes available.
"A truly wise man realizes he knows nothing."
"And thirdly, the code is more what you'd call "guidelines" than actual rules." (Mod points for knowing where that quote comes from.. )
Given the physics of how flash actually works, I'm guessing that we will see a more step wise improvement in storage density. But Moore's law is about increasing complexity, not density. So the logical size of flash devices will continue to go up, even if the density is not improved.
I don't think we are close to a point where Moore's law is going to be proven false, not by a long shot.
"File to fit, pound to insert, paint to match" - Aircraft Maintenance 101
It's equally valid to say SanDisk failed (umm, violated) Moore's law.
Do you even lift?
These aren't the 'roids you're looking for.
So what do you think NAND flash is made of? Tiny spinning hard drives? Magnetic bubbles? Pixie dust? NAND flash is made of (you guessed it) transistors on chip. As such, it is perfectly reasonable to expect it to conform to Moore's law.
Support Right To Repair Legislation.
And when it comes to NAND we all know the dirty little secret they don't like talking about, with each shrink the lifespan gets shorter so they have to add more and more extra space to replace the dying cells and you end up losing any gains you may have made. That is why I hope something new will end up coming out that will let us have the power saving of SSD with the long life of the HDD, because the consumer level MLC chips frankly aren't very good.
ACs don't waste your time replying, your posts are never seen by me.
Moore's law has never been a 'law', it's a historical observation.
It has never claimed that this will be true going forward, merely that at the time it was observed that was the case, and it's largely held up since then.
The fact that it's held true this long is staggering, but the fact that it might be running out is hardly surprising. Moore never claimed this would continue forever.
Lost at C:>. Found at C.
Every word he said is wrong, including "and" and "the".
Since Moore's Law is silent on "minimum feature size", then no observation of that metric can be contradictory to that law.
I am not an engineer. So, you engineers out there--are we nearing the theoretical limit on these things? I mean, 19 nm is pretty darn small. It seems to me that at some point Moore's law has to fail simply because you can't make a connection less than one atom thick. And making a connection one atom thick would be stupid, I would think, for reliability reasons. So--is Moore's law, as extended to NAND flash memory failing due to the fact that it has nearly reached its lowest theoretical size?
A Moore's law 'fail'.
Or Moore's law "failure", for those of us over the age of 12.
Moore's "Law" has to fail eventually, because if you keep doubling (the amount of transistors on a chip) every couple of years you would soon (in a century or two) have more transitors than there are (elementary) particles in the universe
Of course we will be up to Windows 95 by then...
... can never exceed about 90 mph.
So, that's it folks. That's as fast as humans will ever go.
Perhaps Moors Law is sigmoidal.
Gate oxide thickness at 19nm is about at limit.
Remember bubble memory?
I wonder if that would have kept up with Moore's Law a lot better.
This isn't a new issue to people in the industry. Here's a more useful article from last year: "Is the cost reduction associated with IC scaling over?" "Clearly, dimensional scaling is no longer associated with lower average cost per transistor."
The cost of wafer fabs has been going up with each generation. Intel says that a cutting-edge fab now costs upwards of $10 billion, twice the previous generation. That's why higher densities no longer reduce cost. The upper limits of optical lithography are being reached because light, even "deep ultraviolet" light, is too coarse a tool. "Extreme ultraviolet" (soft X-rays, really) are being tried to get down to 10nm or so, but the processes are currently slow and barely work. Electron beam machines, which can go below 10nm, have been around since the 1980s, but they work by writing the chip with an electron beam, not with a mask, which is very slow for a production process.
This is for mostly-static memory. For active transistors, as in CPUs, heat dissipation is already limiting density. CPU clock speed maxed out between 3 and 4 GHz several years ago. (Yes, 8GHz has been achieved with an AMD CPU running in liquid helium. So?)
With the upper limits of speed and density in sight, work is now focusing on reducing cost and power consumption. Hence the push to use ARM CPUs in more applications.
16 FinFET is basically the same size as 22nm but backed in a different orientation to have slightly more transistors in one dimension. But in reality it is a costly stumble as we reach the end of Moore's law.
Those that write 'comment' or 'editorial' articles for magazines and newspapers are the original trolls. They do not intend to inform or engage in worthwhile analysis. They merely write to get a 'rise' from the readership, and in doing so please their editor and publisher.
So called Moore's Law, like all the others found in IT, have the same depth of meaning as Murphey's Law. In other words, they are designed to highlight a common current phenomenon that acts as a simplistic understanding of far more complex sets of inter-relating systems.
Making 'chips' is an industrial process with many dimensions.
-cost of new fabs
-cost of R+D that will lead to creating the new machines that will operate in these fabs
-cost of R+D that will create/discover the new physical/chemical processes that will be used by the above machine tools (eg., methods of lithography and chemical materials for the semiconducting elements
-current need for more transistors per chip
-current need for faster chips
-current need for lower power chips
-size of current market for a given type of chip
-size of future potential market for a given type of chip
Moore's Law was a simplistic observation of the cycle of chip process improvements at the beginning of the integrated circuit industry. It essentially observed that one could continuously improve a chip making process by small increments (in the early stages- at some point one would obviously bump against atomic limits), using the growth of chip sales to fund the cycle.
Only a complete cretin would fail to notice the long term problems of such a prediction (and I don't mean the problems of atomic limits). For a start, the sales of chips must continuously grow if the expense of producing smaller process plants also grows. However, no product marketplace ever grows indefinitely. More interesting is that one must constantly find a need/use for ever larger numbers of transistors per chip.
Take Intel. Once it became apparent that the dreadful OS from Microsoft could NOT usefully support more than THREE CPU cores, Intel gave up the idea of mainstream parts with ever growing numbers of cores, and settled on 4 core designs with ever more transistors thrown at each core. Doubling the number of cores doubles the performance in an ideal computer setting (as seen with your GPU hardware). Doubling the number of transistor per core, however, tends to give only improvements of around 5% (these days), but even that may be mitigated by having a core that no longer clocks as fast. In other words, if you ignore Intel's dreadful integrated GPU, two process shrinks have given Intel no real worthwhile improvement in its 4-core high-end CPU parts.
Memory, the lazy-man's attempt to keep up with Moore's Law, suffers from the same issue. Take your PC RAM. That RAM is really just the L4 cache (before the L5 cache of your SSD). Computer Science theory explains that there is an optimal size for any given level of cache, beyond which you are just wasting energy and money. More RAM also means slower RAM, which at some point will eliminate ANY advantage of adding more RAM. My point is when the useful RAM limit is reached, growth in the memory market depends on growth in new markets, not in selling more memory to existing markets.
Companies making memory (tradition RAM or Flash) now need to focus on cost and/or power usage far more than capacity. So called SoC designs are soaking up increasing numbers of transistors (at least for the time being), but they too now have a major focus on power usage and cost.
Nvidia famously published a paper a year or so back where it showed the cost of moving to a new process could soon overwhelm the advantage of staying on the previous process. Technologies like FD-SOI allow an existing process to out perform a new process that follows the same dumb/lazy methods of semiconductor design. We are in an age (at last) where thinking smart is far more useful than simply banging out another dumb shrink. Intel has found this out to its cost with the notable failures of the technology used to build both the Ivybridge and Haswell parts. In neither case did 'shrink' over 'think' prove advantageous.
What does "SanDisk sampling" mean? Does it mean SanDisk *is* sampling? What are they sampling? I don't understand what it means in this context. And Moore's Law is not a "self-fulfilling" prophecy. In what way it is "self-fulfilling"? To be self-fulfilling the prophecy would have to make itself come true somehow. Are you saying if people hadn't been trying to keep up with Moore's law, electronic technology would have advanced more slowly? The story summary is complete gibberish. Why was it posted like this?
The author is little late with his observations.
Flash is slow (compared to sram), power hungry (requires cap banks to tunnel electrons), unreliable (tunneling eats oxide) and not terribly dense. I wish one of the replacements (memristers or whatever) would come online into mass production with much haste.
Anyway moore's law really boils down to cost per widget... If you can pack more things into a component and use less die it is cheaper and u win. These days cost/power are really the objective functions here not process size exclusivly.. If you can't build smaller features then you can always add more layers/stack shit to your hearts content. Even something as easy to understand as using larger wafers reduces cost.
For transistors that act like capacitors, see Dynamic logic and Floating-gate transistor and 1T DRAM.
What about motherboards, BIOS, DDR, harddrives & toasters? Moors law doesn't seem to apply there
For DDR, it's already up to the limits of how fast the player's feet can move. It took a few years to get from the original Dance Dance Revolution, whose hardest song "Paranoia" had bursts of six steps a second, to the 10 Hz bursts of "Max 300" in DDRMAX: Dance Dance Revolution 6th Mix. It took even longer to get to runs of over 13 Hz in "Fascination Maxx" in Dance Dance Revolution Supernova.
It is also common in the industry, to improve on something on an existing process. Intel does it every tingle process with their tick-tock strategy. They make a new process and release a chip with a largely existing design on it, then they make a new design on that process that is more efficient, then a new process, and so on. You see it with the GPU makers, they release updates outside of when the process shrinks and so on.
There was never anything saying that increasing transistor count was the only way to increase performance.
Maybe its time to ditch binary logic and think up of something new. Quantum computing as limited as it is is a candidate but there can be other options.
Or how about making 3D wafers and figure out a way to deal with noise and heat.
Well, to be fair the only experiment we know of resulted in some weird temporal displacement effect, followed by a spectacular crash...
For the sake of argument, Moore's Law states that transistor COUNT will double, not their density on the chip. They could've doubled the surface area of the die, theoretically.. though that would be of limited use due to the constant demand for shrinking everything.
So, since he had the qualifier "for at least ten years", can we now say Moore's Law was proven to be correct?
It gripped her hand gently. 'Regret is for humans,' it said.
Silicon atom covalance radix is 0.11 nm, which means that at 19 nm, tracks are a large of around a hundred of atoms. Obviously we are not going to shrink a lot further.
Do you actually have a plan to do that? A plan, as opposed to a nebulous wish? I had such a plan. My plan was 15% increase per month for 12 months. I then followed that plan, thereby increasing my income by 15% per month.
On the other hand, if Intel planned to increase by 20% every 18 months, that would practically guarantee they wouldn't double in that time period. I'm guessing your plan for the next 18 months, to the extent that you have one, is to leave your income exactly the same as it is today. Is that more accurate?
Two internets for you, and for GP
1st, physical law has limitations too, for example newton's law doesn't apply when your speed is near the speed of light. 2nd, there're speculations that physical law does change over time, for example fine structure constant may not be constant at all.