Slashdot Mirror
End of Moore's Law Forcing Radical Innovation
dcblogs writes "The technology industry has been coasting along on steady, predictable performance gains, as laid out by Moore's law. But stability and predictability are also the ingredients of complacency and inertia. At this stage, Moore's Law may be more analogous to golden handcuffs than to innovation. With its end in sight, systems makers and governments are being challenged to come up with new materials and architectures. The European Commission has written of a need for 'radical innovation in many computing technologies.' The U.S. National Science Foundation, in a recent budget request, said technologies such as carbon nanotube digital circuits will likely be needed, or perhaps molecular-based approaches, including biologically inspired systems. The slowdown in Moore's Law has already hit high-performance computing. Marc Snir, director of the Mathematics and Computer Science Division at the Argonne National Laboratory, outlined in a series of slides the problem of going below 7nm on chips, and the lack of alternative technologies."
The party's over. Get to work on efficient code. As for the rest of all you mothafucking coding wannabes, suck it! Swallow it. Like it! Whatever, just go away.
Its more of a prediction, that has mostly been on target cause of its challenging nature
Now the blind ants (researchers) will need to explore more of the tree (the computing problem space)... there are many fruits out there yet to discover, this is just the end of the very easy fruit. I happen to believe that FPGAs can be made much more powerful because of some premature optimization. Time will tell if I'm right or wrong.
The really sad thing regarding this "Moore's Law" thing is that, while the hardware had kept on getting faster and even more power efficient, the software that runs on them kept on becoming more and more bloated.
Back in the days of pre-8088 we already had music notation softwares running on Radio Shack TRS-80 model III.
Back then, due to the constraints of the hardware, programmers had to use every trick on the book (and off) to make their programs run.
Nowadays, even the most basic "Hello World" program comes up in megabyte range.
Sigh !
Muchas Gracias, Señor Edward Snowden !
We might even stop writing everything in Javascript?
How can I believe you when you tell me what I don't want to hear?
They should have called it "Moore's Trend" or "Moore's Observation."
..took us in directions we hadn't considered.
Forget the exact quote, but what a time to be alive. My first computer program was written on a Vic-20. Watching the industry grow has been incredible.. I am not worried about the demise of traditional lithographic techniques.. I'm actually expecting the next generation to provide a leap in speed as now there's a strong incentive to look at different technologies.
Here's to yet another generation of cheap CPU.
..don't panic
We might stop seeing ridiculous gains in computing power, and might have to start making gains in software efficiency.
Perhaps the focus on portable computing has pulled research money away from high-end chips. The demand for high-end PC's is flattening out, and so R&D money is being pulled out of there and into ARM-level chips to have existing CPU power in smaller, battery-friendly boxes.
Table-ized A.I.
Ridiculous, talking about Moore's law as if it were a part of nature. When in fact it is the product of innovation and ingenuity, which you say it hampers. Its not Moore's law that's slowing down innovation, its lack of innovation that is invalidating Moore's law.
This is ok. For many purposes, software improvements in terms of new algorithms that are faster and use less memory have done more for heavy-dute computation than hardware improvement has. Between 1988 and 2003, linear programmng on a standard benchmark improved by a factor of about 40 million. Out of that improvement, about 40,000 was from improvements in software and only about 1000 in hardware improvements (these numbers are partially not well-defined because there's some interaction between how one optimizes software for hardware and the reverse). See this report http://www.whitehouse.gov/sites/default/files/microsites/ostp/pcast-nitrd-report-2010.pdf. Similar remarks apply to integer factorization and a variety of other important problems.
The other important issue related to this, is that improvements in algorithms provide ever-growing returns because they can actually improve on the asymptotics, whereas any hardware improvement is a single event. And for many practical algorithms, asymptotic improvements are occurring still. Just a few days ago a new algorithm was published that was much more efficient for approximating max cut on undirected graphs. See http://arxiv.org/abs/1304.2338.
If all forms of hardware improvement stopped today, there would still be massive improvement in the next few years on what we can do with computers simply from the algorithms and software improvements.
Some implications:
Moore's yawn ... er, law. It has ended, again again. It must be the co-joined twin of Voyager which has left the solar system 78 times in the past 14 years.
Wake me up when some real news gets in.
This quote from the article: " But it may be a blessing to say goodbye to a rule that has driven the semiconductor industry since the 1960s." is surely the dumbest thing I've read all day.
Seriously? It's like, people wake up and say, "it would be such a blessing if I could never get a faster computer." Does that make sense at all?
"First they came for the slanderers and i said nothing."
In my mind it was an interesting statistical coincedence, *when it was first discussed*
Then the hype took over, and we know what happens when tech and hype meet up...
Out of touch CEO's get hair-brained ideas from non-tech marketing people about what makes a product sell, then the marketing people dictate to the product managers what benchmarks they have to hit...then the new product is developed and any regular /. reader knows the rest.
It's bunk. We need to dispel these kinds of errors in language instead of perpetuating them, because it has tangible effects on the engineers in the lab who actually do the damn work.
Part of what made the Moore's "Law" meme so sticky is how it was used, usually in a simple line graph, by "futurists" who barely can check their own email to pen mellodramatic, overhyped predictions about *when* we would have 'AI'.
AI hype is tied to computer performance, and Moore's "Law" was something air-head journalists could easily source, complete with a nice graph from a tech "expert"
I know my view of AI as a fiction is in the minority, but IMHO we need to grow up, stop with the reductive notion that computing is progressing towards some kind of 'AI' singularity and focus on making things that help people do work or play.
Our industry looses **BILLIONS** of dollars and hundreds of thousands of work-hours chasing a fiction when we could be making more useful, powerful, and imaginitive things that meet actual, real world human needs.
To bring this back to Moore's Law, let's work on better explaining the value of tech to non-techies. Let's give air-headed journalists something to sink their teeth into that will help our industry progress, not play the bullshit/hype game like every other industry.
Thank you Dave Raggett
The defining characteristic of the 7nm is that it's the one after the 10nm node. I can't remember the last time I worked in a process where the was a notable dimension that matched the node name, either drawn or effective.
Marc Snir gets bogged down in an analysis of gate length reduction which is quite besides the point. If it gets harder to shrink the gate than to do something else, then something else will be done. It worked on processes with the same gate length as the "previous" process, and I've probably even worked on a process that had a larger gate than the previous process. The device density still increased, since gate length is not the only dimension.
I agree in sprit but you perpetuate a false dichotomy based on a misunderstanding of *why* software bloat happens, in the broad industry-wide context.
Just look at Windows. M$ bottlenecked features willfully because it was part of their business plan.
Coders, the people who actually write the software, have always been up to the efficiency challenge. The problem is the biggest money wasn't paying for ninja-like efficiency of executing user instructions.
It was about marketing and ass-backwards profit models forced onto the work of making good code.
I've often observed that in order to do the most desirable work, a coder would have to sacrifice the very thing that made them want to work on the best software in the first place...
Thank you Dave Raggett
If, just if anything can be done, Governments will not play any significant role in this process. They do not even seem to understand the problem, how could they ever be part of the solution? And that is just it: It is quite possible that there is no solution, or that it may take decades or centuries for that one smart person to be in the right place at the right time. Other than blanket research funding, Governments cannot do anything to help that. Instead, scientific funding is today only given to concrete applied research that promises specific results. That is not going to help making any fundamental breakthrough, quite the opposite.
Personally, I expect that is it for computing hardware for the next few decades or possibly permanently. I do not see any fundamental issue with that. And there would be quite a bit of historic precedent for a technology to slowly begin to mature.
As software is to incredible unrefined these days, that would be a good thing. It would finally be possible to write reasonable standard components for most things, instead of the bloated, insecure mess so common these days. It would also be possible to begin restricting software creation to those that actually have a gift for it, instead of having software created by the semi-competent and the incompetent (http://www.codinghorror.com/blog/2010/02/the-nonprogramming-programmer.html). In the end, the fast process of computing (which burned a few centuries of fundamental research) was not a good thing. Things will be moving much slower while new fundamental research results will be created instead of merely consumed.
Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
One of the reasons our brains are more powerful is because we are 3d and a chip is still stuck in 2d.
When will we see 3d or many layered chips turning simple instructions into complex ones similar to a neural net?
Or if machines are fast enough (witness those on 10 year old XP boxen hold outs) no one will care? When can we have our own 3cpos?
http://saveie6.com/
Picture a super computer so massive that components fail as fast as we can replace them. Now that is a big super computer. This is the issue. Super computers have a physical limit. If the node power doesn't grow we will reach a limit on simulation power. It will be interesting to see how the CPU matures. That means more features will be developed beyond raw power.
I thought they finally killed that off last month. Or the other 7K times it was declared dead.
Serious? Seriousness is well above my pay grade.
Misleading.
Yes, I've got 100 fold improvments on a single image processing algorithm. It was pretty easy as well.
However, that only speeds up that one algorithm, 10x faster hardware speeds everything 10x.
Use of interpretted languages and bloated code has more than equalled the point gains in algorithms.
The net result overall is that 'performance' increase has been mostly due to hardware, not software.
It feels like we have been coasting a bit with technology the past few years so this doesn't surprise me. Everything is just a smaller/faster version of the year before.
... time to stop writing garbage in visual basic, man up, and use proper languages again that are actually efficient, isn't it?
I run: Windows, OS X, Linux, FreeBSD. Just because you have a hammer, doesn't mean everything is a nail.
From the 50s-90s businesses ran fine with basically no PCs or 8086 machines.
Once the wall hits businesses will run fine too.
What I expect to see more of refinements of technology that make things work better.
Example: in the 70s you could buy a small portable tire repair kit. Problem was there was a metal seal across the opening of the tube of vulcanizing material, on the street it was a bitch to open. By the 80's the dsmr kit was available but on the outsige edge of the cap was a small spike. You unscreded the cap put it on backwards and pushed in. The spike would puncture the metal seal.
On possible future example. A build process for a linux kernel that scans equipment and creates the optimal kernel configuration.
So... TFA says that the end of Moore's law will result in a ton of new innovations aimed at... making computers faster??? Last time I checked making computers faster is what Moore's law is all about.
I think a better title would be "End of Traditional Silicon/CMOS Technology Forcing Radical Innovation to Keep Moore's Law Going!"
Come on now, someone always predicts an end to Moore's Law every so often, but it never happens.
There is no beginning. There is no end. There is only Moore.
https://www.youtube.com/watch?v=cugu4iW4W54
This year, she and her team announced they had made the first ever single atom transistor.
Perhaps writing efficient code will come back into style.
Coders, the people who actually write the software, have always been up to the efficiency challenge
The minority of coders, you mean. Most coders are and always have been very, very crappy
Back in the bad old days when hardware constrains were REAL only crazy people with diamond-hardened will dared to code.
For every project we undertook it had been an uphill battle, and for every project completed we had that rush of fulfillment, knowing that we had accomplished another impossibility.
Nowadays, the rocky and windy road that we used to travel on, has become a 6-lane highway. Every Tom, Dick and Harry (and with Mary Anne thrown in) can code with minimal difficulties.
They no longer have to be challenged by all the impossibilities, that is why, most coders of today can't even begin to compare to the coders back then.
I see many emerging technologies that promise further great progress in computing. Here are some of them. I wish some industry people here could post some updates about their way to the market. They may not literally prolong the Moore's Law in regards to the number of transistors, but they promise great performance gains, which is what really matters.
3D chips. As materials science and manufacturing precision advances, we will soon have multi-layered (starting at a few layers that Samsung already has, but up to 1000s) or even fully 3D chips with efficient heat dissipation. This would put the components closer together and streamline the close-range interconnects. Also, this increases "computation per rack unit volume", simplifying some space-related aspects of scaling.
Memristors. HP is ready to produce the first memristor chips but delays that for business reasons (how sad is that!) Others are also preparing products. Memristor technology enables a new approach to computing, combining memory and computation in one place. They are also quite fast (competitive with the current RAM) and energy-efficient, which means easier cooling and possible 3D layout.
Photonics. Optical buses are finding their ways into computers, and network hardware manufacturers are looking for ways to perform some basic switching directly with light. Some day these two trends may converge to produce an optical computer chip that would be free from the limitations of electric resistance/heat, EM interference, and could thus operate at a higher clock speed. Would be more energy efficient, too.
Spintronics. Probably further in the future, but potentially very high-density and low-power technology actively developed by IBM, Hynix and a bunch of others. This one would push our computation density and power efficiency limits to another level, as it allows performing some computation using magnetic fields, without electrons actually moving in electrical current (excuse me for my layman understanding).
Quantum computing. This could qualitatively speed up whole classes of tasks, potentially bringing AI and simulation applications to new levels of performance. The only commercial offer so far is Dwave, and it's not a classical QC, but so many labs are working on that, the results are bound to come soon.
The end of Moore's law is a concern only if it is assumed that transistors will be used indefinitely.
The International Technology Roadmap for Semiconductors is published regularly and has information on the maturity of emerging technologies like carbon. There are many possibilities for "more than Moore" improvement. http://www.itrs.net/Links/2012ITRS/Home2012.htm
You may see them, but no actual expert in the field does.
- 3D chips are decades old and have never materialized. They do not really solve the interconnect problem either and come with a host of other unsolved problems.
- Memristors do not enable any new approach to computing, as there are neither many problems that would benefit form this approach, nor tools. The whole idea is nonsense at this time. Maybe they will have some future as storage, but not anytime soon.
- Photonics is a dead-end. Copper is far too good and far too cheap in comparison.
- Spintronics is old and has no real potential for ever working at this time.
- Quantum computing is basically a scam perpetrated by some part of the academic community to get funding. It is not even clear whether it is possible for any meaningful size of problem.
So, no. There really is nothing here.
Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
The refinement of process has postponed this for a long while, but the time has come to explore new architectures and technologies. The Mill architecture is one such example, and aims to bridge the enormous chasm of inefficiency between general purpose CPUs and DSPs. Conservatively, they are expecting a tenfold improvement in performance/W/$ on general purpose code, but the architecture is also well suited to wide MIMD and SIMD.
Another area ripe for innovation is memory technologies, which have suffered a similar stagnation limited to refinement of an ancient technology. The density of both cache and main memory can be significantly improved on the same process with Thyristor-RAM or Z-RAM. Considering the potential benefits and huge markets, it is vexing that more resources aren't expended toward commercializing better technologies. Some of the newer technologies also scale down better.
Something to replace the garbage which is NAND flash would also be welcome, yet sadly there appears to be no hurry there either. One point is certain, there is a desperate need to find a way to commercialize better technologies rather than perpetually refining inferior ones. Though examples abound, perhaps none is more urgent than the Liquid fluoride thorium reactor. Molten salt reactors could rapidly replace fossil fuels with clean and abundant energy while minimizing environmental impact, and affordable energy is the basis for all prosperity.
I would put a big star next to the energy efficiency claim for photonics. It is currently limited greatly by the electrical-to-optical conversion efficiency of the source, e.g. a semiconductor laser. Where photonics really shines (pun intended) at the moment is in longer links. On the other hand, there is a lot of ongoing research that promises to improve efficiency and bring it in line with the decreasing efficiency of electrical interconnects at higher data rates and the physical densities required by such.
No more Less is More
No more Moore's Law
We're more or less
Left only multi-core.
Artificial intelligence is the study of how to make real computers act like the ones in the movies.
- 3D chips are decades old and have never materialized.
24-layer flash chips are currently produced by Samsung. IBM works on 3D chip cooling. Just because it "never materialized" before, doesn't mean it won't happen now.
- Memristors do not enable any new approach to computing, as there are neither many problems that would benefit form this approach, nor tools. The whole idea is nonsense at this time. Maybe they will have some future as storage, but not anytime soon.
Memristors are great for neural network (NN) modelling. MoNETA is one of the first big neural modelling projects to use memristors for that. I do not consider NNs a magic solution to everything, but you must admit they have plenty of applications in computation-expensive tasks.
And while HP reconsidered its previous plans to offer memristor-based memory by 2014, they still want to ship it by 2018.
- Photonics is a dead-end. Copper is far too good and far too cheap in comparison.
Maybe fully photonic-based CPUs are way off, but at least for specialized use there are already photonic integrated circuits with hundreds of functions on a chip.
- Spintronics is old and has no real potential for ever working at this time.
MRAM uses electron spin to store data and is coming to market. Application of spintronics for general computing may be a bit further off in the future, but "no potential" is an overstatement.
- Quantum computing is basically a scam perpetrated by some part of the academic community to get funding. It is not even clear whether it is possible for any meaningful size of problem.
NASA, Google and NSA, among others, think otherwise.
So, no. There really is nothing here.
I respectfully disagree. We definitely have something.
Oh boy have I got news for you! Apple is rumoured to be releasing something shiny! A random weather event happened and people used it to support their arguments! A government did an IT project and it overran! A new tool/language/paradigm holds promise for people who weren't using the previous one properly! Quantum computing! AI! Flying cars! Mars rocket! Someone just found out we're getting them all in 2014! Or at least we'll DEFINITELY make an important step toward them in 2014 or VERY SOON afterwards!!!
Regardless of Moore's Law ending or not, new generations of processors had being more and more closer in performance to it's predecessors for years... I recently replaced my six year old Core 2 Quad Q6600 for a brand new i5 4670K to gain about 38% of overall system performance. I mean... WHAT? Try comparing the performance difference from a AT-386 to a AT-486 or from a Pentium to a Pentium II... there where massive improvements in performance from one generation to another. I can't understand why information like that doesn't make headlines.
Why not de-bloat-ify software by removing the bloat layer (Java JVM, .NET runtime) and getting back to compiling binaries for the processors they run on? Oh, that's right, a decade and a half of en-bloat-ing software has made us reliant on the library functions supplied by the bloat layer. We'd have to start from scratch building native APIs. I'm a low-level C programmer, but haven't written any C in years because no one uses it (outside of embedded systems programming, which is its own self-contained world). The only way I can get paid is to write Java.
Well, because I'm lazy, I won't look up all of them, but:
3D chips, yes, they're used in storage, but storage capacity has never been an issue. To increase storage now, you always have the option of simply making it bigger. Can you use 3D chips to make logic devices?
If memrisistors are only useful in neural networks, then they're worthless. Neural networks are what's called a supervised learning method which was all the hot new rage in the 90s, but since they're difficult to design and are not easily adaptable, and become extremely complex for not terribly complex problems, they're almost never used except for the most remedial cases. They're largely being replaced by Support Vector Machines as they're simpler to design, more flexible, and so far there have been no problems which NNs can do that SVMs can't.
I respectfully disagree. We definitely have something.
That there's research into exotic alternatives is fine, but just because they've researched flying cars and fusion reactors for 50 years doesn't mean it will ever matrialize or be usable outside a very narrow niche. If we hit the limits of copper there's no telling if any of these will materialize or just continue to be interesting, but overall uneconomical and impractical to use in consumer products. Like for example supersonic flight, it exists but all commercial passengers go on subsonic flights since the Concorde landed. You can't have exponential growth forever, not even in computers.
Live today, because you never know what tomorrow brings
Moore's law does not predict performance gains, but an increase of the number of transistors per chip. And it's not yet dead. The problem is how to get performance increases out of this transistor count increase, given that frequencies are stagnating and that "sequential" processors have reached their limits.
Moore's Law can be applied to Moore's Law (if you generalise it): the rate at which Moore's Law is declared dead doubles roughly every two years.
No colour or religion ever stopped the bullet from a gun
It's true that we may not see another 90s-style MHz race on our desktops. But there is ongoing need for faster, bigger, better supercomputers and datacenters, and there is technology that can help there. I did quote some examples where this technology is touching the market already. And once it is adopted and refined by the government agencies and big data companies, it will also trickle down into consumer market.
I/O will get much faster. Storage will get much bigger. Computing cores may still become faster or more energy-efficient. New specialized co-processors may become common, for example for NN or QC. Then some of them may get integrated, as it happened to FPUs and GPUs. So the computing will most likely improve in different ways than before, but it is still going to develop fast and remain exciting.
And some technology may stay out of the consumer market, similar to your supersonic flight example, but it will still benefit the society.
Memristors... The whole idea is nonsense at this time. Maybe they will have some future as storage, but not anytime soon.
They're slated to come out with DDR3 of these this year and SSDs next year as 3 major manufactures have been finishing up their final phase of retooling to start mass production. What do you mean "not any time soon"?
Photonics is a dead-end.
IBM showed off a fully functioning prototype of a 2tb/s/fiber integrated photonics with 90nm tech with a 1km range. They said it directly integrates easily into CPUs, and it's CHEAPER than copper. The only issue is retooling board manufactures to manage fiber instead of traces, but IBM says this is still cheaper because the board designs are much less complex.
IBM said to expect these in cell phones in the near future, because it will cut down on cost and power usage while increasing speed.
You may see them, but no actual expert in the field does.
- 3D chips are decades old and have never materialized. They do not really solve the interconnect problem either and come with a host of other unsolved problems.
- Memristors do not enable any new approach to computing, as there are neither many problems that would benefit form this approach, nor tools. The whole idea is nonsense at this time. Maybe they will have some future as storage, but not anytime soon.
- Photonics is a dead-end. Copper is far too good and far too cheap in comparison.
- Spintronics is old and has no real potential for ever working at this time.
- Quantum computing is basically a scam perpetrated by some part of the academic community to get funding. It is not even clear whether it is possible for any meaningful size of problem.
So, no. There really is nothing here.
And all that does is remind me of the times when a previously-useless technology suddendly became mainstream because someone finally found a proper use for it.
I'd like to see a link to the original paper you speak of, AC
Thank you Dave Raggett
bullshit/hype exists everywhere humans are communicating about something...
I think you misunderstood what I meant when I said, "To bring this back to Moore's law"..."this" is my comment...not sure what you thought I was saying...
Thank you Dave Raggett
I looked up some companies by name (too bad you posted as AC and didn't mention them), and here is what I found:
Intel reveals a neuromorphic chip design based on memristors and spintronics
HP and Hynix postpone memristor-based memory to avoid cannibalizing their flash business
This pearl deserves to be quoted:
"In terms of commercialization, we will have something technologically viable by the end of next year. Our partner, Hynix, is a major producer of flash memory, and memristors will cannibalize its existing business by replacing some flash memory with a different technology. So the way we time the introduction of memristors turns out to be important," said Stan Williams, Hewlett-Packard senior fellow and director of the company's cognitive systems laboratory, during a conversation at the Kavli Foundation.
SanDisk and Toshiba are testing a ReRAM (memristor memory) chip
HP working with AMD, Intel, ARM and others to release memristor-based "nanostores".
A working memristor has already been proven in the lab by HP and they are now working with AMD, Intel, ARM and others to release what they call "nanostores". A chip that combines the memristor and logic of the CPU can prove to replace all current microprocessors and memory architectures.
A startup named "Crossbar" will try to beat HP to market with memristor-based ReRAM.
Moore's yawn ... er, law. It has ended, again again. It must be the co-joined twin of Voyager which has left the solar system 78 times in the past 14 years.
Wake me up when some real news gets in.
It took me several moments to realize you weren't talking about Star Trek. I need to go rethink my life.
Well yes, but he was asking how to do it BETTER as well as quicker and in vaster quantity.
When someone says '22 nm', it doesn't resemble at all what one would have assumed it to mean 15 years ago. It means 'roughly equivalent' to what 15 year old technology would do if it hypothetically 22 nm. For example, the slide refers to different structures, and those are already happening (Intel 22 nm is kind of '2 and a half D'). If the argument is 'that won't count', then that statement has come way too late since the node name has already cease to mean anything particularly specific.
A bigger challenge will be whether the phenomenon of computing pretty much being 'good enough' for the mass market will make it unprofitable to do the R&D to advance the state of the art in semiconductors to keep pushing things forward. I know there is significant demand in a lot of places, but will that demand be tempered if that market has to bear the cost of those advancements on their own without the mass market to help justify the cost.
XML is like violence. If it doesn't solve the problem, use more.
GPU is the rational market response to the Xeon premium. DEC would be shipping kilo-socket, kilo-core Alphas, 256 way threaded, with no GPU, and kilo serial memory channels. Think 4096 1 Ghz cores per socket at 100 W.
Hmm... yes good points. A bit off topic, but flying machines too are nonsense. No expert in the field sees them happening. People have been talking about flying machines, and we've had balloons for decades, but flying contraptions didn't materialize. And they don't solve any problem really that we don't already have a solution to, but do introduce new problems, like falling. The whole idea is nonsense. It's a dead end too! Ships and trains are far too cheap to ever let flying machines even be competitive. It's old and has no real potential for ever working. It's basically a scam prepetuated by some bike builders, and it's not clear it will ever be useful for any meaningful problem.
In conclusion, you're right. There's no chance of any revolutionary computing technology coming forward, and there's no chance that humans will ever fly.
Yep. I remember reading articles with exactly this tone (the imminent end of Moore's law) in the technology trades back in the late 1980s.
Everything is just a smaller/faster version of the year before with slight improvements, just like always.
The Kruger Dunning explains most post on
wow, you must be a hoot at parties
The problem with software efficiency has always been this: There are millions of applications, programs, libraries, etc. created, often redundant and amateur. They all run on one of a handful of CPU core designs. Spending the effort to optimise a CPU speeds everything that runs on it. Spending the effort to optimise a program speeds up one program (most libraries, maybe a handful, and only sometimes).
There are still possibilities for CPU improvement. Transfer triggered architectures, dataflow, counterflow, asynchronous, content addressable memory, smart memory - many ideas had promise, but Moore's Law (and incompatability) meant that established techniques improved CPU speeds faster than the new ones could be commercialised (you might remember RISC as the only one that made it, barely). Without Moore's Law, there will be opportunty to work on the alternatives.
The "computing power" trend of doubling is fairly constant throughout history, but just the medium went from marks on the cave wall, to marks in sand, to marks on paper, etc up to silicon wafers.
So "the end of Moore's Law" cry is really quite silly. If we get to the limits of silicon then someone will come up something else.
Thanks, I was feeling lazy. I've had to correct others on this same topic, I was tired, didn't feel like looking the info up again. +1 informative.
The problem with 3D chips, let's assume TSV, as that is the example you listed:
TSV (through silicon via) chips are made on conventional process lithography lines, which have up to 15~30 mask expose steps and hundreds of process steps per layer. The difference is that vias are drilled through the wafer and plated with metal. The scaling of this is basically:
number of steps per device = number of steps per layer * number of layers + wafer bonding + packaging.
When you consider that most of the cost is in those photolithographic steps, and that this cost scales almost linearly* with number of layers, and the defect rate increases with number of layers, it's kind of obvious that this process will not scale to thousands of layers. This is ignoring the enormous problem of heat dissipation in a large stack, not as much a problem with NV memory, but hugely problematic with CPU or DRAM devices.
* as with all mask based photolithographic processes, the cost decreases with unit quantity, as each photomask can cost more than $100k.
I hate to point this out, but we already have AI. Not HAL mind you, but we have cars that drive themselves and neural nets that break captcha's (Google Google's house address images or something). That stuff didn't exist 10 years ago. Look. I shouldn't really care what you say because I know the truth is the business are going to spend the billions because they want to resolve the most expensive cost on their spreadsheets. Labor. Its why you talk to automated voice prompts. They will make AI and it will have your job.
Most importantly, 10x faster hardware makes optimal code faster (ideally 10x, but let's be realistic). Nothing you can do in software can make optimal code more optimal.
Nothing you can do in software can make optimal code more optimal.
As the saying goes ... "There are more than one way to skin a cat" there are more than one method to accomplish a certain task in software.
Just because Mr. A is the world champion in skinning a cat in a particular way doesn't mean Mr. B couldn't be faster by using another method.
http://agtb.wordpress.com/2010/12/23/progress-in-algorithms-beats-moore’s-law/
http://beta.slashdot.org/story/145484
How is the name of any and all is Moore's law dead... The size of the electron is the only thing that can delimit lithography, uh.... we have ,in fact, achieved photon based computers. That means resonance channels that can be carved much smaller than electrical pathway-thus extending Moore's Law. And according to 'Bell Labs' we shouldn't even need to abandon POSIX classical computing... what the Itanium was trued designed for.
Since when did morons in computer science and programming think that digital was all there was, when they undermined the true 60yr road map back to analog computer. SPECTRUM NOT SIGNAL... calculator bandwidth of under heard of proportions... thinking machines... brain and biological augmentation... wicked cool stuff that make your 'puter look like a tablet...wait...
So the answer to my original question is the 'Pentium D' my friends... think about it techogeeks.
Why are there so many articles about Moore's law and how the industry need to speed up or slow down? Moore's law is an observation, Gordon E. Moore, saw a trend back in the 60's. So F-ing what? Why does anyone care? So what if the processor speed is "on pace" with the law or not?