When Moore’s Law reaches a limit, will rapid advancements stop or will new methods of computing set the industry on breakneck development speeds?

[KelleyNice]

Moore's law is the observation that the number of transistors in a dense integrated circuit (IC) doubles about every two years. Since a transistor cannot be smaller than an atom, the transistor is approaching the point where it is simply as small as we can ever make it and have it still function.

Updates

+1 y

Moore's Law can be extended several generations by making the chips 3 dimensional instead of 2 dimensional. That is, chips with 2 layers, then 4 layers, then 8 layers, until there is to much heat.

[RavVid]

There have been a few points where it has stalled and became harder to go smaller which has led to a few companies dominating. TSMC for making chips and ASML for lithography gear essential for etching. Building transistors has already gone vertical quite some time back.

The nM measures have stop being sensible and more what the company choses to rate them as. 7 nM is commercial and IBM is claiming 2 nM and describes it as narrower than a strand of DNA. Cosmic rays can already cause unexpected results and will with quantum computers also. Mainframes have processors pairs and vary off line if they don't agree.

Quantum computing and material other than silicon have already been mentioned so I won't. There is not much point going more than 64 bits.

I tend to think that we aren't that good at using what we have.

I make a good living tuning software, getting 90%, 95% improvements is a matter of course because most software is written badly.

Most laptops would have dual cores or quad cores and can potentially do 2 or 4 things at once but mostly don't. My 4 cores are under utilized at the moment as I write this.

Which brings up the concept of good enough computing which for what you and I do, we reached many years ago. A faster processor is not going to us much good on GAG or emails or web browsing.

Parallel computing was a big thing quite some time ago but we just aren't good at it and I don't see commercial software written that way. We look instead to the operating system to do that. Processors help in that respect too.

Contrast that with my brain walking down the street. I am considering my finances, double checking what I definitely need to buy today, compromising on what I will buy today, evaluating potential threats from homeless people and all the time scanning for cute girls. It would be challenging to write that software. Do-able but not easy.

[abc3643]

Moore's Law is already beginning to fail.

But, what will eventually happen - likely in your lifetime, but not mine - is that quantum computing will reach capabilities that will be commercialized. That will revolutionize computing and the world.

For instance, right now, "crypto" is bullshit. There's no real "crypto" because they can track everything. If you want to make an untraceable purchase, use cash.

BUT, with quantum computing, crypto may become viable.

Think of it like this: If the authorities were hunting for an illegal crypto transaction, it's like looking for a needle in a haystack. With quantum computing, it might become like looking for 1 needle that may exist among all the haystacks that have ever existed throughout the known universe.

Then again, the authorities will have quantum computers too so it is like they will have zillions of hunters too.

So, who knows?

[abc3643]

TITLE: No more transistors: The end of Moore’s law
SUBTITLE: We can’t make transistors any smaller, is this the end of Moore’s Law?
By John Loeffler
Feb 13, 2022

There has been a lot of talk about the end of Moore’s Law for at least a decade now and what kind of implications this will have on modern society.
...
In the end, Moore’s Law was never a "law" to begin with but more of a self-fulfilling aspiration. We expected transistor density to double every year, then every two years, and so we looked for how we could accomplish that task.

Whatever the next thing is, whether it’s quantum computing, machine learning and artificial intelligence, or even something we don’t even have a name for yet, we’ll find a new aspiration to drive that innovation forward.

Ultimately, our fascination with Moore’s Law was never really about the density of the transistors. Most people who’ve heard of Moore’s Law couldn’t even begin to explain what transistor densities even mean, much less how interlocking transistors form logical circuits or how the smartphone in their pocket works (or even a 1970s pocket calculator, for that matter). For most of us, Moore’s Law was always about our expectations of progress, and that is something that is largely up to us.

Moore’s Law might be on its last legs, but we’ll find a new Moore’s Law if we want it badly enough.
interestingengineering.com/transistors-moores-law

[abc3643]

RESPONSE TO UPDATE:
Something like that was suggested in a talk I went to at RPI in 1991.
That may not be economically worthwhile though.

[abc3643]

The talk was given by RPI then-Provost the late James B. Meindl.
news.stanford.edu/.../

[KelleyNice]

@abc3643 What was not economically worthwhile 30 years ago might be today. Even so, 3d transistors also has limits.

[abc3643]

In this case, it's the opposite. The economics were easier for advancement, but, now, chips are so dense, the power they consume is releasing a huge amount of heat. I've been in rooms with banks of computers and test rigs for software and electronics. We had to have a 5-ft diameter fan on the floor just to remove the heat. Companies like data warehouses with oceans of machines are reaching the point where they can't afford to cool these machines anymore. So, even if chips became denser, people aren't going to want them if they can't afford to use them.

This might give you some insight.
www.techtarget.com/.../Liquid-cooling-vs-air-cooling-in-the-data-center

[KelleyNice]

@abc3643 Oh! Okay!

[007kingifrit]

moor's law hit it's limit years a go becasue of a part in a computer called a transistor which has already reached 1 atom in size. you can't go smaller than an atom without losing chemical properties so the computer as a whole isn't getting much smaller.

that being said there are plenty of new advancements to make, synthetic diamonds are becoming cheap enough to replace the quartz in computers which is 17x slower than the semicondutive properties of diamond

and i didn't even mention quantum computing. plenty of development to come

[abc3643]

I was attending a talk given about the limits of Moore's law in 1991.
What was interesting is that, the components were getting so small that soon (back in 1991) they'd have to worry about the magnetic fields generated in one nanowire in a chip affecting the nearby nanowires in the chip. Also, though, they had suggested the possibility of layering - instead of chips being flat, they'd become more like blocks like a stack of chips. In other words, 3-D thinking instead of 2-D.

[MikeInHawaii]

Who says Moore’s Law has a limit? In the 1890’s, some politicians wanted to abolish the Patent Office thinking everything imaginable had already been invented.

[Adaeva]

I suppose, one way or another, computers will keep advancing.

[Sixgunsound]

Quantum processing tech will replace transistor based tech

[MrHopper]

Its not really the same as it was back when the graph was made, things have altered in how we measure processing power, its certainly not exponential any more but is increasing

[IsoUser]

Moores law forgot that technology doubles every six months

[IsoUser]

With all things being considered, when my wife is not here via wifi or otherwise my Artificial Intelligence Robots take on life of their own it's just a dominant feature

[IsoUser]

Once again, your being jealous when you have no reason to be. I have made the same AI available to women

[exitseven]

Most digital circuits just need a way to register two states, on and off. There are scientists now making these circuits out of different materials so that they can be smaller.

[DWornock]

It may stop for a while until new and improved technology is discovered.

[TallAnon]

Nah, we're already busy with the next step, quantum computing

[Anonymous]

I can assure you that remotely controlled vibrators are already here through bluetooth same as for men the only difference is that I actually cum twice insi3

[jahaims]

Quantum computing will be the next step