Moore’s Law hits physics in memory chips – Reuters.com: Reuters delivers an interesting article about the upcoming challenges that memory manufacturers will have to face soon.
Makers of memory chips are looking ahead to a day, not too far off, when technology based on silicon bumps up against the laws of physics and memory can’t be made any smaller, with implications for gadgets like MP3 players and digital cameras.
“You get in to the 25-nanometer regime and there may need to be a new structure for nonvolatile memory,” Mike Splinter, chief executive of Applied Materials Inc., “I’m quite nervous about this because 25 nanometers is not that far away, and if you have to change a process in a couple generations, then that is really challenging”, that would slow the development of things like digital music players and cameras, for which current flash memory — used to store music and images — will not suffice beyond the next couple of years.
Such concerns aren’t far from the mind of Tom Trill, marketing director of Samsung Electronics, “It’s a question we’ve had forever, and we’ve always had an answer… there’s been a resurgence in terms of pessimism … in the last few months.”
“Every two years someone comes up and says they have found better memory technology, but there’s always some technical limitation, and this has gone on for 30 years,” said analyst Jim McGregor from In-Stat.
Two words… Carbon Nanotubes.
I’ve been reading about this looming wall for years. Logically it has to happen eventually but so far the reports of Moore’s Law being dead have been premature.
Anyhow, when we do finally see the next generation of memory, it might not be based on silicon at all. There is some fascinating research into alternative approaches now; journals like Science Daily, New Scientist, Science News and not doubt others cover the subject frequently.
Still, I’d expect both Samsung and the analysts at In-Stat to be on top of this subject. Maybe the wall is really there this time. Where’s quantium/nano/biological/crystal/etc. data storage when we need it?
Moores law never made sence over the long term.
The maximum times you can fold a piece of paper in half is no more than 12 times no matter the size of the paper
A hundred years ago, the known laws of physics wouldn’t have allowed most of the technology we have today. It’s only a matter of time before new scientific discoveries propel us forward another hundred years.
Well, so what? So miniaturization stops at a certain point. What we’ll have will be more than good enough for what we need to do with it.
Maybe we can start a new paradigm: buying equipment that is built well and sturdy, and using it for 25 years, like we used to do with TV’s. Seriously. Time to stop filling the landfills with toxic junk just because the news stuff is twice as capacious or fast. Improve the quality of the stuff we make so that we don’t have to throw it away every year or two. That’s part and parcel of the “Amishization” of engineering that needs to start, now, to switch over to a maintainable impact on the life of the planet.
When we finally hit that brick wall, architecture will stabilize. This may actually be of some benefit to the consumer:
— We won’t feel the need to upgrade to the latest and greatest. That two-year old home computer will still be quite capable of playing the latest games.
— Software competition may actually return. It’s hard to compete with MS when you are also running against time. In today’s market, it is damn near impossible for a dozen software engineers to throw together a product that isn’t obsolete by the time it makes it to the store shelves.
On the downside (for the US):
— Competition for hardware sales will be driven solely by production costs. Intel may lose a lot of its market share to companies that can bypass the cost of R&D. China, for instance, could easily take advantage of, say, an eight-year hiatus of Moore’s Law to drive Intel and AMD completely out of the market.
According to intel.com, Moore’s law is “In 1965, Intel co-founder Gordon Moore saw the future. His prediction, now popularly known as Moore’s Law, states that the number of transistors on a chip doubles about every two years.”
If you can’t fit more transistors into the same space (because of limits of the laws of physics) you just make the chip bigger. Anyone notice that a Core2Duo chip is quite bigger than an 8088 ?
Of course, if you double the chip size, after 16 iterations the chip 327.68 meters square. D’oh.
You guys are missing the whole point of the article – it’s hitting a wall on conventional memory design.
The good news is that conventional memory design (trapping a charge in a capacitor) is being replaced by things like polymer memory and new materials that do not trap a charge and not capacitor based.
Dvorak: Please refrain from posting technical points here. The audience here are dumb ass tech wannabees that really want to talk about politics and gay marriage,
#6 &7 And an entire industry implodes because they aren’t selling anything. I’m not saying you aren’t right just what the outcome will be.
I don’t see how the ascertion that this is going to spell disaster for MP3 players and digital cameras can be true. Are you telling me that digital camera technology is going to move to storing photos that are 100 Meg any time soon?
I can get 4 Gig CF cards that hold nearly 1000 photos for my digital camera; Considering Digital cameras are an upgrade from a technology that allowed me to take a lousy 36 shots at a time, I think it will be a long time before current memory technology isn’t sufficient.
I do like the idea of some technology stability though; It’ll never happen though 🙂
#9 – Dvorak: Please refrain from posting technical points here. The audience here are dumb ass tech wannabees that really want to talk about politics and gay marriage,
Imagine that… people can talk about other stuff than technical mambo-jambo… And btw… Of what I know, this site was not intended for one-subject minds like yourself.
Cheers.
#6, #7
I agree with you in general. I wish quality was on people’s minds.
My take on Moore’s law is that it is / was a self fulfilling prophesy. He came up with it because they needed a predictable road map for planning. Since Intel was doing the chips, they just needed to follow Moore’s law to be doing just fine.
What about Gallium Arsenide chips?
Ultimately, Moore cannot continue. In a mere 6000 years or so, all the Universe’s sub-atomic particles will be needed to continue its inevitable prediction. As for now? Clever minds are still mulling the issue. Stay tuned.
Funny how people only think in two dimensions.
When you add Z to the math, commonly known as layers, this explains why 16 iterations from 1965 hasn’t resulted in my Core Duo laptop being the size of my fridge.
What about the implication of what AMD is doing for four independent cores on a single die, whereas Intel simply cheated putting a tiny gate/controller MDM chip to talk to two separate Core Duo’s?
Intel “beats” AMD to four CPU’s inside one GPU by a few months, but the non-tech press don’t realize it’s simply two Core Duo’s sharing the same space.
AMD’s approach will a significant milestone for multicores. This is where Moore’s law is starting a new trend. He never stated the size of the chip, and there’s lots of room to expand yet.
8 gig SD’s from SanDisk… Yea. A HD video camera without tape.
Go technology, go…
very
very
nice.
thanksss