Moore's Law is dead by 2022. Lots of implications for Bitcoin, but let's focus on general business.
How to profit from the death of Moore's Law? Short 10x Intel?
Moore's Law is dead by 2022. Lots of implications for Bitcoin, but let's focus on general business
Nathan Wood
Other urls found in this thread:
en.wikipedia.org
twitter.com
Carson James
>not preparing for Quantum Computing
Moore's law is about to be superceeded.
Landon Barnes
Its a literal double top yo. Better short all tech 20x now before the hedgefunds do it for you. Wassup, hickachicka.
Hudson Smith
this graph goes to 2010
John Murphy
Stop posting outdated graphs faggot
Nathan Nelson
Where's the fresh data?
Nathan Garcia
Chances are that while silicon's semiconducting limits are reaching by miniaturization, chances are that increasing supply line efficiency and production capabilities will continue to halve the price. As production quality increases, clock speeds will be able to increase as the "silicon valley lottery" of chip quality tends towards higher and higher quality nanolithography and construction, which of course determines the maximum clock speed before the chip starts making errors. That will allow more power to be wrung out of existing chips at the semiconducting limits of silicon by allowing faster chips made with existing architectures. What basic moore's law analysis--- which focuses on the limits of silicon semiconductors--- ignores is ever-increasing efficiency in the manufacturing and supply line side of things, which is still very much under way and can yield cheaper and cheaper prices for existing chips, which leads to my next point.
That will be compounded by intel wringing as much efficiency out of the working of existing chips as possible, which is getting better and better. Think of how the invention of hyperthreading increased the capabilities of chips without changing the architecture significantly. And as higher quality, more efficient chips are produced at the same or lower price as previous generations, and running at higher clock speed, the next frontier is simply increasing core count, CPU count, and parallel processing capabilities. There is so much we can't foresee that will prolong moore's law.
Lincoln Walker
>How to profit from the death of Moore's Law?
Buy bitcoin , miners dump most of the supply to upgrade asics , once they are not forced to do that anymore bitcoin moons stronger than in a halving event.
Jose Russell
This. We're in the dead cat bounce. Exit all CPUs. This is not a drill
John Collins
Crazy stuff like 3D silicon with internal pipes for liquid cooling should create a small bull run in those MIPS per clock cycle graphs.
Today designs are limited to be flat planes with a few layers.
Benjamin Ramirez
Parallelization will continue until ai breakthroughs in manufacturing/theory.
The moore's law is dead meme is about decomputerizing the population and destroying the free market to prevent people from having access to compute.
Jayden Sullivan
This thread deserves a bump.
Did a science degree 15 odd years ago, i was obsessed with nanotechnology back then, it seems we don't hear too much about it these days apart from cheap marketing scams about surface coatings. As OP and others point out, the process size for silicon chips has progressively gotten smaller and smaller
see for example: en.wikipedia.org
When you start getting to the 5nm process diameter at the end of Intel's tick-tock(-refresh) roadmap, you're talking about 25 atoms thick.
I'm not an engineer, but it was my understanding we practically needed a new era of manufacturing tools and materials to advance beyond this (just as we moved from stone tools to bronze tools to iron etc). It's not just a limit in size but in the practicality of the semiconductors we currently use. Do we move onto superconducting carbon nanotubes? Will the necessary tech be seeded, or about to be seeded?
Pic related was a 1989 demonstration by IBM scientists who manipulated 35 xenon atoms to write the company logo using a scanning tunneling microscope.
Eli Evans
Invest in ternary
Thomas Anderson
>moore's law dies
>we enter singularity
Xavier Watson
nice post
Evan Rivera
Long amd. Their chiplets will conquer the whole server market and Intel has nothing to hold against it.
Easton Davis
lol you don't know anything about quantum computing
Kayden Gutierrez
Based and redpilled
Isaac Thompson
This. Clock speeds haven't been important since 2006.
Charles Stewart
you think you can bury it just yet? sapphire based semiconductor technology has barely been touched yet it has a potential to add a few magnitudes to capacity.
Luke Collins
Based and knowledgeable
Charles Nguyen
the death of moores law is extremely bearish for crypto in general.
the tps of decentralized networks depends on their physical hardware.
Nathaniel Watson
Do you think Sergey masturbates to memes of himself? I would.
Jayden Rodriguez
no crypto network today is limited by the underlying infrastructure in the slightest
Gabriel Perry
what is the workload is the proper question instead of puffing pipes from an armchair in a spiral of smoke
Aiden Reyes
Don't short intel faggot. The guy who made AMD relevant again is leading intel's next chip design. He's not one to talk out of his ass too. I'm selling my AMD shares in a year or so before the chip gets announced.
Brandon Lopez
It's a new paradigm.