The self-assembling computer chips of the future | Karl Skjonnemand

  • Published on Mar 13, 2019
  • The transistors that power the phone in your pocket are unimaginably small: you can fit more than 3,000 of them across the width of a human hair. But to keep up with innovations in fields like facial recognition and augmented reality, we need to pack even more computing power into our computer chips -- and we're running out of space. In this forward-thinking talk, technology developer Karl Skjonnemand introduces a radically new way to create chips. "This could be the dawn of a new era of molecular manufacturing," Skjonnemand says.
    Check out more TED Talks:
    The TED Talks channel features the best talks and performances from the TED Conference, where the world's leading thinkers and doers give the talk of their lives in 18 minutes (or less). Look for talks on Technology, Entertainment and Design -- plus science, business, global issues, the arts and more.
    Follow TED on Twitter:
    Like TED on Facebook:
    Subscribe to our channel:
  • Science & TechnologyScience & Technology

Comments • 209

  • foldr
    foldr 19 days ago

    There's plenty of room at the bottom.

  • Sujur Ramalingam
    Sujur Ramalingam Month ago

    Extraordinary, well presented 👏👏👏

  • Shincy Rajith
    Shincy Rajith 2 months ago

    What happens when two different layer needed to be connected they will interact with each other and cause disruption

  • The Real LA Reid
    The Real LA Reid 2 months ago

    This technology is already in use. I have a self assembling chip in the left side of my brain and illegal testing is being conducted on my body and brain. I was poisoned by Charlie Evans of Oakland, CA with Smart Dust from 8/2013 - December 2016, after he stole my DNA. On August 6 , 2017 while talking on the phone a piercing blast was sent though my phone and into my left ear. I have suffered with tinnitus since. Now the left side of my head is enlarged. I am monitored and tracked 24/7. These illegal non-consenting experiments are being conducted all across the world but mostly be my previous employer Univiersity of California, here in California. They pick people who they feel lives is of no value ( women and minorities). In addition to the non-consenting experiments they have Community Organized Policing in place to harass us to keep the focus off what is really going on. i have nano piezoelectric crystals in my ears, nano charged film in my eyes, and my eyes now glow in the dark under UV light. In addition they have given bottom-feeder criminals Ronnica Evans of Hayward, CA, Lorena (Rena) Wallace of Oakland, CA, Marjorie Edmond of Oakland, CA and their Organized Crime Ring, all access to watch me through my eyes and monitor me via Remote Neural Monitoring and viewing. Yes these people are hacking bodies. It all sounds so unbelievable! I know what I am saying many will think I am delusional. However, I wish I was delusional so I could get medication and this nightmare would be over. I am speaking my truth and one day soon, everyone will see and know, that I am telling the God's honest truth. Oh, I have also had murder attempts on my life for speaking the truth. I am 26 years clean and sober, no criminal record and I move to a city to try and escape the technology and harassment and it continues. Complete strangers trying to murder a 58 year old woman for no reason is not normal. This entire New World Order is nothing but a depopulation agenda. Chip the human brain, hit the 5G switch, fry us and send in FEMA with their portable crematorium. Here is video proof of the technology in my body and the murder attempt on my life.

  • divinefallfromgrace
    divinefallfromgrace 2 months ago

    If we can use empty DNA strands to store vast amounts of data, is there no way we could apply this to transistors also?

  • Svetlin Totev
    Svetlin Totev 2 months ago +1

    This is flawed on so many levels.....
    First of all the main problem with the miniaturisation of transistors is a quantum mechanical problem that can simply be described as the probability of the transistor giving the wrong answer or the voltage required to make it give the correct answer is just impractical.
    Then there is the problem that processors or other silicon components are not made of repeating patterns. They are made of MOSTLY repeating patterns that are occasionally broken. The complexity of editing the perfectly repeating patterns to make the mostly repeating patterns is way too high to be worth doing. It is so much easier to literally print the diagrams of the chips.
    The cost of printing chips is not the fact that the initial cost is high. In fact even the maintenece of the machines insn't that high. It's the cost of the production methods that is so high.
    As you described here the method you suggest has many of the complexities of the printing method like requiring ridiculously pure substances which is way easier to achieve when you are filtering out one type of atom rather than filtering a complex molecule.
    No mather what method you use the information of the final pattern of the chip needs to somehow get to the waffer so you can't avoid printing in any way. You can change the method of printing but you can't avoid the need for it.
    A breakthrough in computer technology would come from developing cheaper 3D chip manufacturing or atomic level logic gates and not from fantacising about using a basic manufacturing method for making basic shapes and using it to make a complex shape. The information defining the chip needs to come from somewhere...

  • Say my name? Harambe
    Say my name? Harambe 2 months ago

    Did anyone thought if this can happen, the major problem will be cancer molecules.

  • Secret Recipe
    Secret Recipe 2 months ago

    My guess is we will get fentom computing for daily use in the household before quantum computing

  • Jershna
    Jershna 2 months ago

    Yea, this would help make smaller transistors. But, what about the limiting effects of quantum tunneling?

  • A Day In The Life
    A Day In The Life 2 months ago

    Inefficient hardware and software design won't be solved with quantum computing. More layers, bigger chips, parallel architecture, lower level language programming. A bigger closet just collects more shoes you don't ware.

  • Xi Xu
    Xi Xu 2 months ago

    If it helps me download p*rn faster, I'm in.

  • Franklin R. Futch Jr.
    Franklin R. Futch Jr. 2 months ago

    Someone Get This Guy An Award Quick!

  • Pasta Farii
    Pasta Farii 2 months ago

    must be cold there, i can see his nipples

  • Allucardsin
    Allucardsin 2 months ago

    This is a good idea and if this can be combined with additive manufacturing it can be even better.

  • Damon Blade
    Damon Blade 2 months ago

    So basically this guy is trying to create a computer that can pysicaly grow and mentally on its own.
    Like those machines from battlestar gallactica ? Well I for one welcome our new machine overlords !!

  • bert havermout
    bert havermout 2 months ago

    Computers inside your body "how cool is that". "That we all enjoy today". Speak for yourself!!! Mr, Carl, We may enjoy having a Mobile phone but we also may enjoy having a joint.

  • Taylor Mann
    Taylor Mann 2 months ago

    I wonder if sound could be used to make the nano partucals self a desired pattern

    • Taylor Mann
      Taylor Mann 2 months ago

      @vroomik nice....

    • vroomik
      vroomik 2 months ago

      it's in the comments... It's called cymatics, look at Hans Jenny work. We just need more research

  • Robert Galletta
    Robert Galletta 2 months ago


  • JustMe
    JustMe 2 months ago

    When physical limits reached, then need go for mental limitations ;D :)

  • W Something
    W Something 2 months ago

    Disappointed he didn’t go on to explain how the transistors are attached to, or work with these molecular chains / structures.

  • Left TechnoLibertarian Party of YouTube

    Hardware has always been a limitation on software. It's not new. And the challenge with smaller transistors is not manufacturing cost. It is quantum effects which make transistors unworkable.
    Progress in hardware requires new kinds of chips, not new ways to make them!

  • finlanderxx
    finlanderxx 2 months ago +1

    I thought this talk was about self assembling paperclip replicators which would eat all material in the world and in the universe.

    • W Something
      W Something 2 months ago +1

      finlanderxx that’s the end result of this work 😏

  • Mark Jacobson
    Mark Jacobson 2 months ago

    he says implanted computers are "cool" and that AI software which is a foreign form of "intelligence" and which is entirely devoid of thinking and compassion, is the way for humans to go. that makes him the enemy.

    • Pugpugpugs
      Pugpugpugs 2 months ago

      implanted computers are cool

  • SquillyMon
    SquillyMon 2 months ago +3

    Using queue cards instead of a tablet to describe the miracles of miniaturization of transistors and computing power... lol

  • Jay alva
    Jay alva 2 months ago

    Just so Huawei can go in and steal all this hardwork 😒😒😒😒

  • Joe Delisle
    Joe Delisle 2 months ago


  • Shen Zhou
    Shen Zhou 3 months ago

    The speaker should learn news about DNA self assembly. DNA molecules cam be programmed to any 3d nano structures desirable.

    • W Something
      W Something 2 months ago

      Shen Zhou The issue is probably with consistency and Percision.

  • jose rizal
    jose rizal 3 months ago

    you guys do the hard work and china just do the easy work, copy or steal it! hehehehehehe

  • Rjk
    Rjk 3 months ago

    Hardware got eaten up by Software? My CPU and RAM are constantly on "bored" status. And I'm working with average stuff... The only thing I experience is, that LTE and connectivity are laggy 🤷‍♂️

    • Rjk
      Rjk 3 months ago

      Btw. Transistor development stuck, because we're reaching physical limits...see Nvidias Product presentation in 2017.

  • Rylan W
    Rylan W 3 months ago

    If you have a cross section and a beam combined with a very high frequency and a method of aiming it to the precision of hitting a specific atom. Then it could be possible to fuse manufacture stuff atom by atom.

  • DeePal072
    DeePal072 3 months ago +1

    5:28 and then came Butlerian Jihad ...

  • wassollderscheiss33
    wassollderscheiss33 3 months ago +1

    He didn't get it. Sorry. The truth is, the hardware is just fine for the most part. But software developers take the power of fast hardware as a given and just don't spend much effort on efficency. Actually most computer software today is implemented rather stupid, reducing time to market by not optimizing code. If there was a shortage in hardware performance, software could rather easily be made much faster by implementations that considered performance. There are obvious exceptions like games and video editing and compilers for example. But even there you'll find the trend I mentioned. For example there are more and more games that work directly in the browser, although a browser is a much much less suitable platform for that kind of software than your original operating system is.

  • epSos Premium
    epSos Premium 3 months ago +4

    Beautiful idea. Let's hope it works out.
    We could have replicator technology, if materials assemble in useful forms for us.
    Better tools and better things will come from this magic !

  • Leonidas666
    Leonidas666 3 months ago +2

    Ah Science... Finally Ted.

    • Jimmy Bell
      Jimmy Bell 23 days ago

      Get it and you will be separated from God forever.
      I think you know this already tho.
      Repent in Jesus Name.
      Hells not a party.

  • 4G12
    4G12 3 months ago

    We already have self assembling computers. They're called living organic brains.

  • BurnGiordano
    BurnGiordano 3 months ago +1

    Lower cost and higher efficiency, across how much tech, but never vehicles.. Thats a lot of time wasted.

  • Sonja Brajkina
    Sonja Brajkina 3 months ago +11

    This talk is like startup presentation.

    • anony mous
      anony mous 2 months ago

      Sounds similar to an iPhone presentation lol.
      "I present to you, the iChip"

  • tom7865
    tom7865 3 months ago

    Pitching for investment through TED!?

    • Skull Kid
      Skull Kid 3 months ago

      It's about public awareness, drawing interest, and hopeful endeavors.
      Nothing wrong with that, gotta get the money to fund these projects somehow.

    • Vitor Oliveira
      Vitor Oliveira 3 months ago

      I aways though this was the whole purpose of TED

  • AlphaOmega
    AlphaOmega 3 months ago +1

    sheer insanity

  • Erik Thiart
    Erik Thiart 3 months ago

    Bit of a boring video stating age old facts.

  • Pat Trainor
    Pat Trainor 3 months ago +11

    Wait a second, Karl Skjonnemand. I thought the idea was to allow for imperfections by having multiple, simultaneous/parallel 'cips' or whatever, that are checked against each other. IOW, fault tolerant by design, the kind of design that accepts that there are faults and justs moves on.
    I cannot see a viable future where perfection is the requirement, and imperfection is failure. Humans-and nature-don't work that way. And if you are going to emulate nature, you may as well learn ALL the lessons it is offering you.

  • tester moving
    tester moving 3 months ago

    I am from India. All my student friends are just got to know there is somethings called Quantum computer. We want the world stop spinning for a while.

  • Vasile Sulica
    Vasile Sulica 3 months ago

    I read a article about this in 2004 in scientific American is not about self asambing chip but self asambing mask for graving the chip

  • mmisbach
    mmisbach 3 months ago +1

    Rate of miniaturizing transistors is slowing down, and not keeping up with the rate of software innovation. Complexity of the manufacturing process is the problem. Molecular engineering will help us break-through by using self assembling materials instead of expensive machines. These nano-structure transistors can be manufactured using chemical engineering. #TedTalk

  • Josh Gunn
    Josh Gunn 3 months ago +4

    I bet they're lining up the transistors
    Using cymatics

  • gut rot
    gut rot 3 months ago

    Not good for the human body

  • Different
    Different 3 months ago +1

    nature is insanely complex, chaos everywhere and errors, millions of errors.. And through the chaos comes order, through the errors and successes the total is a structured "thing".
    Meanwhile we already got order in our fabrication processes.
    good luck creating perfect order using molecular engineering and chemistry to create perfect order.

  • Baga Inferno
    Baga Inferno 3 months ago


  • alok Jha
    alok Jha 3 months ago

    jobless creation 😂

  • Rickbearcat
    Rickbearcat 3 months ago +1

    I'm going to assume here that the production of these self-assembling structures is the easy part. The hard part or bottleneck is getting them to assemble properly on a substrate suitable for conducting electricity and signals through it. Also known as the PCB, which is the interface that connects to the motherboard of a conventional PC.

  • Evren Bekdaş
    Evren Bekdaş 3 months ago


  • DjSapsan
    DjSapsan 3 months ago +1

    Speaker moving like a robot

  • maze case
    maze case 3 months ago

    I still dont understand anything

  • maze case
    maze case 3 months ago

    Distance =periodicity

  • maze case
    maze case 3 months ago

    Block po lymer chains hate each other , but we bond them to build frustration. It squirms to form a shape this shape is ina nanoscale.

  • maze case
    maze case 3 months ago

    Summery:- transistors got smaller, but helped us compute ! Transistors are not getting smaller through time. Our hardware could limit our development ! The semiconductor industry is working on quantum computers to robust and efficiancy ! Were lookimg for immediate solution . The complexity of the transistor after development is slowing the process down. We can put smaller transistors in chips . Everything is built on the silicon wafer . The process is improving. Molecular engineering and mimicking !

  • Alpha Strength
    Alpha Strength 3 months ago

    Resourses are extremely important

  • RAV3N
    RAV3N 3 months ago

    Wonderful manufacturing ideas! There is still the quantum problem with increasingly miniaturized transistors. Quantum tunneling is still a problem that needs a solution. Transistors need a breakthrough in design or need a replacement technology. That may spell out the end of manufacturing transistors in this method before it begins or soon after. That being said, I am certain your self assembly technology will be very useful still.

  • Morten Brodersen
    Morten Brodersen 3 months ago

    Those bad "hugry software engineers" hahahaha :D

    • Aj195
      Aj195 3 months ago

      You're one?

  • Lamborghini Husky
    Lamborghini Husky 3 months ago

    Terminator anyone? Something being able to assemble it self is never a good idea

    • Feroce
      Feroce 3 months ago +1

      It doesn't have a brain, so it can only make single predetermined shapes and can't replicate itself. It's harmless.

  • Alex M
    Alex M 3 months ago +1

    8:16 if different are trying to separate, why that on screen is not walling apart? totally the guy leaves important details IMHO

    • Edwin Wiles
      Edwin Wiles 3 months ago

      @Alex M it means he's either lying or seriously mistaken. "He's full of s--t". Just as I could be seriously mistaken in my theory of how what he's actually doing works.

    • Alex M
      Alex M 3 months ago +1

      @Edwin Wiles thanx, 1st part is clear, end of second is not - what means "he's full of it" - he, it?

    • Edwin Wiles
      Edwin Wiles 3 months ago +1

      You can design a very small molecule that is capable of attaching to each of the other materials. But depending on how you design it, it will allow different degrees of freedom to the two repelling chains. This is pretty much how detergents work. There's one side that is hydrophilic (drawn to water, repelled by 'dirt'), and the other side is hydrophobic (repelled by water, but drawn to 'dirt'). That's a pretty horrible simplification, but it works. It allows the detergent to surround the 'dirt', bind to it, and separate it from the material you're trying to clean, because the outside is now strongly drawn to water.

      That's one of the things that bothers me about the images he showed. There should have been 'bulges' or obvious patterns of different materials in the self assembled examples. Sort of like dashed lines across the linear patterns, and rings of different material around the radial patterns. So either my idea is full of it, his images are not of sufficient magnification to show it, or he's full of it. Heck if I know which is right.

  • Krishna Kumar Gupta
    Krishna Kumar Gupta 3 months ago


  • General Tech
    General Tech 3 months ago

    wow amazing

  • Xander Aibas
    Xander Aibas 3 months ago

    Not cool.

  • MTB Thru my eyez
    MTB Thru my eyez 3 months ago

    No disassemble. disassemble!

    • god damn
      god damn 3 months ago

      Short circuit lol

  • Iaman Empoweredone
    Iaman Empoweredone 3 months ago

    This sounds a lot like the behavior of Morgellons disease. Self replicating nano stuff

  • David Durant
    David Durant 3 months ago

    Eric Drexler would be proud.

      ABDICOLE STUDIOS 3 months ago

      David Durant along with Marvin minsky and fineman

  • Skandar
    Skandar 3 months ago

    What’s the benefits of it???????

    • Skandar
      Skandar 3 months ago

      Sebastian Wiesendahl, the future is quantum, I don’t think this will ever come close to quantum computing

    • Sebastian Wiesendahl
      Sebastian Wiesendahl 3 months ago

      Like he said: cheaper to produce. The material is outlining itself, like a compass needle folllows the magnetig field.

  • Guilherme Santos
    Guilherme Santos 3 months ago +2

    How Cool is that
    Surely is a Crazy Russian Hacker reference

  • muhammad amir
    muhammad amir 3 months ago

    Very nice lecture.

  • Rykahnz
    Rykahnz 3 months ago +2

    So what I get out of this is that it can really take a knack out of the manufacturing cost but as far as further miniaturization goes there isn't much room regardless because of things like Quantum tunnelling. Also what do polymers share with boron silicon and phosphorus what I mean is how is the state change achieved.

  • elite AJ
    elite AJ 3 months ago +5

    Holy crap technology is flipping amazing damn.

  • Iñaki Gomez Mayol
    Iñaki Gomez Mayol 3 months ago

    Truly amazing. I love it

  • S Tristan
    S Tristan 3 months ago +1

    But let’s spend billions on a wall.

    • Remy Lebeau
      Remy Lebeau 3 months ago

      No thanks, we would rather hurt the economy by about 11 billion because we shutdown the gov't to bicker over the wall like petulant children.

  • zeramino
    zeramino 3 months ago +2

    There's only so much smaller we can get before we hit quantum uncertainty, quantum computers are a better long-term solution for this problem.

    • zeramino
      zeramino 3 months ago

      @Sebastian Wiesendahl yes, QCs are meant to churn data for now, they're not designed for day-to-day computations, but who knows, in the future we might have a really powerful central computer and less powerful machines connected to it.

    • Sebastian Wiesendahl
      Sebastian Wiesendahl 3 months ago +1

      It depends on the task. Quantum computers are not better at streaming for example or any asynchonous task in general.

  • Scarletpooky
    Scarletpooky 3 months ago

    Do you want Terminators? This is how you get Terminators.

  • skinz
    skinz 3 months ago +3

    technology is drip fed to the people they are always 10 to 20 years ahead of what they sell to us nothing we have is new it has the appearance of being new because technically it is new to market

    • Xi Xu
      Xi Xu 2 months ago

      Do you want to use products in the research phase or finished products?

    • nagualdesign
      nagualdesign 3 months ago

      @ipodgamer99 Exactly.

    • nagualdesign
      nagualdesign 3 months ago +3

      "Drip fed"?! You make it sound like a conspiracy. In truth, things simply take time to develop to the point of being a mass-produced product ready for market. The fact that you or I may not be privy to the entire development process is neither here nor there.

    • ipodgamer99
      ipodgamer99 3 months ago +2

      I hope youre not trying to say thats because some elites are holding it back. Before products can be made in great numbers with profit, a lot of optimization has to happen

  • Adymn Sani
    Adymn Sani 3 months ago

    why do we really need smaller to make it better? For instance in a desktop or laptop there isn't any terrible reason that is good enough to say a CPU package couldn't be 4X current size thus quadrupling CPU transistors. In fact if the chip package and thus the architecture is larger heat wouldn't be a bigger issue. The only issue I think might stand is the desire for increased speed, but we already have pretty much all the speed we need without going to quantum computing. Sure I would love to have a computer that turns on and boots up in a second, but I built one years ago that did it in 10 seconds. I have hand built many desktops and there is plenty of room on the motherboard for a larger cpu chip package. I don't see the issue since we have been building smaller just to get more: give the chip more room to get more transistors. I don't really think that a cell phone needs to be much more powerful than now.

    • [Misery] RaphielOhneF
      [Misery] RaphielOhneF 3 months ago

      What about a GPU? Those are quite expensive today and aren't really fast enough for high-resolution simulations, video rendering or gaming.
      Besides, if the parts get smaller, I'd expect heat to be less of a problem. But maybe you're more of an expert on this than me.
      Also, as I understand the video, smaller parts are more energy-efficient, which might be helpful in making (smartphone) batteries last longer.

  • jonden
    jonden 3 months ago +1

    Ok I understood this

  • mhtinla
    mhtinla 3 months ago +2

    It's ok to be white.

  • Aurélien Carnoy
    Aurélien Carnoy 3 months ago

    Let the material self assemble, that is a good way to think.

  • Animesh Sharma
    Animesh Sharma 3 months ago

    using yin-yang of molecules leading to directed self-assembly #awesome

  • Aurélien Carnoy
    Aurélien Carnoy 3 months ago +44

    Design lighter software

  • Chinmai Naregal
    Chinmai Naregal 3 months ago +2

    This is epic!!

  • Calvin K
    Calvin K 3 months ago

    why dont virtualize hardware

    • Logan
      Logan 3 months ago +2

      Calvin K what will that virtualized hardware run on, other hardware?

  • Tribhuvan Sah
    Tribhuvan Sah 3 months ago

    This technology is useful

  • Tony Ray
    Tony Ray 3 months ago +1

    TED= Transhumanistic Evolutionary Directive

  • Ari Marun
    Ari Marun 3 months ago

    info age

  • MikaelKKarlsson
    MikaelKKarlsson 3 months ago +145

    On the other hand, most types of software are wasting resources like never before.

    • Sand Man
      Sand Man 2 months ago +1

      *cough* java *cough*

    • Bradley P
      Bradley P 2 months ago

      Check out MIT's RAW technology where the entire chip is programmable. The is not field programmable logic it is the programmable path ways to several billion RISC processors on a single die.

    • hytlerson
      hytlerson 3 months ago

      Screw Electron

    • BurnGiordano
      BurnGiordano 3 months ago

      All this innovations and vehicles have been stuck at what for how long?..

    • whatsupbudbud
      whatsupbudbud 3 months ago +7

      As junior developer, it's hard to disagree but try to understand us as well. Business requires a multitude of technologies, everyone wants the best of new stuff - all of that needs to built fast and agile. This means that a lot of the hard stuff that NASA did when they coded machines for going into space decades ago is now delegated to integrated software processes (think memory allocation and such). It's not only coders being lazy, the whole industry is a self-fulfilling prophecy.

  • boson96
    boson96 3 months ago +2

    Singularity 2050 baby!

    • Skandar
      Skandar 3 months ago +1

      boson96, believe me, it’ll be sooner than this

  • Chuck Brenner Indianashogman

    Really....... Your a little behind on the tech , but it makes sense to only tell little at a time.

    • Chuck Brenner Indianashogman
      Chuck Brenner Indianashogman 3 months ago +1

      @Disruptive Times true, alongside interdimensional telekinesis, tough subjects.But somebody godda do it.

    • Disruptive Times
      Disruptive Times 3 months ago +1

      Its for a general audience. Talking about what else already brewing in the labs would take hours and needs an educated audience. So, yes, a little at a time. A story about an neural atomristor network might be a little too much for a 10 min talk.

  • Michael Cook
    Michael Cook 3 months ago +1

    More and more powerful ways to spy on us more and more, and to predict and plan for what we are going to be saying and doing.

  • Paul Teich
    Paul Teich 3 months ago +1


  • tanuj9802
    tanuj9802 3 months ago +2

    As an Electrical Engineer studying Computer Architecture, this is very useful and inspiring.

    • Abhiya
      Abhiya 3 months ago

      @Youssef It will at least I think solved tons at as possible as I can say min at least 100s + like, robot replicate it self to mining on Moon, Mars, that will solve energy requirements itself for the people that thoes who live on thoes planet. We can use same technology to let robot to self replicate itself to solve robot requirement for companies. We can even use it in biology to self replicate some insider our body which required like any DNA repairing this is very advance I know but, it also will possible easily we extend it. Sorry for my poor English... There are more and more as much as I can express ..

    • tanuj9802
      tanuj9802 3 months ago +1

      @Youssef my field of interest is Power System control using AI. So if we can create more sophisticated circuitry within the computers, we can achieve high level control on power grids using computers making their management more feasible. A lot of math prototyping is required though.

    • Youssef
      Youssef 3 months ago +1

      Really interested to hear your take on why it's exciting (I do agree it does sound exciting) but I'm wondering if you know of more use value/areas of application. What would this do to technology? How much further could we go? Apologies and no need to answer if these questions are too vague though!

  • sheamus
    sheamus 3 months ago +1

    Skynet has entered the chat

  • Mr Lurvdjur
    Mr Lurvdjur 3 months ago +22

    And he's info is already out dated, we're down to 7nm. :)

    • Dazer
      Dazer 3 months ago +3

      @Rykahnz Yeah.

    • Rykahnz
      Rykahnz 3 months ago +3

      Is it fabbed by TSMC?

    • Dazer
      Dazer 3 months ago +8

      @Rykahnz There is already a 7nm process. If you want a consumer product, a GPU made by AMD called Radeon VII. It is built on the 7nm node and is available for purchase.

    • Rykahnz
      Rykahnz 3 months ago +1

      As far as I know the 7nm TSMC made AMD CPUs don't hit until later this year so no not quite

    • Darth Pro
      Darth Pro 3 months ago

      really? that's cool! can you share a source?

  • GuitarZombie
    GuitarZombie 3 months ago


    • Feroce
      Feroce 3 months ago +2

      He explained why. It is because we're hitting a wall soon on how small we can manufacture microchips and thus on how much processing power we have. This is one of a few techniques we can use to continue to advance technology.

    • GuitarZombie
      GuitarZombie 3 months ago

      @boson96 It is not needed

    • boson96
      boson96 3 months ago +2

      why not?

  • bgcvetan
    bgcvetan 3 months ago +6

    Now that dosen't scream "Exterminate" at all.

  • myutubechannel
    myutubechannel 3 months ago +69

    hardware has always been the limitation for software, strange to see him make the argument it is something recent. Also the latest version of the chip making machine will cost you 250mln. SDA (self directed assembly) has been around for quite a while now. it still requires a photo lithography chip making machine. Molecules are big and that together with the fact that below a couple of nm's quantum effects kill your transistor is the real limit for cost effective processing power. SDA is not a disruptive technology but complementary, the semi conductor industry has a lot of money invested in current production techniques they cannot just throw all of that away and make a profit. What the industry really is doing to counteract the slowdown of progress in photo lithograph is 3D stacking. Samsung, Intel, TSMC are all doing it. btw e-beam is another example of 'break-through' technology. What we need is chips based on photons.

    • Cody Nemo
      Cody Nemo 2 months ago +1

      I loved the fact that the first couple of google results weren't even about the topic when googling e beam.
      Also it's still a disruptive technology for cost efficiency. 3d stacking and e-beams are great but it'll cost a shitton and increase production times. If this technology is adapted it could - for a short while reverse the ever increasing costs.
      And yea photonic chips would be great but we're as far off from those as star trek pads were from ipads. Sure it definitely will exist and soon and we sorta have parts of the technology, but we so do not know how to actually efficiently do it much less have the details down of how to produce it not to mention at a scale, so it's still a couple decades away.

    • Marc Jackson
      Marc Jackson 3 months ago

      Addressing the heat issues, already we're turning off over 50% to reduce heat, 3D stacking is increasing the density, where's the heat going to go, can we put in graphene pathways

    • Rykahnz
      Rykahnz 3 months ago +1

      @Edwin CC Vital my old professor is working on it but the guy was so up there that I didn't really understand it in detail. He referred to it as trapping electrons in an elevator to turn on the transistor. Presumably you use high energy boundaries to trap the electron in it's waveform but how you then lower the boundary to release it again and how to detect it idk. He said if it works he can operate at terahertz. I'm guessing there are other efforts

    • Edwin CC Vital
      Edwin CC Vital 3 months ago

      @Rykahnz I'd like to know more about it

    • John Hillman
      John Hillman 3 months ago

      @Balaji Annamalai the real solution is magical thinking.

  • J J
    J J 3 months ago +9

    I’ve experienced the molecular frustration in stacking chips when I get the fever for the flavor of a Pringles.