Added: 4 years ago
From: channelintel
Views: 17,753
Sort by time | Sort by thread (beta)

Link to this comment:

Share to:

All Comments (30)

Sign In or Sign Up now to post a comment!
  • We should continue the chips with graphene! 

  • Keep going Intel. I want to affordably run Linux Mint. Which is running in 6 x 4320p. Which is running Windows 8 which in VirtualBox in 4320p. Which is running high end games in max settings. While recording that game at the same time in FRAPS. Which is also running Sony Vegas or Final Cut Pro in another window in 1080p+. While running a whole bunch of nice programs in Linux like Transmission or KTorrent. While having 50 tabs open in Firefox. You can do it Intel, we're counting on you.

  • A horse in a cleanroom at 1:05?! If that happened, that place would neve again be useful as a cleanroom for chipmaking... >.<

  • Its currently 32 nm. 22nm is just next year. I think it was 17nm soon after that

  • @CruelAngels 22nm is already working. Intel already demonstrated 22nm ivy bridge processors. :D

  • Moore's law will probably decline between 2020-30. Now it's already 30 nm and 20 nm or below are to be expected in the next decade or so.

  • graphene next with much research?

  • the density part of moore's law is set to run out very soon - around ~2035. Well, in between 20 to 30 years.

    Then, we will have graphene. that's 2nm manufacturing process.

    But it doesn't mean the moore's law will stop. Next we have 3d CPUs. Then we have non-atom based transistors. Then extreme mega cloud powered supercomputers...

    but we'll run out of everything by the end of this century. If not earlier. Or if there won't be some big war. Then we would have perfect computer in ~2110

  • Moore's Law lasts only as long as the silicon holding the transistors together. Graphene could take silicon's place, keeping Moore's Law going some more years. But that will most likely be used in high-performance devices, not everyday electronic applications.

    But, come on...3D molecular computing is way too important to ignore. It alone could facilitate quantum computing, getting us closer to the you-know-what.

    Some people are treating Moore's Law like the Beatles: nobody wants it to end.

  • gordon moore is the man

  • What a fag video. 2007, I seriouly hope you died, cause of that utube hasn't deleted you the fuck out !

  • try more like another 10 years. but still not bad. keep up the good work.

  • THIS video totaly ROFL

  • Moore's Law as we know it (i.e. CMOS VLSI) will run out long before 2050

    45 nm scale transistors are now standard and 32 nm scale demo chips are being fabbed

    20 nm is expected in 3 to 5 years

    Somewhere around 10 nm there are fundamental physical limits.

    This means:

    Depending on the doubling time for the number of transistors (i.e.2 or 3 years) the end is 10 to 15 years from now.

  • glad someone brought that up

  • Just means we'll have to do it another way. How? Beats the hell out of me. I never would have thought of HAFNIUM, for Pete Squeaks!

    I wouldn't be surprised if there's some molecule that can act as a transistor. There's already talk of using electrons (not atoms, electrons) for memory storage.

  • Another way = quantum computing

  • Yeah. That won't be the continuation of Moore's Law in the strict "number of transistors" form, but it will allow the continued doubling of processor speed which has been going on since the days of mechanical adding machines.

  • yeah

  • the end for hafnium you mean? maybe they will find something else that allows to go even smaller

  • wt u said might be true, but no one noes wt will happens in the future, we see physical limitations atm, doesn't mean the limitation is invincible, 40years ago, ppl probably dun even noe there's a length unit- nano metre. I support your thoughts basically, but technology, hmm no one knows.

  • 13nm exactly is limit for a silicon, but with some other material it is possible to continue with Moore's law ;)

  • @nepssis umm even at 90nm we dont use silicon. we use SiGe or strained silicon. it has considerably higher electron mobility.

  • yep..but till there well find a solution..i ve seen they are investigating hardcore ins everal diferent solutions already for a problem 10-15 years ahead

  • @westhighlander There are plenty of sub-nanometer structures that could be used as a transistor.

  • It fluctuates a bit. There was a while when it was only taking a year, and for a while it seemed to settle in at eighteen months.

    I wanted to write something set in 2045. To see what sort of capability they would have, I assumed a Moore's Law of THREE years. Still turned out that the average person could have a computer that runs at 24 THz and has 4 TB of RAM.

    And that was me being ridiculously conservative.

  • I hope a new law comes that says each year instead of each two.

  • well the truth is that is can only get so small before the intermolecular forces work against size, and then we wont be able to shrink computer sizes anymore. but likely by that time, a May 2008 supercomputers abilities will be capable in something about the size of a grain of rice...

    I guess we will just have to start working on doing it some other way...

  • By that time they will probably have Quantum computers, and then they can start making those faster and faster...haha :P

    Quantum computers are still many years in the future though... they only made some tests so far, and those were Quantum computers that could only do simple math... but just image what the technology could do one day :)

  • Yes well, asml can now make 32 nm =) But its not a really sold IC. So see you next time in the 32nm series..

Loading...
Alert icon
0 / 00Unsaved Playlist Return to active list
    1. Your queue is empty. Add videos to your queue using this button:
      or sign in to load a different list.
    Loading...Loading...Saving...
    • Clear all videos from this list
    • Learn more