Advertisement
If you have a new account but are having problems posting or verifying your account, please email us on hello@boards.ie for help. Thanks :)
Hello all! Please ensure that you are posting a new thread or question in the appropriate forum. The Feedback forum is overwhelmed with questions that are having to be moved elsewhere. If you need help to verify your account contact hello@boards.ie
Hi there,
There is an issue with role permissions that is being worked on at the moment.
If you are having trouble with access or permissions on regional forums please post here to get access: https://www.boards.ie/discussion/2058365403/you-do-not-have-permission-for-that#latest

Moore's Law - only 10 years left!?

  • 15-09-2012 12:44am
    #1
    Banned (with Prison Access) Posts: 21,634 ✭✭✭✭


    :eek: http://news.softpedia.com/news/Intel-s-Experts-Give-Moore-s-Law-Another-10-Years-292139.shtml

    Wtf will happen once we reach 10nm?....optical?, Intel plan to go from 22nm straight to 14nm and bypass 16nm. The next mildstone is 10nm.
    I suppose multiple cores larger then 16 might come into play.

    Intel ain't stoopid, even if they reach and technical bloackade they'll just rehash or retweak existing designs, more cache...improved pipelines etc.

    But tis mad to think in 10 years we'll reach the techincal limitations of current CPU design.


Comments

  • Closed Accounts Posts: 1,455 ✭✭✭RUCKING FETARD


    But tis mad to think in 10 years we'll reach the techincal limitations of current CPU design.
    Then they'll start making them like that electric blue one Summer Glaus character in Alphas made and everyone will be like, Moore who!:pac:


  • Registered Users, Registered Users 2 Posts: 3,537 ✭✭✭SickBoy


    I was watching something or reading something, can't remember which but it was to do with manufactured diamonds. The claim was that silicon is becoming a bottleneck and diamond has better semiconductor properties. It was also suggested that the thermal properties will also be significantly better also and require far less cooling at far higher speeds.
    The future is bright, sparkling even :D


  • Closed Accounts Posts: 1,455 ✭✭✭RUCKING FETARD




  • Closed Accounts Posts: 5,377 ✭✭✭zenno


    The future is hologram projection and it's going very well so far.

    Soon you will come home, and awaiting for you can be any person holographically tuned for your liking...



  • Closed Accounts Posts: 5,377 ✭✭✭zenno


    :eek: http://news.softpedia.com/news/Intel-s-Experts-Give-Moore-s-Law-Another-10-Years-292139.shtml

    Wtf will happen once we reach 10nm?....optical?, Intel plan to go from 22nm straight to 14nm and bypass 16nm. The next mildstone is 10nm.
    I suppose multiple cores larger then 16 might come into play.

    Intel ain't stoopid, even if they reach and technical bloackade they'll just rehash or retweak existing designs, more cache...improved pipelines etc.

    But tis mad to think in 10 years we'll reach the techincal limitations of current CPU design.

    Never mind that, but soon they will master 2.5 nano metres and that will be amazing, less heat and wear but the goal is 0.5 nano metres and i wish the scientific community the best of luck as it will be accomplished soon.


  • Advertisement
  • Closed Accounts Posts: 1,455 ✭✭✭RUCKING FETARD


    The Moore's Law Moon Shot
    The computer industry's future depends on a behind-schedule technology that's proving tough to get working.

    intel.asmlx616.jpg
    Next generation: A technician at ASML works on a prototype machine for manufacturing computer chips based on extreme ultraviolet technology, a process that promises to keep chip components shrinking as they have for decades.


    It is seemingly a fact of life that every new generation of computing gadget will be significantly more powerful than the one before, but a looming technical roadblock threatens to undermine that.

    That's why the world's largest chip maker, Intel, announced on Monday that it has invested $4 billion in Dutch company ASML, which makes equipment for manufacturing computer chips.

    The two companies are trying to instigate a collaboration involving the world's largest computing companies—in a kind of silicon moon shot—to ensure that chips keep getting faster by perfecting the tools needed to make smaller features on silicon chips.

    The technology that promises to keep Moore's Law going after 2013 is known as extreme ultraviolet (EUV) lithography. It uses light to write a pattern into a chemical layer on top of a silicon wafer, which is then chemically etched into the silicon to make chip components. EUV lithography uses very high energy ultraviolet light rays that are closer to X-rays than visible light. That's attractive because EUV light has a short wavelength—around 13 nanometers—which allows for making smaller details than the 193-nanometer ultraviolet light used in lithography today. But EUV has proved surprisingly difficult to perfect.

    Until 2007, Intel believed that EUV would be used to make the 22-nanometer chips that came out this year, but it opted instead to make fixes that extended the life of lithography based on 193-nanometer light. As recently as 2010, the company hoped to use EUV techniques for the 11-nanometer generation of chips that are still several years away, but once again, technical innovations meant that existing lithography had its retirement canceled.

    Delays have also beset EUV. Mostly these have originated with the companies such as ASML and Nikon that develop and sell lithography equipment. One of the biggest challenges has been in making powerful enough sources of EUV light. All types of matter absorb light at these wavelengths, so a source needs to be bright enough to ensure that sufficient light reaches the wafer being worked on. EUV lithography machines are designed to have the beam pass through a vacuum as much as possible, so that air molecules don't get in the way. Special structures have also been developed to direct and focus the light with minimal absorption, but still, by the time the beam reaches the wafer to be worked on, over 90 percent of the original EUV light has been lost.

    ASML's most complete prototypes can etch components on silicon wafers, but have beams roughly half as strong as what would be needed to mass-produce chips economically. At the same time, the company is trying to make progress on the second generation of EUV lithography technology, which will use complex reconfigurable mirrors to create even finer details on chips. It is also trying to switch to using silicon wafers with just over twice the surface area of those used today, so more chips can be made in a single batch, but this will require new equipment.

    Although the Intel deal is formally focused on the next generation of EUV and the switch to larger wafers, the resources it brings should also help ASML address its more immediate problem. "Clearly, there is no next generation if we don't get this generation working," says ASML spokesman Ryan Young. "We obviously have to get the first gen finished first."

    In a video statement posted online earlier this week, ASML's chief financial officer, Peter Wennink, encouraged other chip makers to join the effort to support EUV's development by signing deals similar to Intel's, which commits Intel to supporting a large increase in research and development in return for a 25 percent stake in its supplier. Wennink said the whole industry would reap the profits from such agreements. "This is about making sure that the technology that is needed for the next generation chips comes forward faster," he said. "Ultimately, of course, it will be the consumer that benefits."

    Samsung and Taiwan Semiconductor Manufacturing Company, the two largest chip-making companies after Intel, are the most likely preferred partners, but neither has so far publicly signaled interest. Both Intel and ASML are hoping that these companies and others will put aside their usual competition and pool the funding and expertise needed to get past the roadblock that threatens a coming generation of gadgets.


  • Moderators, Education Moderators, Technology & Internet Moderators Posts: 35,125 Mod ✭✭✭✭AlmightyCushion


    The demise of Moore's Law has been called for a while now and it's still going strong (was it ever really meant to be a law, I thought it was more and observation he made of the industry at the time). Newer technology will come along that will enable us to push chips further and further and postpone the demise for another 5-10 years. Actually, aren't IBM working on using graphene to replace silicon?


  • Registered Users, Registered Users 2 Posts: 893 ✭✭✭U_Fig


    There is research being done on using Carbon nanotubes..and some research is being done into sub atomic range


  • Moderators, Computer Games Moderators, Technology & Internet Moderators Posts: 19,242 Mod ✭✭✭✭L.Jenkins


    U_Fig wrote: »
    There is research being done on using Carbon nanotubes..and some research is being done into sub atomic range

    Quantum state processing/computing is currently in development. Once that becomes a viable technology, I'd give it 10-15 at most to over take current processing technologies. So Moore's law may not have to be cast aside completely.


  • Registered Users, Registered Users 2 Posts: 86,729 ✭✭✭✭Overheal


    February 42? Arizona will be waiting a while


  • Advertisement
  • Registered Users, Registered Users 2 Posts: 86,729 ✭✭✭✭Overheal


    Thats - actually kind of frightening. Makes me think of Ghost in the Shell and AI. The idea that anybody rolling by can hack into your - anything. This would make building a Closed System that much more challenging.


  • Registered Users, Registered Users 2 Posts: 4,020 ✭✭✭Hijpo




  • Closed Accounts Posts: 1,455 ✭✭✭RUCKING FETARD


    Hijpo wrote: »





    He doesn't really say anything that's not fairly obvious...could be this, could be that, sometime when, I should be a Theoretical physicist! I'd put money down on it.

    How would mass production be a problem for Molecular computing once the Tech is perfected?


  • Registered Users, Registered Users 2 Posts: 4,020 ✭✭✭Hijpo


    Hijpo wrote: »





    He doesn't really say anything that's not fairly obvious...could be this, could be that, sometime when, I should be a Theoretical physicist! I'd put money down on it.

    How would mass production be a problem for Molecular computing once the Tech is perfected?

    I only posted it to be discussed, i didnt mean for it to be taken for absolute fact


Advertisement