1. Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

What comes after nanometer?

Discussion in 'General Hardware' started by hat, Jun 17, 2009.

  1. Mussels

    Mussels Moderprator Staff Member

    Joined:
    Oct 6, 2004
    Messages:
    42,485 (11.47/day)
    Thanks Received:
    9,765
    one thing to consider wafer, is location. whats cheaper near you, may not be so in other countries. Around here AMD is cheaper than intel for budget machines, but in midrange and above intel have better price to performance. its also a lot easier (read, cheaper) to overclock intel machines as well. (any board with voltage control, will get decent overclocks out of even the cheap CPU's)
     
  2. Velvet Wafer

    Velvet Wafer New Member

    Joined:
    Jun 16, 2009
    Messages:
    5,123 (2.58/day)
    Thanks Received:
    990
    Location:
    North of Germany
    The integration of the Memory Controller in the Cpu was one of the most intelligent things to do when building this thingy.

    it had nothing to do with production process... as you said, there are bigger/better foundries regarding chip construction.

    i just thought loud. sorry for that^^

    @ mussels

    Here in Germany AMD CPUs are cheap as hell, and i find that even an a64x2 is easily overclockable in a not too complicated and expensive setup, regard my rig, it cost me not as nearly as much as an i7 965 would cost alone. Even with my soon-to-be 955 im well within the margin.

    EDIT: I think there are amd fabs here in Germany, i always had Cpus from them which were diffused here. the low price may be due to this... (got my phenom 955 for 190 euro, yesterday, that are 332.99 Austr. Dollar)
     
    Last edited: Jun 17, 2009
  3. Kreij

    Kreij Senior Monkey Moderator Staff Member

    Joined:
    Feb 6, 2007
    Messages:
    13,881 (4.87/day)
    Thanks Received:
    5,616
    Location:
    Cheeseland (Wisconsin, USA)
    After nanometer comes ... the 2nd nanometer! (rimshot) Sorry couldn't resist. :D

    I am not so sure I would make statements like "not in our lifetime" or "several hundred years" or "no one will ever need more than 640k" (I had to through that in, it's one of my favorites), as the advancement in technology is limited by the current materials and processes that are in place. One amazing discovery, and we could be at picometer scale integration within a decade.

    Just my 2 cents :toast:
     
  4. Velvet Wafer

    Velvet Wafer New Member

    Joined:
    Jun 16, 2009
    Messages:
    5,123 (2.58/day)
    Thanks Received:
    990
    Location:
    North of Germany
    where we start with nanotubes and super-conductivity (Supraleitung)
     
  5. hat

    hat Maximum Overclocker

    Joined:
    Nov 20, 2006
    Messages:
    16,954 (5.79/day)
    Thanks Received:
    2,074
    Location:
    Ohio
    Honestly I see Intel as being the best choice in every situation for an overclocker, unless there is some sort of brand loyalty towards AMD or some urge to check out what's on the AMD side of the pond.
     
    Crunching for Team TPU
  6. hat

    hat Maximum Overclocker

    Joined:
    Nov 20, 2006
    Messages:
    16,954 (5.79/day)
    Thanks Received:
    2,074
    Location:
    Ohio
    640k? That's outrageous! With that much memory, I could install my entire hard drive on that!
     
    Crunching for Team TPU
  7. Velvet Wafer

    Velvet Wafer New Member

    Joined:
    Jun 16, 2009
    Messages:
    5,123 (2.58/day)
    Thanks Received:
    990
    Location:
    North of Germany
    i used several intel procs, but i love my AMDs... ever since i touched them.
     
  8. btarunr

    btarunr Editor & Senior Moderator Staff Member

    Joined:
    Oct 9, 2007
    Messages:
    28,860 (11.07/day)
    Thanks Received:
    13,715
    Location:
    Hyderabad, India
    Here we go again Intel vs. AMD :rolleyes:

    I think it will be a while before we can enter the picometre domain, if solid-state electronics is even possible at that small scale.

    Thankfully Intel isn't the only company deep into VLSI technology research. It finds competition in IBM (which was the first to reach 90 nm (<100 um)) so competition will accelerate things.
     
  9. Velvet Wafer

    Velvet Wafer New Member

    Joined:
    Jun 16, 2009
    Messages:
    5,123 (2.58/day)
    Thanks Received:
    990
    Location:
    North of Germany
    we need a second underdog here!
     
  10. Kreij

    Kreij Senior Monkey Moderator Staff Member

    Joined:
    Feb 6, 2007
    Messages:
    13,881 (4.87/day)
    Thanks Received:
    5,616
    Location:
    Cheeseland (Wisconsin, USA)
    Both companies (Intel and AMD) have come a long way since the inception of their early processors. That's the beauty of competition. I never want to see one of them "win", for as long as they keep battling it out we, as consumers, are the winners.
     
  11. Velvet Wafer

    Velvet Wafer New Member

    Joined:
    Jun 16, 2009
    Messages:
    5,123 (2.58/day)
    Thanks Received:
    990
    Location:
    North of Germany
    but intel should lower their prices so amd could follow, so everyone could go and build a rig
    at least this is a wet dream of me :p
     
  12. btarunr

    btarunr Editor & Senior Moderator Staff Member

    Joined:
    Oct 9, 2007
    Messages:
    28,860 (11.07/day)
    Thanks Received:
    13,715
    Location:
    Hyderabad, India
    Underdogs are for the weak. Competitors are good for everyone.

    Let's suppose there are two companies, A and B, A being the stronger forward-thinking one which is a few steps ahead of B, while B is the typical underdog one that keeps catching up, just so people have a binary choice. Do you think choosing B is going to do you any good? No. You are willingly choosing something that's a generation or two behind, and not giving credit to innovation, which ends up being the causality. If you begin to think rationally, and give credit to A, by accepting it, you are forcing B to do something radical, or get out. Right now Intel is the peerless leader with AMD providing the binary choice. The two almost traded places between 2004 and 2006. We don't need any more underdogs. We need a real competitor to Intel, one that has the potential to push it to at least 50-50, or 40-60 (market-share), not something that stays at 80-20. To do so, there needs to be a competitive machine architecture to x86 which is also popular. You can't fight Intel with its own technology.
     
  13. MRCL

    MRCL

    Joined:
    May 31, 2008
    Messages:
    5,791 (2.44/day)
    Thanks Received:
    860
    Location:
    Switzerland, Heart of Europe
    Maybe IBM is secretly working on a new CPU...

    You're right bta, choosing the cheapest is not always the best solution. However Intel needs a competitor. There need to be a minimum of two manufacturers, or its economicly unhealty.
     
  14. btarunr

    btarunr Editor & Senior Moderator Staff Member

    Joined:
    Oct 9, 2007
    Messages:
    28,860 (11.07/day)
    Thanks Received:
    13,715
    Location:
    Hyderabad, India
    There is no real match to Core i7 920. Yet its price has remained stable. AMD can't influence Intel's pricing to a big extant. Despite there being sub-$150 quad-core chips by AMD, Core 2 Quads are still holding a greater market share.
     
  15. Kreij

    Kreij Senior Monkey Moderator Staff Member

    Joined:
    Feb 6, 2007
    Messages:
    13,881 (4.87/day)
    Thanks Received:
    5,616
    Location:
    Cheeseland (Wisconsin, USA)
    There is nothing wrong with more competition, but if anyone wants to enter into the battle with Intel and AMD they had better have all their ducks in a row or their efforts will be fruitless.
    Look at nVidia and ATI. If Intel joins the fray with Larrabee(sp?) in the dicreete GC market, the whole playing field will change wildly. I am greatly looking forward to that.
     
  16. cray86

    cray86 New Member

    Joined:
    Mar 18, 2008
    Messages:
    97 (0.04/day)
    Thanks Received:
    10
    Location:
    Springfield, IL
  17. Odin Eidolon

    Odin Eidolon New Member

    Joined:
    Feb 12, 2008
    Messages:
    921 (0.37/day)
    Thanks Received:
    156
    Location:
    Rockclimbing somewhere in Italy
    next step is nanotubes + high temperature superconductive metals + qubits
     
  18. Velvet Wafer

    Velvet Wafer New Member

    Joined:
    Jun 16, 2009
    Messages:
    5,123 (2.58/day)
    Thanks Received:
    990
    Location:
    North of Germany
    and heatsinks out of buckypaper!
     
  19. Kreij

    Kreij Senior Monkey Moderator Staff Member

    Joined:
    Feb 6, 2007
    Messages:
    13,881 (4.87/day)
    Thanks Received:
    5,616
    Location:
    Cheeseland (Wisconsin, USA)
    Nanotubes are not all they are cracked up to be. I ordered a ten pack from NewEgg and accidentally spilled the contents when I was opening the package. I can't find them anywhere. My cat, however, has been able to do intra-spacial repositioning of its litter to a spot out in the back yard. This may not be a bad thing.

    On Topic : I think we will see 22 or 23 nm hit the market pretty soon and then half that again in aboput two years. We should be into single digit nanometer chips relatively soon.
     
  20. Mussels

    Mussels Moderprator Staff Member

    Joined:
    Oct 6, 2004
    Messages:
    42,485 (11.47/day)
    Thanks Received:
    9,765
    i think 22nm is when the slowdown will start. the steps they shrink it by will get smaller and smaller, with more time between - at least until some new tech comes out that lets them do it better.

    65nm to 45nm = 24% smaller.
    45nm to 32nm = 29%
    32nm to 22nm = 32%
    22 to ?


    well whatever they do, its always good for us in general. soon* we'll be playing crysis 4 on a holographic projector at 9,0001 FPS and controlling it with our minds.

    *Soon is objective. It will happen in the blink of an eye if you're already a few hundred years old.
     
    Last edited: Jun 17, 2009
  21. btarunr

    btarunr Editor & Senior Moderator Staff Member

    Joined:
    Oct 9, 2007
    Messages:
    28,860 (11.07/day)
    Thanks Received:
    13,715
    Location:
    Hyderabad, India
    22 to 16, 16 to 12, 12 to 8, 8 to 6 (6.5), 6 to 4 (4.5), 4 to 3 (3.2)...
     
  22. Mussels

    Mussels Moderprator Staff Member

    Joined:
    Oct 6, 2004
    Messages:
    42,485 (11.47/day)
    Thanks Received:
    9,765
    i was just noting (as i had just realised it myself) that while the shrinks are smaller differences number wise, they're getting larger and larger, % wise.
     
  23. Odin Eidolon

    Odin Eidolon New Member

    Joined:
    Feb 12, 2008
    Messages:
    921 (0.37/day)
    Thanks Received:
    156
    Location:
    Rockclimbing somewhere in Italy
    yeah i noticed it too.

    PS according to your avatar i should eat my subwoofer. ouch! :laugh:
     
  24. IceCreamBarr New Member

    Joined:
    Jun 17, 2009
    Messages:
    76 (0.04/day)
    Thanks Received:
    22
    Location:
    Montreal, Canada
    Hi all,

    New here. A little math I did - may be incorrect but let's give it a shot: We will be at the atomic scale in 36 years assuming a shrink every 2.5 years. Atom is 250 picometers and it would take roughly 15 generations to reach.

    Generation Picometer pm^2 Years
    65,000 4,225,000,000
    45,000 2,025,000,000 1
    1 32,000 1,024,000,000 3.5
    2 22,000 484,000,000 6
    3 15,556 242,000,000 8.5
    4 11,000 121,000,000 11
    5 7,778 60,500,000 13.5
    6 5,500 30,250,000 16
    7 3,889 15,125,000 18.5
    8 2,750 7,562,500 21
    9 1,945 3,781,250 23.5
    10 1,375 1,890,625 26
    11 972 945,313 28.5
    12 688 472,656 31
    13 486 236,328 33.5
    14 344 118,164 36
    15 243 59,082

    Barr

    EDIT: ?? the spacing in my post doesn't match the spacing in the viewable response? Hope you can make out that jumble.
     

Currently Active Users Viewing This Thread: 1 (0 members and 1 guest)

Share This Page