"real" 3D Processors

• Osayidan wrote on 2011-09-07 20:55
  IBM and 3M teaming up to stack many layers of "2D" processors into a 3D one. Seems like this will be interesting.
  
  

  
  3D hype is fast wearing out its welcome, but there's at least one area of industry where the buzzed about term could usher in true innovation. Announced today as a joint research project, IBM and 3M will work towards the creation of a new breed of microprocessors. Unlike similar three-dimensional semiconductor efforts by Intel, the two newly partnered outfits plan to stack up to 100 layers of chips atop one another resulting in a microchip "brick." Under the agreement, IBM will contribute its expertise on packaging the new processors, while 3M will get to work developing an adhesive that can not only be applied in batches, but'll also allow for heat transfer without crippling logic circuitry. If the companies' boasts are to be believed, these powerhouse computing towers would cram memory and networking into a "computer chip 1,000 times faster than today's fastest microprocessor enabling more powerful smartphones, tablets, computers and gaming devices." That's a heady claim for a tech that doesn't yet exist, but is already taking swings at current faux 3D transistors. Official presser and video await you after the break.
  

  [Image: http://www.blogcdn.com/www.engadget.com/media/2011/09/ibm-3m.jpg]

  
  

  
  
  http://www.engadget.com/2011/09/07/ibm-and-3m-join-forces-to-fab-3d-microchips-create-mini-silicon/
• Conor wrote on 2011-09-07 20:58
  ohboyohboyohboyohboyohboyohboy.
  I hope this works/has the amazing results that they're predicting it to.
• Osayidan wrote on 2011-09-07 21:10
  My only concern is heat, but they'll probably develop better materials and heatsinks to take care of that.
• Kyreffer wrote on 2011-09-07 21:31
  Hope its not 100 times more expensive. 10000 dollar processors? nty
• Osayidan wrote on 2011-09-07 21:37

  Quote from Kyreffer;581249:

  Hope its not 100 times more expensive. 10000 dollar processors? nty

  
  
  Will be at first like any new technology, but as the manufacturing process gets cheaper and it becomes standard it won't be any more expensive than what we have today.
• Conor wrote on 2011-09-07 21:38

  Quote from Osayidan;581253:

  Will be at first like any new technology, but as the manufacturing process gets cheaper and it becomes standard it won't be any more expensive than what we have today.

  
  
  How long would you estimate that it takes for the price to drop to something affordable?
• TA wrote on 2011-09-07 21:41
  I hope it works out.
  
  At this point, basically all that's been said is that they're going to start trying to make it.
  
  Hell, we've been trying to cure cancer, come up with immortality, time travel and invent lightsabers for years now, and we still haven't done it. Just because they're going to try doesn't necessarily mean they will succeed.
  
  Then again, it's nowhere near as complex as that, so I hope they do succeed.
• Osayidan wrote on 2011-09-07 21:56

  Quote from Conor;581254:

  How long would you estimate that it takes for the price to drop to something affordable?

  
  
  Once it's actually done, maybe 2 - 3 years.
  Keeping a niche market instead of going everywhere would not be as profitable. So they're going to push this as much as they can, likely by licensing it to other chip manufacturers (if those guys don't R&D their own, which they probably are. yay competition).
  
  The only way I can see this technology being ignored on the general business and consumer market is if cloud computing picks up very quickly and drives a huge demand for tons of power in as dense a space possible (it already is but not on the scale needed to make IBM ignore consumer markets).
  
  I also don't see very much stopping this from happening. IBM already has crazy R&D going on in just about every field needed to get this done, for them to partner up with 3M kind of tells us they're confident.
• whocares8128 wrote on 2011-09-07 22:00
  Layering has been around in integrated circuits since before the invention of the double cheeseburger. This is just like trying to make a centuple cheeseburger and hoping there's a mouth out there big enough to be able to eat it.
  
  So my question is...what will it take to power this thing? The more power it draws, the harder it will be to keep cool, and we know cooling is a major concern.
• Osayidan wrote on 2011-09-07 22:15

  Quote from whocares8128;581273:

  So my question is...what will it take to power this thing? The more power it draws, the harder it will be to keep cool, and we know cooling is a major concern.

  
  
  That's what they're working on. It mentioned mobile devices so we can be sure they have intentions (and probably already good data on how to do so) to keep it low powered and relatively cool.
• Joker wrote on 2011-09-07 22:53
  /inb4itcosts100xTheNormalStandardAndBurnsOutInAMonth
• Sumpfkraut wrote on 2011-09-08 02:24
  I feel the blood pumping down there.
• Zid wrote on 2011-09-08 03:46
  I'm more interested in that adhesive.
  
  But sounds good overall.
• Osayidan wrote on 2011-09-08 06:40

  Quote from Zid;581651:

  I'm more interested in that adhesive.
  
  But sounds good overall.

  
  
  Got some plans to stick many layers of extremely thin circuitry together?
• Lolicon wrote on 2011-09-08 11:25
  The thing looks like it'll break when you touch it.
  Then again, that applies to almost every small, expensive things.