|
|
|
|
|
|
recent entries "My take on Thanksgiving" "Belichick used results from dynamic programming!" "I know he's not to blame, but..." "In sickness and in health..." "Fall Beautiful"
|
Dec 03, 2008 Some perspectiveI was recently puttering around the internet and found this story about computers that are breaking the petaflop barrier. (As an aside, my spell check currently recognizes “gigabyte,” but thinks I’m making stuff up with “petabyte.” Technology, know thyself.) The article talks about how having such powerful computers could really change scientific research as we know it. Modeling becomes not just a support tool for theory and experimentation, but the research itself. Pretty cool. But what really got me are the pictures. You just see huge racks of electronics, filling a basement. Haven’t we seen that before? Oh that’s right. This is what computers used to look like. Thanks to the transistor (see you later, vacuum tubes! It’s been fun!) and the integrated circuit, we no longer need an entire basement to house a computer. 50 years after the creation of the ENIAC, researchers at Penn (home of the first computer) decided it would be fun to put all of the ENIAC’s functionality on a silicon chip. The result was smaller than a dime. Let’s take another step back. It took us 50 years to go from a computer that filled a basement to a dime-sized silicon chip. Right now, we have a petaflop computer that fills a basement. 50 years from now, we’ll have… ? The possibilities are astounding. I have no idea what our computers will look like or where they’ll be (did anyone think we’d have computers in cars?) 50 years from now. But being in grad school means I’m in on the ground floor of that innovation. It’s more exciting than any job I could have gotten. There are other perks of grad school: the flexibility of your schedule, the freedom to be very creative, the chance to meet some of the smartest people alive, etc. But those are all really secondary. 50 years from now, my grandchildren will be walking around with petaflop computers on their watches that enable holograms of their friends to pop up and talk to them. They’ll think it’s normal. And I’ll be able to say that some of my research made it possible. I’ll be the coolest grandma around.
3 Responses To This Entry:Post A Response
|
do you think we’ll have the same progress in 50 years as they did during the last 50 years, since we can’t scale anymore? or will Ge/III-Vs get us there
?
Posted by: Ivan on December 3rd, 2008 at 12:37 pm
That’s a very interesting question - and one that I tried to address pithily in my post.
The ENIAC was based on vacuum tubes. The transistor and integrated circuit made them obsolete and completely changed the way we compute. So, taking a lesson from history, it seems that the transistor as we know it could be replaced.
What it could be replaced with is anybody’s guess. I’m in the integrated photonics field, so I have to think that we’ll all be using light-based computers in 50 years. But there are folks working to design quantum computers. Who knows what those would look like? Or maybe someone will finally find an application for carbon nanotubes.
Posted by: ndilello on December 3rd, 2008 at 4:04 pm
yay holograms!! fyi, there’s a cool new holography exhibit at the MIT museum. i encourage everyone to check it out.
Posted by: Laura on December 6th, 2008 at 6:03 pm