Kernel Hacker’s Bookshelf: Ultimate Physical Limits of Computation

I finally made the time to finish my latest Kernel Hacker’s Bookshelf article, this time on the physical limits of computation. Computers are made of matter and energy and stuff and so are subject to the rules governing them, and it turns out that physicists know a lot about how matter and energy and stuff behave. When a computer programmer asks “Will Moore’s law ever run out?” you know they’re just trying to start a philosophical argument at a geek party. When a physicist asks that question, they haul out the textbooks and get calculatin’.

Fortunately for the computer programmers without PhDs in physics (and I know at least two personally), Dr. Seth Lloyd decided to share the results of his calculatin’, and I wrote an article based on it:

Content is subscriber only until next Tuesday, unless you fork over a whole $5/month in cash to Jon Corbet – translation: WORTH IT. But if you ask nicely I’ll give you a free link over email.

If you read the original paper in Nature, I added a few extra tidbits, including the computational limits of “smart dust” – computers a cubic micrometer in size – and the relative speed at which we’re approaching the limit of storage capacity versus operations per second. Turns out, maybe there won’t be much file systems work going on in a few decades because we’ll all be busy researching compression algorithms to squish our data into our stagnant data storage technology.

2 thoughts on “Kernel Hacker’s Bookshelf: Ultimate Physical Limits of Computation”

  1. Liked it

    Liked the article. It’s a fun romp.

    Would’ve liked the numbers for max-performance at roomtemperature, 100W, non-reversible done, dunno if those are in the paper.

Comments are closed.