Getting the point of parallel maths

What’s 1000.2 minus 1000? It’s not a trick question, but you might find your computer tells you the answer is 0.200012, and not the 0.2 you might expect. It’s a perfect example of the difference between precision and accuracy: the computer’s got it wrong to six decimal places. Tim Mattson and Ken Strandberg have written [...]

Do spreadsheets offer the solution to the challenge of parallel programming?

Every once in a while somebody writes a blog post that really changes your perspective. I Programmer has changed the way I look at a spreadsheet, perhaps forever, by pointing out what goes on underneath the hood, and showing how it relates to parallel programming. I’ve been using spreadsheets for over 20 years. I first [...]

Gustafson speaks about reevaluating Amdahl’s Law

I’ve written in the past about Amdahl’s Law, and it still comes up from time to time at conferences as part of the conventional wisdom of parallel programming. The law basically says that the speedup possible from using multicore is limited by the serial part of a program. I haven’t written so much about Gustafson’s [...]

Follow

Get every new post delivered to your Inbox.