Do you have a big endian or little endian problem?

on Oct 27, 10 • by Brendan Harrison • with No Comments

Ok… bad pun but question still stands. We wanted to try and answer that question so we worked with the team at VDC Research to try and quantify some of these questions. You can download their full report on the multicore and multiprocessor landscape, but here’s one piece of data that I thought might be...

Home » General Coding » Do you have a big endian or little endian problem?

Ok… bad pun but question still stands. We wanted to try and answer that question so we worked with the team at VDC Research to try and quantify some of these questions. You can download their full report on the multicore and multiprocessor landscape, but here’s one piece of data that I thought might be interesting. Basically, heterogeneous processor architectures are growing quickly and the number of projects using simple processor architectures is diminishing fast.

Multi-processor development

Data from VDC Research showing trend in multiprocessor development

Really, backs-up what we all instinctively know and understand but nice to see some empirical evidence to add to the mix of customer feedback, social media chatter, and articles we’ve been seeing on the topic. We also tried to collect a bunch of other resources on this topic in one spot with a previousĀ endian blog post we published.

Of course, our focus is on the code analysis tooling available to address these issues. If you’re doing a port, the prospect of manually (or semi-manually) crawling through thousands of instances where you need to modify code to support a different endian format really sucks. Some automation would certainly be nice. Our CTO Gwyn Fisher will be talking in more detail on this topic next week. He has a presentation on improving software reliability in multicore and multiprocessor architectures. Check it out.

Related Posts

Leave a Reply

Your email address will not be published. Required fields are marked *

Scroll to top