A very cool and strange thing (for me at least) has happened. Having spent a great deal of time digging through my Google analytic reports, I can say with certainty that my personal site traffic is on the rise. What’s most interesting to me is that a good portion of that new readership linked into my site from my corporate blog. Even more interesting, our corporate blog over at www.discursivelabs.com has far eclipsed the readership of my personal blog here at www.hunterdavis.com. As such, I thought I’d share the link to my newest article.

In the fifth article in our ongoing series on low power compilation clusters, things are really starting to get interesting. I run you through a distributed Java, Fortran, and ARM->X86-64 compilation using an updated set of scripts we created in our previous articles, as well as post an update on the real-world Pogoplug compilation cluster (hint: twice as fast).

 

As the title implies in our latest article over at Discursive Labs we walk you through the creation of a fully distributed compilation system (i.e. a fully federated system not based on DistCC, Sun’s DMAKE, or other existing distributed compilation tools). The scripts are available in the article and can be dropped into an existing compilation node or as a base for future development. While I have posted a few articles over at Discursive Labs since I last posted here, I thought this one in particular may be of interest to anyone wishing to make their own cluster for compilation or scientific computing. If you’re interested in cross-compiling, low power ARM clusters, virtual clusters or distributed computing and you’re not reading our continuing series then you should really catch up.
Active Compile on 6 Virtual Nodes

© 2012 HunterDavis.com Suffusion theme by Sayontan Sinha