I've been following the progress of Snabbswitch for a while now, and I think it is absolutely fascinating technology. While its geared for network i/o, I think that the methods of reaching the performance gains attained by Snabb can be applied to other forms of software projects, too .. from my point of view, the architecture of Snabb would be very applicable to other applications in the embedded Linux space - particularly, for pro Audio/DSP applications. I think kernel bypass has a lot of potential to open up ways of developing apps for this market segment which I hope to continue to explore, and maybe .. one day .. commercialize.
I applaud Luke and the Snabb team - you've opened doors that I hope others will walk through, and see that the technology you've pioneered has value far beyond the networking realm.
I used to work in networking. One of the things that always amazed me is people's willingness, or excitement to be pinned to one proprietary networking system. On the proposal of a network that utilized primarily open standards, they were very put off. I hope Snabb succeeds where I failed.
I think Luke's work on Snabb is impressive and it's cool to see efforts to make a marketplace for Snabb based solutions. I don't know if it will work out since companies can be fickle when choosing solutions, but I'm rooting for Snabb nonetheless. It's great to see people building non-trivial network things with kernel bypass.
I applaud Luke and the Snabb team - you've opened doors that I hope others will walk through, and see that the technology you've pioneered has value far beyond the networking realm.