It's been a while since an acquisition was worthy of comment, but this week's Intel-Sarvega acquisition is one that's very, very interesting. The XML/Web Services security/management space has been in flux for the past two years, with several vendors vying for the top spot in the field and others grooming themselves for acquisition. Sarvega is the first to be scooped up, and by the most unlikely of buyers, Intel. Intel, if you remember back far enough, picked up XML player iPivot in the 90's. The XML Director was ahead of its time and didn't go far, eventually ending up in the hands of Tarari when it was spun off from Intel a few years ago and entered back into the XML market with its own XML acceleration technology in a PCI form factor.
Now it appears Intel is heading back into the XML game, and Sarvega is an interesting purchase given its architecture. It's a software based solution, so it's possible that Intel is planning on doing exactly what we've been hoping to see - XML accelerated parsing built right into the chipsets that make up most enterprise motherboards. This would be a boon, in the end, for XML security and management vendors to pick up some performance gains without the need for custom silicon and our own testing earlier this year proved that Sarvega certainly has the chops for such a task in terms of the latency introduced by its parsing technology - which is say nearly none.
It would give its leading competitors Data Power, Forum Systems, Reactivity and Actional something to think about. If such a thing were to come to fruition, Actional and Reactivity would surely benefit the most, given their solutions are software only platforms and could easily take advantage of integrated XML acceleration, unlike ForumSystems and DataPower who use custom hardware to generate the horsepower necessary for enterprise class performance when handling XML. It will be interesting to see if Intel is going to rip apart Sarvega's technology and integrate it into chipsets or whether it will keep the technology as is and attempt to compete against its partners yet again. Past experience has hopefully taught Intel that competing against its OEM channels is a bad idea, but sometimes it's hard to teach and old dog new tricks.