Big and Fast Data Blog

Way back five years ago, just as the iPhone came out, Karthik Ramasamy ( @karthikz ) and I ( @pateljm ) started talking about building a combined real-time streaming and real-time analytical platform to power enterprise and consumer mobile services. We started a company called Locomatix to build such a platform. Sanjeev Kulkarni ( @sanjeevrk ), a founding member of the Google AdSense team, joined Locomatix soon after it was formed.  Next came Chris Kellogg ( @cckellogg ), who joined in his last semester at the University of Michigan. Today, I’m very happy to announce that the Locomatix team is joining Twitter ! Special thanks to our families and friends for supporting us throughout this incredible journey. We simply wouldn't have made it without you! Sanjeev, Karthik and I all got our graduate degrees from the University of Wisconsin, so a big part of Locomatix was powered by UW grads. Go Badgers! We are working on converting Chris into a Badger fan -- its still work in (ve...

So, I have finally decided to start a blog. Why now, you might ask? The first reason is that I have been too lazy to start a blog till now. Since blogs often matter as much as publications, this oversight is inexcusable. Better late than never. The second reason to start a blog is that U. Wisconsin doesn’t, AFAIK, have a blog in the big data space. This is sad given the tradition that our university has in producing key data processing technologies, like the GAMMA parallel database system (which continues to influence what we now call “big data systems”), and BIRCH which provided a key pivot point for data mining (the buzzword-compliant term for that today is “deep data analytics”). This blog is a humble start in trying to get some of what Wisconsin does today in the big data space out into the blogosphere. So, the topic for today is – Exploiting intra-cycle parallelism for query processing . First, a quick background behind this line of thinking. If you zoom into the processor...