1. 12
  1. 3

    A few highlights for me that illustrate great, software design compared to most analytic software:

    “Q and kdb+ are shipped as a single binary that has a small memory footprint and is capable of running in the L2 and L3 caches of modern CPUs making the system very performant. “

    “The x86 version for Linux is a 587 KB binary that can run without any real setup process on most flavours of Linux. There is also support for Solaris, Windows and Mac OSX.”

    “These query times are by far the fastest I’ve ever seen on any CPU-based system and the third query finished faster than on any other system I’ve ever benchmarked before.”

    “It’s also refreshing to not have a large configuration overhead when launching kdb+/q. A binary that just works well right off the bat and can be tweaked with a couple of flags is a welcome relief from hours of tuning desperate configuration files spread all over a cluster.”

    Stories with similar links:

    1. 1.1 Billion Taxi Rides on kdb+/q & 4 Xeon Phi CPUs via friendlysock 5 years ago | 17 points | 4 comments