View Sidebar
SiSense makes big data analytics possible on a Chip

SiSense makes big data analytics possible on a Chip

11/20/2013 7:38 am0 comments

Thriving on the concept of ‘big data meets business intelligence,’ a company in Israel has worked out a way to allow enterprises to put analytics in cache memory of a CPU in order minimize cost of hardware purchase that enterprises incur. SiSense big data analytics on a chip If you were to ask someone, what is big data’s biggest drawback? They would outrightly claim it as ‘more hardware’meaning more expenditure! Since, organizations see their big data levels rise more than twice each year, they need ways to store this data. Not just store, but also process, search and analyze data from time to time. All this requires IT infrastructure, which involves a lot of cost. As an innovative stride forward to help organizations save considerably on this cost of hardware, SiSense, Big Data Analytics Companyhas designed a way to do analytics in the cache memory of Central Processing Unit – Analytics on a Chip i.e. The idea of rethinking business intelligence – with high speed and smaller sets on multicore processors began way back in April when SiSense received $10 million in Series B funding from Battery Ventures along with Opus Capital and Genesis Partners. SiSense has basically worked out a technique to make analytic software work on multicore processors, say a parallel computer cluster on a Chip. Explaining the efficiency of the system, SiSense CTO EldadFarkash says that their big data analytics on a chip system can run queries against almost 20 terabytes of data on the Chip. According to Farkash, their technology works on Intel and AMD multicore 64-bit architectures and is probably the first to support Intel’s new Haswell architecture. Explaining the applicable use of the technology in the future, Farkash says an analytics system like theirs would be something we will carry in the palm of our hands, it will be a system that will work on new age iPads and Android tablets with terabytes of storage. Seen from a broader prospective, SiSense could be a deviation from high-performance computing ofbig data on Hadoop Hive. This could really save enterprises a lot that they end up spending on maintaining their data and analyzing it using various platforms. SiSense makes for a good way to diverge big data analytics beyond Hadoop hives, your take? Via: InformationWeek

Leave a reply