Need for speed
Instarea works with some of the biggest telco providers in Europe and the Middle East. Our marketing and analytics platform processes vast amounts of telco data in a web app and the response times need to be quick. It’s this necessity to query long tables filled with event records and display results in real time that spawned the development of a GPU accelerated database – the fastest databases for complex polygonal operations. We call it qikkDB.
We’ve utilized new hardware opportunities and integrated modern graphic cards that enable fast and efficient processing of vast amount of data. This year we have joined Nvidia’s Inception program designed to propel companies forward with powerful GPU tools and we’ve been very busy ever since.
The benchmarking results show the query speeds we’ve been able to achieve with our database.
Because of smart terms & conditions of most databases, we can’t compare ourselves to our competition, but we came up with some codenames for the big players and you can get an idea of how much faster can qikkDB process queries.
GiraffeDB is a leading GPU database, CatDB the leading columnar database and RacoonDB the leading relational database.
We already know qikkDB works great with telco events data (counting number of sim cards in certain area with certain criteria), but we cannot wait to speed up other industries and operations such as web logs (computing funnel analysis of certain website visitors), finance data, automotive data and many others.
Beta version of the database is available for testing. We are looking forward to use-cases where big datasets need to be processed in real time. If you’re interested, you can sign up on qikk.ly and get a download link.
Data story telling
One particular use case that we’re excited about is data story telling. Until now, big datasets had to be processed slowly and reports created subsequently. With qikkDB enormous datasets can be the basis of beautiful data stories with charts and infographics, all editable in real time.
Find out more about qikkDB and sign up for the beta at qikk.ly