Ryft Virtual, a heterogeneous cloud-based version of its core Ryft ONE technology, enables AWS users to get fast, actionable insight from their data.
Ryft, a provider of heterogeneous compute applications, has announced a new partnership with Amazon Web Services (AWS). The new partnership will bring fast, simple data discovery to users of AWS’s new F1 FPGA-based heterogeneous cloud instance.
Users will now have access to Ryft ONE, which will enable them to get latency-free, actionable insights from their data at speeds that far exceed what is possible with traditional CPU-based infrastructures, according to Ryft. Along with the increased speed, users will get an easy-to-navigate interface.
In their December 6 announcement, Ryft said it is crucial that businesses embrace heterogeneous computing in order to solve the unique problems that today’s fast and varied data creates. However, FPGAs aren’t easy for a data scientist to understand, let alone an enterprise user, so an FPGA-based heterogeneous cloud just isn’t enough. That’s where the Ryft partnership comes in. It provides the supercharged analytics companies needs along with a virtual application that is simple to use, the company said.
“Amazon’s decision to put FPGAs in its cloud infrastructure is a powerful step forward to putting the vast benefits of heterogeneous compute in all hands,” said Des Wilson, CEO of Ryft. “With our history of delivering supercharged yet easy-to-use heterogeneous compute accelerators to a broad range of customers, Ryft’s virtual instance of the Ryft ONE is a natural fit to enable AWS’ cloud customers to gain value from this new offering.”
The initial release of Ryft Virtual will happen in the AWS Marketplace in Q1 of 2017. It will leverage Elasticsearch, a popular open-source search engine built into most logging, security and analytics systems. It’s optimized for full text querying and complex Levenshtein distance searches of up to a distance of two. For searches requiring more than that, Ryft Virtual enables additional capabilities, the company stated, including eliminating indexing requirements, searching any field in Elasticsearch JSON records, and speeding up searches to many gigs per second through true hardware parallelization.
A free trial is available here.
