FPGA Acceleration for Big Data & AI Computing
Grovf offers products that accelerate big-data analytics and enable organizations to search and analyze the most critical data at speeds approaching real-time.
The Grovf platform takes advantage of the versatile FPGA HWs to accelerate the processing of computationally intensive big-data. We do this by accelerating middleware, so applications running on top of middleware get an automatic boost without any changes on the user application.
GCache is an FPGA-based key-value database to efficiently store and retrieve cached data. Network stacks are implemented on FPGA to receive data from fast speed networks completely bypassing any type of operating systems and software layers. Key-value store algorithms are implemented completely on FPGA to handle Memcached compatible operations. Direct connection with DRAM from FPGA is used to stream data to memory for storage and retrieve the data on demand.
Regular expressions are specially encoded text strings used as patterns for matching sets of strings. GRegeX is an implementation of a standard regular expression algorithm on FPGA chip achieving 12.8 GB/s throughput with a single IP core. A wide range of supported regular expression functions allows developers to configure desired rules which can be handled in a chip without reducing the throughput. The target applications include Smart firewalls, Security and Log Text analysis.
In-memory computing is taking over traditional disk-based platforms. This puts new requirements for memory size, speed, and cost. Grovf develops technology that extends the memory of the existing servers to several TBs per server while accelerating the in-memory computing workloads.