InAccel provides FPGA-Accelerated ML suite for distributed systems as a fully integrated AMI/AFI that in used to speedup machine learning applications for Apache Spark™. The suite provides all the required APIs that allows the utilization of the Amazon F1 instances without any changes of the original code.
InAccel provides both the IP cores and a novel scalable resource manager that allows the scalability of the FPGA instances in the cloud or on-premise data centers utilizing Xilinx FPGA cards. InAccel accelerated ML-suite offers up to 3x speedup compared to multi-threaded contemporary processors and 2x lower OpEx.
Download the Datasheet to learn more about the Alveo Solution.