The need for low latency, real time response to data is pushing compute to the edge of the network where it can process and extract intelligence from data in real time. The rapidly evolving discipline of AI is being used to extract insight from the plethora of data being generated and this in turn is exponentially increasing the compute requirements. New and innovative compute architectures are evolving to meet this need but in a world of shrinking silicon process geometries and correspondingly increasing development costs, the derived devices find it increasingly difficult to produce a return on their enormous development costs. FPGAs have long been used to substitute for the high cost of custom silicon creation but traditionally, FPGAs have prohibitively high part prices and power consumption that prevents them from being used in volume production.
Efinix has solved this impasse with its Quantum technology. Quantum delivers efficient, low power FPGA devices priced for high volume production. For the first time, an application can be taken from concept, through prototype phase and into mass deployment using the same efficient device. Years can be taken off the development cycle and many tens of millions of development cost saved. Furthermore, the agility of FPGA design means that the design, its algorithms, features and AI capabilities can be adapted to meet system requirements right up to, and even after deployment. Using Quantum Acceleration, hardware / software partitioning can be explored leaving the majority of the design in familiar software constructs while performance critical portions are implemented within accelerators at hardware speed.
Efinix is redefining product development at the edge. High performance, low power and flexible devices are now available with fast time to market and low development cost. This reality will speed time to innovation for the edge AI community leapfrogging traditional design approaches.
This talk will provide an overview of this revolutionary capability from Efinix.
The future is now