Well, this is something I wasn’t expecting, but it does make a lot of sense. As reported by Dylan McGrath the folks at Xilinx have just announced the acquisition of the high level synthesis company AutoESL Design Technologies, Inc. (Click Here to see Dylan’s article).
AutoESL's flagship high level synthesis tool, AutoPilot, allows users to capture their designs at a high level of abstraction in C, C++ or SystemC. The AutoPilot tool is optimized for Xilinx FPGA architectures and intelligently generates register transfer-level (RTL) code that produces the best possible QoR to meet throughput, power, area, and timing design goals.
In addition to dramatically speeding design capture and verification, this type of technology opens up the use of FPGAs to a much wider range of users beyond traditional hardware design engineers.
In many ways this is “a match made in heaven,” because AutoPilot has already been adopted by leading semiconductor and systems companies to enhance productivity and speed time-to-market for video, wireless, and high performance computing applications, of whom more than 25 are already Xilinx customers and Alliance Program members.
In addition to AutoPilot’s use with traditional FPGAs, one area of particular interest will be the forthcoming Xilinx Extensible Processing Platform, which combines 28 nm programmable fabric with a dual-hard-core ARM Cortex-A9 processor. A tight integration of AutoPilot with the Xilinx tool suite will allow users to capture a complete system in C/C++; and to then run the control portions of the program on the ARM processors and to translate the dataflow portions of the program into RTL for hardware acceleration in the programmable fabric.
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.