Xilinx Unveils Vivado ML, an FPGA EDA Tool
Compilation times reduced by 5x and result quality improved by 10%.


Today's EDA designers are struggling with increasing design complexity.

Xilinx announced on the 24th the launch of Vivado® ML Edition, an FPGA EDA tool that reduces design time and costs based on machine learning (ML) optimization algorithms and a team-based design flow. Compared to the existing Vivado HLx Edition, Vivado ML Edition promises five times faster compilation times and an average of 10% higher result quality for complex designs.

“Machine learning can accelerate the design process and deliver quality benefits,” said Nick Ni, marketing director for Xilinx’s software and AI solutions division. “Vivado ML helps developers shorten design cycles and achieve higher productivity.”

Vivado ML Edition supports machine learning-based algorithms that accelerate design deadlines. This enables ML-based logic optimization, latency prediction, intelligent design execution, and automated strategies to reduce timing deadline iterations.

Xilinx introduced the "Abstract Shell" concept, allowing users to define multiple modules within a system to be compiled incrementally or in parallel. This reduces compilation times by an average of 5x and up to 17x compared to traditional full-system compilation times.

Additionally, Abstract Shell helps system integrators providing value-added solutions such as FPGA-as-a-Service protect their customers' IP by hiding critical design details outside the module.

Additionally, Vivado ML Edition enhances collaborative design with the Vivado IP Integrator, which enables modular design with its "Block Design Container" feature. This feature facilitates team-based design methodology and enables collaborative processing of large-scale designs across multiple development sites using a divide-and-conquer strategy.

Additionally, adaptive features like 'Dynamic Function eXchange (DFX)' enable more efficient use of silicon resources by dynamically loading custom hardware accelerators, such as runtime over-the-air (OTA). DFX can load design modules in milliseconds, enabling new use cases such as switching between different vision algorithms during frame processing in automotive applications or performing genome analysis by switching between different algorithms in real time during DNA sequencing.

The Vivado ML Edition Standard Edition is available for free. The Enterprise Edition starts at $2,995 MSPR (recommended retail price). More details can be found on our website .