Xilinx Vitis Platform Delivers Benefits of Adaptable Hardware to Software Developers, AI Engineers

October 1, 2019 Brandon Lewis

Xilinx has launched Vitis, a free and open-source development platform that automatically adapts software and algorithms to Xilinx hardware without the need for VHDL or Verilog expertise. Vitis packages hardware modules as software-callable functions, and is compatible with standard development environments, tools, and open-source libraries.

Vitis leverages a stack-based architecture and standard libraries. The base layer is comprised of a board and pre-programmed I/O. A second layer includes the Vitis core development kit, which integrates Xilinx’ open-source runtime library to facilitate data transfer between different domains and subsystems. The second layer also integrates development tools like compilers, analyzers, and debuggers, from Xilinx and others.

A third layer houses eight Vitis libraries that provide more than 400 open-source applications. The libraries allow developers to call pre-accelerated functions using standard APIs. They include:

  • AI
  • Basic Linear Algebra Subprograms (BLAS)
  • Solver
  • Security
  • Vision
  • Data Compression
  • Database
  • Quantitative Finance

Vitis AI, for instance, is a domain-specific, machine-learning plugin that optimizes Xilinx hardware for frameworks such as TensorFlow, Caffe, and PyTorch. According to Xilinx, the process of optimizing, compressing, and compiling trained AI models onto Xilinx ACAP or Zynq platforms requires roughly one minute.

Domain-specific architectures from partners like Illumina for GATK genome analysis, BlackLynx for big datta analytics, and other proprietary DSAs are also part of the Vitis AI offering.

Vitis will be available for download in November. A developer site will also launch that provides tutorials, documentation, etc. Interested parties can sign up to get alerts on the Vitis platform.

For more detailed information on the Vitis platform, visit www.xilinx.com/products/design-tools/vitis/vitis-platform.html.

 

About the Author

Brandon Lewis

Brandon Lewis, Editor-in-Chief of Embedded Computing Design, is responsible for guiding the property's content strategy, editorial direction, and engineering community engagement, which includes IoT Design, Automotive Embedded Systems, the Power Page, Industrial AI & Machine Learning, and other publications. As an experienced technical journalist, editor, and reporter with an aptitude for identifying key technologies, products, and market trends in the embedded technology sector, he enjoys covering topics that range from development kits and tools to cyber security and technology business models. Brandon received a BA in English Literature from Arizona State University, where he graduated cum laude. He can be reached by email at brandon.lewis@opensysmedia.com.

Follow on Twitter Follow on Linkedin Visit Website More Content by Brandon Lewis
Previous White Paper
Farm to Freezer, Real-Time Analytics Cut Food Waste
Farm to Freezer, Real-Time Analytics Cut Food Waste

More than 1.2 billion tons of food spoil each year before reaching the consumer.

Next Article
Micron Xccella Flash to Drive Xilinx Versal ACAP AI Acceleration Platforms

Xcella flash is built on a 45 nm NOR flash process, and meets AEC-Q100 requirements. The 256 MB to 2 GB dev...