The Data Center AI Platform Supports industry-standard frameworks
You can bring your own trained model or start with one from our model zoo
Xilinx ML suite provides comprehensive optimization for optimal FPGA implementation, together with a runtime and hardware DSA
Target a Xilinx Alveo accelerator card, your own custom card, or FPGA-as-a-Service such as Amazon AWS
The Data Center AI Platform software and hardware overlay is called ML Suite
ML Suite provides a comprehensive AI/ML solution allowing you to read in models from supported frameworks, optimize them and map them to Xilinx infrastructure
The provided runtime and DSA allow you to benefit from Xilinx hardware acceleration, without needing to be an FPGA expert
xfDNN middleware is a high-performance software library with a well-defined API which acts as a bridge between deep learning frameworks such as Caffe, MxNet, Tensorflow, and xDNN IP running on an FPGA.
xfDNN software is currently the only available method for programming and using xDNN IP and assumes a system running SDAccel reconfigurable acceleration stack compliant system.
xfDNN not only provides simple Python interfaces to connect to high level ML frameworks, but also provides tools for network optimization by fusing layers, optimizing memory dependencies in the network, and pre-scheduling the entire network removing CPU host control bottlenecks
Once these optimizations are completed per layer, the entire network is optimized for deployment in a "One-Shot" execution flow.
xfDNN Quantizer enables fast, high-precision calibration to lower precision deployments to INT8 and INT16. These Python tools are simple to use.
Xilinx Data Center AI Platform supports a number of industry-standard frameworks, highlighted in the table below.
|TensorFlow is an open-source framework developed by Google.||✔|
|CAFFE is an open-source framework developed at UC Berkley.||✔|
|MXNet is an open-source framework developed by Apache.||✔|
|Darknet is an open-source framework developed by Joseph Redmon.||✔|
|Keras is an open source high-level API capable of running on top of several other frameworks.||✔|
|Onnx is an open-source graph model and standardized operator definition. It works in conjunction with several frameworks. It was created by Facebook and Microsoft.||Coming Soon|
The Xilinx Data Center AI Platform supports the AI/ML Models as shown below.
|Task||Design Example & Descriptions|
|Object Detection||Yolo v3 (ADAS Detection)|