cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

State-of-the-art AI at Alibaba is Powered by Xilinx

Xilinx Employee
Xilinx Employee
1 0 1,599

At XDF China in December 2019, Mr. Jiansong Zhang, Staff Engineer of Alibaba, gave a great talk about AI Platforms and Heterogenous Computing.

Alibaba developed a deep learning stack on top of Xilinx FPGAs including IP, shell, runtime, driver, compiler and models to enable various AI workloads.

ali1.png

This solution features -

  • Optimized and customized hardware design
    • On-chip streaming structure
    • 3D-systolic-array conv. engine which makes full use of DSP carry chain and supertile design @ 600MHz
    • Configurable dimensions of parallelism
  • Resource allocation & task scheduling using a runtime
  • Software-hardware co-optimization in compiler
  • Model parsers for ONNX and Tensorflow

Mr. Zhang also introduced the following 4 key use cases to demonstrate their excellent results.

Case 1: OCR (Optical Character Recognition) in Public Cloud Services

  ali2.PNG

 

Case 2: Edge solution for Smart Retail

ali3.png

 

Case 3: Private Cloud Service

~7x TCO saving achieved in replacement of CPU servers

ali4.png

 

Case 4: Speech Synthesis

Speech synthesis is an iterative task in which 16,000 iterations are needed to generate one second of audio. NN-based TTS (Text-to-speech) can be indistinguishable from human speech. Alibaba developed a Xilinx FPGA-based solution for real-time WaveNet, a state-of-art NN model for TTS. With the customized autoregressive low-latency IP in hardware and customized on-chip loop implemented in a compiler, they achieved 150x speed-up compared to a GPU implementation!ali5.png

ali6.PNG

This is another great example of adaptable AI inference powered by Xilinx!

Tags (3)