AI chips: FPGA

Generality provides flexibility at the cost of complexity

What is FPGA? (Optional)

Source
Source

FPGA in DL

Source
Source (Intel)
Source
Source Intel Stratix product line
Source
Source: Wikipedia
Source (Intel)
Source (Xilinx)
Source

Intel Vision Accelerator Design with Arria 10 FPGA (For AI inferencing)

Source: Intel
Source: Intel
Source
Source
Source
Source Intel (Programming FPGA with HDL)
Source
Source: Intel
  • Deep Learning Development Toolkit (DLDT) in optimizing and deploying ML models,
  • FPGA Deep Learning Acceleration (DLA) and bitstreams for DL,
  • Tools, libraries, and accelerators for OpenCV, OpenVX, OpenCL and Intel Media SDK,
  • Plugins for Intel devices, and
  • A model zoo with pre-trained models for DLDT.
Source
Source
  • Model optimizer and
  • Inference engine.
ResNet269 block before and after optimization
Stride optimization

Intermediate representation Quantization

Intel FPGA Deep Learning Acceleration (DLA) Suite

Source
Source
Source
Source
Source
Source
Source
Source (Groups of layers to be processed in the pipeline concept)
Source
Source
Source (Process the 3rd and 2nd last layers)
Source
(Source)
Source
Naming conventions for the bitstream file
Source
Source
Source
Source
Source
Source
Source

Optimization

Source
Source
Source
Source (Right diagram: Handle a batch of images to increase concurrency and decrease the loading of the weights)
Source
Source

Intel Stratix 10 NX FPGA (For AI inferencing)

Source
Source (Tensor block)

Intel Agilex FPGA

Source: Google
Source

Xilinx Versal

Source
Xilinx Versal
Xilinx Versal ACAP Functional Diagram
AI Engine Tile
Vitis AI

Credit & References

--

--

Deep Learning

Love podcasts or audiobooks? Learn on the go with our new app.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store