Ztachip is a Multicore, Data-Aware, Embedded RISC-V AI Accelerator for Edge Inferencing running on low-end FPGA devices or custom ASIC.
Acceleration provided by ztachip can be up to 20-50x compared with a non-accelerated RISCV implementation on many vision/AI tasks. ztachip performs also better when compared with a RISCV that is equipped with vector extension.
An innovative tensor processor hardware is implemented to accelerate a wide range of different tasks from many common vision tasks such as edge-detection, optical-flow, motion-detection, color-conversion to executing TensorFlow AI models. This is one key difference of ztachip when compared with other accelerators that tend to accelerate only a narrow range of applications only (for example convolution neural network only).
A new tensor programming paradigm is introduced to allow programmers to leverage the massive processing/data parallelism enabled by ztachip tensor processor.
Ztachip consists of the following functional units tied via an AXI Bus to a VexRicsv CPU, a DRAM and other peripherals as follows
- The Mcore, a Scheduling Processor
- A Dataplane, to stream the next data and instruction to the Tensor Engine .
- A Scratch-Pad Memory to temporarily hold data
- A Stream Processor to manage data IO
- Tensor Engine with 28x Pcores that can be configured to act like a systolic array to perform in memory compute each containing a Scalar and Vector ALU, with 16 Threads of execution on private memory.
The software provided consists of
- Ztachip DSL C-like compiler
- AI vision libraries
- Application examples
- Micropython port and examples
.
├── Documentation Overview on HW/SW and programmer's guide for ztachip, pcore, visionai and tensor
├── HW Hardware
│ ├── examples Reference Design: Integration of Vexriscv, Ztachip, DDR3, VGA, Camera, LEDs & Buttons
│ ├── platform Memory IP depenedencies for different FPGA synthesis (e.g. XIlinx, Altera) or ASIC
│ ├── simulation RTL Simulation
│ └── src RTL of Ztachip's top design, Scalar/Vector ALU, Dataplane, Pcore, SoC integration etc
├── LICENSE.md
├── micropython Micropython Support
│ ├── examples edge_detection, image_classification, motion_detect, object_detect, point_of_interest etc
│ ├── micropython micropython
│ └── ztachip_port ztachip micropython port
├── README.md
├── SW Software
│ ├── apps AI kernel libraries of canny edge detector, harris corner, neural nets, optical flow etc
│ ├── base C runtime zero, Ztachip application libraries and other utilities
│ ├── compiler Ztachip C-like DSL compiler that generates instructions for the tensor processor
│ ├── fs File for data inference to be downloaded together with the build image
│ ├── linker.ld linker script for Ztachip
│ ├── makefile Main project makefile
│ ├── makefile.kernels Kernel makefile
│ ├── makefile.sim Makefile to test Kernels
│ ├── sim C source to test kernels
│ └── src SW Main (visionai and unit test entry points), SoC drivers and Zta's micropython API
│ This is a good place to learn on how to use ztachip prebuilt vision and AI stack.
└── tools openocd and vexriscv interface descriptions
In HW/platform, a generic implementation is also provided for simulation environment. Any FPGA/ASIC can be supported with the appropriate implementation of this wrapper layer. Choose the appropriate sub-folder that corresponds to your FPGA target.
Also, in SW/apps, many prebuilt acceleration functions are provided to provide programmers with a fast path to leverage ztachip acceleration. This folder is also a good place to learn on how to program your own custom acceleration functions.
The build procedure produces 2 seperate images.
One image is a standalone executable where user applications are using ztachip using a native [C/C++ library interface] (https://github.com/ztachip/ztachip/raw/master/Documentation/visionai_programmer_guide.pdf)
The second image is a micropython port of ztachip. With this image, applications are using ztachip using a Python programming interface
sudo apt-get install autoconf automake autotools-dev curl python3 libmpc-dev libmpfr-dev libgmp-dev gawk build-essential bison flex texinfo gperf libtool patchutils bc zlib1g-dev libexpat-dev python3-pip
pip3 install numpy
The build below is a pretty long.
export PATH=/opt/riscv/bin:$PATH
git clone https://github.com/riscv/riscv-gnu-toolchain
cd riscv-gnu-toolchain
./configure --prefix=/opt/riscv --with-arch=rv32im --with-abi=ilp32
sudo make
git clone https://github.com/ztachip/ztachip.git
export PATH=/opt/riscv/bin:$PATH
cd ztachip
cd SW/compiler
make clean all
cd ../fs
python3 bin2c.py
cd ..
make clean all -f makefile.kernels
make clean all
You are required to complete the previous build procedure for standalone image even if your target image is micropython image. Below is procedure to build micropython image after you have completed the standalone image build procedure.
git clone https://github.com/micropython/micropython.git
cd micropython/ports
cp -avr <ztachip installation folder>/micropython/ztachip_port .
cd ztachip_port
export PATH=/opt/riscv/bin:$PATH
export ZTACHIP=<ztachip installation folder>
make clean
make
-
Download Xilinx Vivado Webpack free edition.
-
Create the project file, build FPGA image and program it to flash as described in FPGA build procedure
The following demos are demonstrated on the ArtyA7-100T FPGA development board.
-
Image classification with TensorFlow's Mobinet
-
Object detection with TensorFlow's SSD-Mobinet
-
Edge detection using Canny algorithm
-
Point-of-interest using Harris-Corner algorithm
-
Motion detection
-
Multi-tasking with ObjectDetection, edge detection, Harris-Corner, Motion Detection running at same time
To run the demo, press button0 to switch between different AI/vision applications.
Reference design example required the hardware components below...
Attach the VGA and Camera modules to Arty-A7 board according to picture below
Connect camera_module to Arty board according to picture below
If you are running ztachip's micropython image, then you need to connect to the serial port. Arty-A7 provides serial port connectivity via USB. Serial port flow control must be disabled.
sudo minicom -w -D /dev/ttyUSB1
Note: After the first time connecting to serial port, reset the board again (press button next to USB port and wait for led to turn green) since USB serial must be the first device to connect to USB before ztachip.
In this example, we will load the program using GDB debugger and JTAG
sudo apt-get install libtool automake libusb-1.0.0-dev texinfo libusb-dev libyaml-dev pkg-config
git clone https://github.com/SpinalHDL/openocd_riscv
cd openocd_riscv
./bootstrap
./configure --enable-ftdi --enable-dummy
make
cp <ztachip installation folder>/tools/openocd/soc_init.cfg .
cp <ztachip installation folder>/tools/openocd/usb_connect.cfg .
cp <ztachip installation folder>/tools/openocd/xilinx-xc7.cfg .
cp <ztachip installation folder>/tools/openocd/jtagspi.cfg .
cp <ztachip installation folder>/tools/openocd/cpu0.yaml .
Make sure the green led below the reset button (near USB connector) is on. This indicates that FPGA has been loaded correctly. Then launch OpenOCD to provide JTAG connectivity for GDB debugger
cd <openocd_riscv installation folder>
sudo src/openocd -f usb_connect.cfg -c 'set MURAX_CPU0_YAML cpu0.yaml' -f soc_init.cfg
Open another terminal, then issue commands below to upload the standalone image
export PATH=/opt/riscv/bin:$PATH
cd <ztachip installation folder>/SW/src
riscv32-unknown-elf-gdb ../build/ztachip.elf
Open another terminal, then issue commands below to upload the micropython image.
export PATH=/opt/riscv/bin:$PATH
cd <Micropython installation folder>/ports/ztachip_port
riscv32-unknown-elf-gdb ./build/firmware.elf
From GDB debugger prompt, issue the commands below This step takes some time since some AI models are also transfered.
set pagination off
target remote localhost:3333
set remotetimeout 60
set arch riscv:rv32
monitor reset halt
load
After sucessfully loading the program, issue command below at GDB prompt
continue
If you are running the standalone image, press button0 to switch between different AI/vision applications. The sample application running is implemented in vision_ai.cpp
If you are running the micropython image, Micropython allows for entering python code in paste mode at the serial port.
To use the paste mode, hit Ctrl+E then paste one of the examples to the serial port, then hit ctrl+D to execute the python code.
Hit any button to return back to Micropython prompt.
Click here for procedure on how to port ztachip and its applications to other FPGA/ASIC and SOC.
First build example test program for simulation. The example test program is under SW/apps/test and SW/sim
export PATH=/opt/riscv/bin:$PATH
cd ztachip
cd SW/compiler
make clean all
cd ..
make clean all -f makefile.kernels
make clean all -f makefile.sim
Then compile all RTL codes below for simulation
HW/src
HW/platform/simulation
HW/simulation
The top component of your simulation is HW/simulation/main.vhd
main:reset_in must be driven low for few clocks before going high.
main:clk_x2_main must be twice the speed of main:clk_main and in phase.
The main:led_out should blink everytime a test result is passed.
This project is free to use. You can open an issue or a discussion on github. But for business consulting and support, please contact us
Follow ztachip on Twitter