1 December 2012 in Review of Scientific Instruments
Gas puff imaging (GPI) offers a direct and effective diagnostic to measure the edge turbulence structure and velocity in the edge plasma, which closely relates to edge transport and instability in tokamaks. A dual GPI diagnostic system has been installed on the low field side on experimental advanced superconducting tokamak (EAST). The two views are up-down symmetric about the midplane and separated by a toroidal angle of 66.6°. A linear manifold with 16 holes apart by 10 mm is used to form helium gas cloud at the 130×130 mm (radial versus poloidal) objective plane. A fast camera is used to capture the light emission from the image plane with a speed up to 390 804 frames/s with 64×64 pixels and an exposure time of 2.156 μs. The spatial resolution of the system is 2 mm at the objective plane. A total amount of 200 Pa.L helium gas is puffed into the plasma edge for each GPI viewing region for about 250 ms. The new GPI diagnostic has been applied on EAST for the first time during the recent experimental campaign under various plasma conditions, including ohmic, L-mode, and type-I, and type-III ELMy H-modes. Some of these initial experimental results are also presented.
2 February 2022 in Wireless Communications and Mobile Computing
To meet the changing real-time edge engineering application requirements of CNN, aiming at the lack of universality and flexibility of CNN hardware acceleration architecture based on ARM+FPGA, a general low-power all pipelined CNN hardware acceleration architecture is proposed to cope with the continuously updated CNN algorithm and accelerate in hardware platforms with different resource constraints. In the framework of the general hardware architecture, a basic instruction set belonging to the architecture is proposed, which can be used to calculate and configure different versions of CNN algorithms. Based on the instruction set, the configurable computing subsystem, memory management subsystem, on-chip cache subsystem, and instruction execution subsystem are designed and implemented. In addition, in the processing of convolution results, the on-chip storage unit is used to preprocess the convolution results, to speed up the activation and pooling calculation process in parallel. Finally, the accelerator is modeled at the RTL level and deployed on the XC7Z100 heterogeneous device. The lightweight networks YOLOv2-tiny and YOLOv3-tiny commonly used in engineering applications are verified on the accelerator. The results show that the peak performance of the accelerator reaches 198.37 GOP/s, the clock frequency reaches 210 MHz, and the power consumption is 4.52 w under 16-bit width.
1 January 2022 in Infrared Physics & Technology