29 7.eight 0.12 A5 259 three.9 0.12 A6 246 four.1 0.13 A7 492 2.0 0.13 A8 140 7.1 0.Future Net 2021, 13,16 of120 A1 – (13,eight)2-Bromo-6-nitrophenol Protocol number of
29 7.eight 0.12 A5 259 three.9 0.12 A6 246 four.1 0.13 A7 492 two.0 0.13 A8 140 7.1 0.Future Online 2021, 13,16 of120 A1 – (13,eight)Number of Cores60 A8 – (13,four) 40 A6 – (4,8) A3 – (13,two) 20 A7 – (four,four)A4 – (eight,eight);A2 – (13,four)A5 – (8,four)0,two,4,6,0 eight,0 10,0 Frames per Second (FPS)12,14,16,Figure 9. The number of cores versus frames per second of every single configuration from the architecture. The graphs indicate the configuration as quantity of lines of cores and number of columns of cores).Table 9 presents the Tiny-YOLOv3 network execution instances on various platforms: Intel i7-8700 @ three.two GHz, GPU RTX 2080ti, and embedded GPU Jetson TX2 and Jetson Nano. The CPU and GPU final results have been obtained applying the original Tiny-YOLOv3 network [42] with floating-point representation. The CPU result corresponds for the execution of Tiny-YOLOv3 implemented in C. The GPU result was obtained from the execution of Tiny-YOLOv3 in the Pytorch environment working with CUDA libraries.Table 9. Tiny-YOLOv3 execution instances on numerous platforms. Software Version Floating-point Floating-point Floating-point Floating-point Fixed-point-16 Fixed-point-8 Platform CPU (Intel i7-8700 @ 3.two GHz) GPU (RTX 2080ti) eGPU (Jetson TX2) [43] eGPU (Jetson Nano) [43] ZYNQ7020 ZYNQ7020 CNN (ms) 819.2 7.5 140 68 FPS 1.2 65.0 17 1.two 7.1 14.The Tiny-YOLOv3 on desktop CPUs is too slow. The inference time on an RTX 2080ti GPU showed a 109 speedup versus the desktop CPU. Making use of the proposed accelerator, the inference occasions had been 140 and 68 ms, in the ZYNQ7020. The low-cost FPGA was 6X (16-bit) and 12X (8-bit) more rapidly than the CPU using a compact drop in PF-06454589 Technical Information accuracy of 1.4 and two.1 points, respectively. In comparison with the embedded GPU, the proposed architecture was 15 slower. The benefit of utilizing the FPGA is definitely the power consumption. Jetson TX2 includes a energy close to 15 W, though the proposed accelerator includes a power of about 0.5 W. The Nvidia Jetson Nano consumes a maximum of 10 W but is around 12slower than the proposed architecture. five.3. Comparison with Other FPGA Implementations The proposed implementation was compared with previous accelerators of TinyYOLOv3. We report the quantization, the operating frequency, the occupation of FPGA sources (DSP, LUTs, and BRAMs), and two performance metrics (execution time and frames per second). Furthermore, we considered three metrics to quantify how efficientlyFuture Online 2021, 13,17 ofthe hardware sources were becoming applied. Considering the fact that diverse options commonly possess a various number of sources, it is actually fair to consider metrics to somehow normalize the results just before comparison. FSP/kLUT, FPS/DSP, and FPS/BRAM determine the number of each and every resource that is certainly applied to make a frame per second. The higher these values, the greater the utilization efficiency of those sources (see Table ten).Table ten. Overall performance comparison with other FPGA implementations. [38] Device Dataset Quant. Freq. (MHz) DSPs LUTs BRAMs Exec. (ms) FPS FPS/kLUT FPS/DSP FPS/BRAM ZYNQZU9EG Pedestrian indicators 8 9.6 104 16 one hundred 120 26 K 93 532.0 1.9 0.07 0.016 0.020 18 200 2304 49 K 70 [39] ZYNQ7020 [41] [40] Ours ZYNQVirtexVX485T US XCKU040 COCO dataset 16 143 832 139 K 384 24.four 32 0.23 0.038 0.16 100 208 27.5 K 120 140 7.1 0.26 0.034 0.8 100 208 33.4 K 120 68 14.7 0.44 0.068 0.The implementation in [39] may be the only previous implementation using a Zynq 7020 SoC FPGA. This device has significantly fewer sources than the devices made use of inside the other works. Our architecture implemented in the exact same device was 3.7X and 7.4X more quickly, rely.