Skip to main content
. Author manuscript; available in PMC: 2022 Jan 1.
Published in final edited form as: IEEE Trans Ultrason Ferroelectr Freq Control. 2020 Dec 23;68(1):164–177. doi: 10.1109/TUFFC.2020.3001848

Fig. 8.

Fig. 8.

Benchmarking for sparse matrix operation. (a) Computational time of CPU and GPU implementations for the sparse matrix multiplication and standard DAS using the NHP in vivo data set for integrated sample NT varying from 10 to 1500 (Z: 2000 samples, N: 128 elements, Nx: 128 lateral pixels, Nz: 2000 axial pixels). (b) Sample processing rate at the same conditions. (c) Computational time with NT = 30 for different sizes of FOV, where the lateral map size Nx remained constant equal to 128 pixels and the axial map size varied: Nz = 50, 100, 200, 500, 1000, and 2000 pixels. The benchmarking was performed offline in a Dell Precision T7910 workstation (dual processor Intel Xeon CPU E5–2650 v4 @ 2.20GHz, 128 GB of RAM) equipped with a GPU (NVIDIA Quadro P6000, 24 GB memory, 3840 cores, driver: 392.56) running MS Windows 10 Pro 64-bits and Matlab 2017b.