SWRA667 January 2020 CC1312PSIP , CC1312R , CC1314R10 , CC1352P , CC1352P7 , CC1352R , CC2642R , CC2642R-Q1 , CC2652P , CC2652R , CC2652R7 , CC2652RB , CC2652RSIP
To showcase the power and speed benefit of using the hardware accelerated drivers in comparison to a software based implementation, we have run several benchmarks. The metrics of interest are:
To compute these metrics, we collected the following data points for each benchmark:
The comparison metrics Duration Improvement and Energy Efficiency Improvement both describe the relative performance increase obtained by using the hardware accelerated drivers compared to the pure software implementation. They are computed by dividing the mbed TLS duration or energy used by the equivalent accelerated driver value.
The larger the number, the more efficient it is to use the drivers.
All benchmarks were run on a SimpleLink multi-standard CC26x2R wireless MCU LaunchPad™ development kit (LAUNCHXL-CC26X2R1) running at 48 MHz and powered with 3.3 V. The benchmark project was compiled with SimpleLink CC13x2-CC26x2 SDK version 3.20.00.68. The mbed TLS library was built with mbed TLS version 2.7.9. The project and mbed TLS library were compiled with the TI ARM C/C++ 18.12.1.LTS compiler using -O3 optimization settings. Current measurements were taken using an Agilent N6705B Power Analyzer. Time measurements were made with a Saleae Logic 8 measuring GPIO signals. Results are comparable between the CC26x2 and CC13x2 device families.