Making sure embedded processors are up to the mark

Called the FPMark, the suite uses 10 kernels to generate 53 workloads, each of which self verify to ensure correct execution of the benchmark. The kernels include a mixture of general purpose algorithms, including Fast Fourier Transform and linear algebra, and complex algorithms, such as a neural network routine. "Until now, the industry has lacked a reliable, useful and consistent floating point benchmark," said EEMBC president Markus Levy. "In the same way that EEMBC CoreMark was intended to be a 'better Dhrystone', FPMark provides an improvement over the easily manipulated Whetstone and Linpack. The FPMark will expose and highlight the performance gains from innovations in FPU development in terms of real application performance." FPMark contains single (32bit) and double (64bit) precision workloads, as well as a mixture of small to large data sets to support microcontrollers to high end processors, respectively.