Although being very fast and having relative good accuracy for estimating power consumption, the proposed analytical modeling presents some limitations for estimating execution time, which in turn affects the energy consumption estimation as shown in their experimental results [4].
In multiprocessor SoC performance evaluation, simulating the profiled or estimated execution times (or number of operations) of tasks on abstract HW resource models is an effective way of observing combined effects of task execution times, mapping, scheduling, and HW platform parameters on resulting task response times, response time jitters, and processing element utilizations.
As we illustrate in Fig. 10, the estimated execution time is very close to the actual, so applying this prediction model for estimating the jobs' execution time is a valid choice.
The debug run techniques run the job twice and they estimate the execution time of the second invocation using the execution time observed in the first run.
The way to estimate extra execution time using the reference assembly instruction set is also introduced.
It is also difficult for users to estimate application execution times, especially on heterogeneous environments.
Source code profiling is a technique to estimate the execution cost and memory cost of source code.
When the correlation between the frames of the same type is high, we do not use the profiler and just estimate by execution time of the previous frame.
Software-based solutions usually follow the general approach of classic profiling tools, like GPROF [17], the GNU statistical profiler able to estimate the execution time of functions that compose an application and to count the number of times they have been called.
We first evaluate the number of instructions of each step of a typical fingerprint verification algorithm, and estimate the execution time of several cryptographic algorithms to guarantee the security/privacy of the fingerprint data transmitted in the smart card with the client-server environment.
