… of the execution time can be allocated for profiling at runtime … Our implementation of ATM
only shows speedup for four bench… This approach is limited to task-based programs while our …

… went through several modifications to improve execution time, performance, and energy …
an approximate task memoization (ATM) technique for run-time systems storing all the tasks in a …

… Approximate Task Memoization [2] is a runtime system for dynamic and approximate memoization
of tasks… This can be attributed to memoization being the only approximation technique …

… used for a wide range of tasks. Sensors that generate billions of … for computation and data
approximation on both the … computer systems spend much of their execution time and power …

… system for embedding reduction is a challenging task. If done naïvely using a dense array,
the cost of memoization … height denote runtime speedup without memoization. In Remapped, …

… ATM approach [19] performs runtime task memoization, relying on hashing functions to store
the task … decide whether to use memoization or execute the task. The programmer needs to …

… information captured by task dataflow programming models … in the run-time system of task
dataflow programming models to … of parallel applications in the runtime system and guides the …

… ATM employs both dynamic redundancy and approximation for a tasklevel computing that
utilizes the previous task … output-based approximate memoization function that provides …

… execution time of task pipelines in distributed CPPS with the help of speculative computing.
The system learns how to map a task … This approach, Approximate Task Memoization as they …

… ATM approach [68] performs runtime task memoization, relying on hashing functions to store
the task … decide whether to use memoization or execute the task. The programmer needs to …