As more processing cores are added to embedded systems processors, the relationships between cores and memories have more\r\ninfluence on the energy consumption of the processor. In this paper, we conduct fundamental research to explore the effects\r\nof memory sharing on energy in a multicore processor. We study the Memory Arrangement (MA) Problem. We prove that\r\nthe general case of MA is NP-complete. We present an optimal algorithm for solving linear MA and optimal and heuristic\r\nalgorithms for solving rectangular MA. On average, we can produce arrangements that consume 49% less energy than an all shared\r\nmemory arrangement and 14% less energy than an all private memory arrangement for randomly generated instances. For DSP\r\nbenchmarks, we can produce arrangements that, on average, consume 20% less energy than an all shared memory arrangement\r\nand 27% less energy than an all private memory arrangement.
Loading....