No abstract

We present a run-time system for a multi-grained reconfigurable processor in order to provide a dynamic trade-off between performance and available area budgets for both fine-as well as coarse-grained reconfigurable fabrics as part of one reconfigurable processor. Our run-time system is the first implementation of its kind that dynamically selects and steers a performance-maximizing multi-grained instruction set under run-time varying constraints. It achieves a performance improvement of more than 2x compared to state-of-the-art run-time systems for multi-grained architectures. To elaborate the benefits of our approach further, we also compare it with offline-and onlineoptimal instruction-set selection schemes.

We propose a novel scheme for run-time management of mixedgrained reconfigurable fabric for the purpose of simultaneous multi-tasking in multi-core reconfigurable processors. Traditionally, reconfigurable fabrics are allocated to distinct tasks in order to improve the overall performance without considering quality of service of the entire application (e.g. 30 fps in video conferencing application). We employ a new concept of task criticality that is based on performance constraints of each task at functional block level. Our scheme significantly reduces the number of deadline misses by dynamically evaluating the criticality of each task and performing an efficient load balancing of a mixed-grained reconfigurable fabric at run-time. We use a comprehensive video conferencing application as a benchmark to evaluate our scheme. Compared to the state-of-the-art our scheme reduces the number of deadline misses by 6x (on average) and improves the performance by 1.3x (on average).