Monomorphism-based CGRA Mapping via Space and Time Decoupling
By: Cristian Tirelli, Rodrigo Otoni, Laura Pozzi
Potential Business Impact:
Makes computer chips faster and use less power.
Coarse-Grain Reconfigurable Arrays (CGRAs) provide flexibility and energy efficiency in accelerating compute-intensive loops. Existing compilation techniques often struggle with scalability, unable to map code onto large CGRAs. To address this, we propose a novel approach to the mapping problem where the time and space dimensions are decoupled and explored separately. We leverage an SMT formulation to traverse the time dimension first, and then perform a monomorphism-based search to find a valid spatial solution. Experimental results show that our approach achieves the same mapping quality of state-of-the-art techniques while significantly reducing compilation time, with this reduction being particularly tangible when compiling for large CGRAs. We achieve approximately $10^5\times$ average compilation speedup for the benchmarks evaluated on a $20\times 20$ CGRA.
Similar Papers
Mapping code on Coarse Grained Reconfigurable Arrays using a SAT solver
Hardware Architecture
Finds best way to speed up computer tasks.
An MLIR-based Compilation Framework for Control Flow Management on CGRAs
Software Engineering
Makes flexible chips run complex code faster
Re-thinking Memory-Bound Limitations in CGRAs
Hardware Architecture
Makes computers run complex tasks much faster.