To mitigate the performance gap between the CPU and the main memory, multi-level cache architectures are widely used in modern processors. Therefore, modeling the behaviors of the downstream caches becomes a critical part of the processor performance evaluation in the early stage of Design Space Exploration (DSE). In this paper, we propose a fast and accurate L2 cache reuse distance histogram model, which can be used to predict the behaviors of the multi-level cache architectures where the L1 cache uses the LRU replacement policy and the L2 cache uses LRU/Random replacement policies. We use the L1 reuse distance histogram and two newly proposed metrics, namely the RST table and the Hit-RDH, that describing more detailed information of the software traces as the inputs. The output of our model is the L2 cache reuse distance histogram, based on which the L2 cache miss rates can be evaluated. We compare the L2 cache miss rates with the results from gem5 cycle-accurate simulations of 15 benchmarks chosen from SPEC2006. The average absolute error is less than 5%, while the evaluation time can be sped up by 7.38 times. We also extend our model into a multi-core architecture in which two cores share a unified L2 cache. The error of our model in this case is less than 7%.
To mitigate the ever worsening "Power wall" and "Memory wall" problems, multi-core architectures with multilevel cache hierarchies have been widely used in modern processors. However, the complexity of the architectures makes the modeling of shared caches extremely complex. In this paper, we propose a data-sharing aware analytical model for estimating the miss rates of the downstream shared cache in a multi-core environment. Moreover, the proposed model can also be integrated with upstream cache analytical models with the consideration of multi-core private cache coherent effects. The integration avoids time-consuming full simulations of the cache architecture, which are required by conventional approaches. We validate our analytical model against gem5 simulation results under 13 applications from PARSEC 2.1 benchmark suites. We compare the L2 cache miss rates with the results from gem5 under 8 hardware configurations including dual-core and quad-core architectures. The average absolute error is less than 2% for all configurations. After integrated with the upstream model, the overall average absolute error is 8.03% in 4 hardware configurations. As an application case of the integrated model, we also evaluate the miss rates of 57 different cache configurations in multi-core and multi-level cache scenarios.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.