No abstract

No abstract

Abstract-Heterogeneous architectures consisting of generalpurpose CPUs and throughput-optimized GPUs are projected to be the dominant computing platforms for many classes of applications. The design of such systems is more complex than that of homogeneous architectures because maximizing resource utilization while minimizing shared resource interference between CPU and GPU applications is difficult. We show that GPU applications tend to monopolize the shared hardware resources, such as memory and network, because of their high thread-level parallelism (TLP), and discuss the limitations of existing GPU-based concurrency management techniques when employed in heterogeneous systems. To solve this problem, we propose an integrated concurrency management strategy that modulates the TLP in GPUs to control the performance of both CPU and GPU applications. This mechanism considers both GPU core state and system-wide memory and network congestion information to dynamically decide on the level of GPU concurrency to maximize system performance. We propose and evaluate two schemes: one (CM-CPU) for boosting CPU performance in the presence of GPU interference, the other (CM-BAL) for improving both CPU and GPU performance in a balanced manner and thus overall system performance. Our evaluations show that the first scheme improves average CPU performance by 24%, while reducing average GPU performance by 11%. The second scheme provides 7% average performance improvement for both CPU and GPU applications. We also show that our solution allows the user to control performance trade-offs between CPUs and GPUs.

In this paper, we present techniques that coordinate the thread scheduling and prefetching decisions in a General Purpose Graphics Processing Unit (GPGPU) architecture to better tolerate long memory latencies. We demonstrate that existing warp scheduling policies in GPGPU architectures are unable to effectively incorporate data prefetching. The main reason is that they schedule consecutive warps, which are likely to access nearby cache blocks and thus prefetch accurately for one another, back-to-back in consecutive cycles. This either 1) causes prefetches to be generated by a warp too close to the time their corresponding addresses are actually demanded by another warp, or 2) requires sophisticated prefetcher designs to correctly predict the addresses required by a future "far-ahead" warp while executing the current warp.We propose a new prefetch-aware warp scheduling policy that overcomes these problems. The key idea is to separate in time the scheduling of consecutive warps such that they are not executed back-to-back. We show that this policy not only enables a simple prefetcher to be effective in tolerating memory latencies but also improves memory bank parallelism, even when prefetching is not employed. Experimental evaluations across a diverse set of applications on a 30-core simulated GPGPU platform demonstrate that the prefetch-aware warp scheduler provides 25% and 7% average performance improvement over baselines that employ prefetching in conjunction with, respectively, the commonlyemployed round-robin scheduler or the recently-proposed two-level warp scheduler. Moreover, when prefetching is not employed, the prefetch-aware warp scheduler provides higher performance than both of these baseline schedulers as it better exploits memory bank parallelism.