Task Scheduling on Adaptive Multi-Core

Abstract

Abstract-Multi-cores have become ubiquitous both in the general-purpose computing and the embedded domain. The current technology trends show that the number of on-chip cores is rapidly increasing, while their complexity is decreasing due to power and thermal constraints. Increasing number of simple cores enable parallel applications benefit from abundant threadlevel parallelism (TLP), while sequential fragments suffer from poor exploitation of instruction-level parallelism (ILP). Recent research has proposed adaptive multi-core architectures that are capable of coalescing simple physical cores to create more complex virtual cores so as to accelerate sequential code. Such adaptive architectures can seamlessly exploit both ILP and TLP.The goal of this paper is to quantitatively characterize the performance potential of adaptive multi-core architectures. Previous research have primarily focused on only sequential workload on adaptive multi-cores. We address a more realistic scenario where parallel and sequential applications co-exist on an adaptive multi-core platform. Scheduling tasks on adaptive architectures reveals challenging resource allocation problems for the existing schedulers. We construct offline and online schedulers that intelligently reconfigure and allocate the cores to the applications so as to minimize the overall makespan under the constraints of a realistic adaptive multi-core architecture. Experimental results reveal that adaptive multi-core architectures can substantially decrease the makespan compared to both static symmetric and asymmetric multi-core architectures.

A Study of Dynamic Phase Adaptation Using a Dynamic Multicore Processor

“…Previous work on Dynamic Multicore Processors includes Bahurupi [16,17], and CoreFusion [9]. These architectures use a standard ISA and either fetched xed sized instruction windows [9] or an entire basic blocks [16].…”

“…In the work of Pricopi et al [17], they show how dynamic recon guration is bene cial when it comes to scheduling tasks. However, they do not discuss any method of automatically deciding the optimal core-composition beyond a 4 core fusion.…”

Dependency aware ahead of time static scheduler for multicore

“…Task scheduling on adaptive multi-core [11] has both online and offline scheduling methods. Grudenic [13] summarizes the tools for parallel processing and different techniques for scheduling tasks on multicore processors.…”

“…According to this equation, task k is categorized as ready provided execution of tasks represented by all the non-zero columns of ( , ) is completed. Out of these tasks, the task having minimum execution time is given more priority and it is selected for execution as shown in equation (11). The reason for selecting a task with minimum execution time is that it can release other dependent tasks faster to make the scheduler more efficient.…”

“…There are many algorithms proposed in the past to improve the performance of scheduler. Every scheduling algorithm chose to optimize one or more parameters [2] like cache [3], memory [4], dependency information [5], [6] energy, number of stages [7], [8] task decomposition [9], [10], [11], [12] etc. However, very few schedulers use the dependency information to generate optimal schedule.…”