Smart hill climbing for agile dynamic mapping in many-core systems

Fattah, Mohammad; Daneshtalab, Masoud; Liljeberg, Pasi; Plosila, Juha

doi:10.1145/2463209.2488782

Cited by 80 publications

(69 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Benefits of centralized NoC management methods over their distributed counterparts have been shown in some previous work [3,25]. We address the main problem of centralized methods, i.e.…”

Section: Introductionmentioning

confidence: 89%

Improving the performance of packet-switched networks-on-chip by SDM-based adaptive shortcut paths

2015

View full text Add to dashboard Cite

“…Benefits of centralized NoC management methods over their distributed counterparts have been shown in some previous work [3,25]. We address the main problem of centralized methods, i.e.…”

Section: Introductionmentioning

confidence: 89%

Improving the performance of packet-switched networks-on-chip by SDM-based adaptive shortcut paths

2015

View full text Add to dashboard Cite

“…Several resource allocation approaches have been proposed while following different policies. Most of these approaches map communicating tasks of each application close to each other such that communication overhead and power are reduced [2], [7], [8], [9], [11], [21], [23], [31], [34]. Some of these approaches also reduce computation power of the cores by employing voltage/frequency scaling [9].…”

Section: Related Workmentioning

confidence: 99%

Bubble Budgeting: Throughput Optimization for Dynamic Workloads by Exploiting Dark Cores in Many Core Systems

Wang

Singh

et al. 2018

IEEE Trans. Comput.

View full text Add to dashboard Cite

Abstract-All the cores of a many-core chip cannot be active at the same time, due to reasons like low CPU utilization in server systems and limited power budget in dark silicon era. These free cores (referred to as bubbles) can be placed near active cores for heat dissipation so that the active cores can run at a higher frequency level, boosting the performance of applications that run on active cores. Budgeting inactive cores (bubbles) to applications to boost performance has the following three challenges. First, the number of bubbles varies due to open workloads. Second, communication distance increases when a bubble is inserted between two communicating tasks (a task is a thread or process of a parallel application), leading to performance degradation. Third, budgeting too many bubbles as coolers to running applications leads to insufficient cores for future applications. In order to address these challenges, in this paper, a bubble budgeting scheme is proposed to budget free cores to each application so as to optimize the throughput of the whole system. Throughput of the system depends on the execution time of each application and the waiting time incurred for newly arrived applications. Essentially, the proposed algorithm determines the number and locations of bubbles to optimize the performance and waiting time of each application, followed by tasks of each application being mapped to a core region. A Rollout algorithm is used to budget power to the cores as the last step. Experiments show that our approach achieves 50% higher throughput when compared to state-of-the-art thermal-aware runtime task mapping approaches. The runtime overhead of the proposed algorithm is in the order of 1M cycles, making it an efficient runtime task management method for large-scale many-core systems.

show abstract

“…The global level optimization uses the Hill Climbing algorithm [20,21]. The algorithm moves each breakout to new locations on its adjacent nodes of the road map to get different breakout sets.…”

Section: Global Level Optimizationmentioning

confidence: 99%

A methodology to enable automatic 3D routing of aircraft Electrical Wiring Interconnection System

2017

View full text Add to dashboard Cite

Harness 3D routing is one of the most challenging steps in the design of aircraft Electrical Wiring Interconnection System (EWIS). This is due not only to the intrinsic complexity of the EWIS, but also to the increasing number of applying design constraints and its dependency on any change in the design of the airframe and installed systems. The current routing process employed by EWISdesign is largely based on the manual work of expert engineers, partially supported by conventional CAD systems. As a result, the routing process is quite inefficient, error prone and unable to deliver optimal solutions. Although many harness components are selected from catalogues and the design process is largely repetitive and rule based, it has been found that none or very limited automation solutions, which can significantly decrease the workload of engineers and increase their efficiency, are currently available. In this paper, an innovative approach is proposed to solve the 3D routing automation as an optimization problem. Knowledge Based Engineering (KBE) and optimization methods are proposed to achieve minimum cost routing solutions that satisfy all relevant design rules and constraints. The proposed solution is scalable in terms of constraints, can be deployed on any type of routing environment, and, thanks to the achieved level of automation, able to reduce the process lead time drastically. The basic idea is to achieve optimal EWIS routing solutions by optimizing the position of the harnesses clamping points, which are used as way-points to route the harnesses inside the aircraft digital mock-up. The challenge to solve this optimization problem is that the number and initial value of design variables, namely the number and position of clamping points, are not known a priori. To handle this challenge, a two-step, hybrid optimization strategy has been devised. The first step, called Initialization, uses a road map based path finding method to generate a preliminary harness definition, including the required number and preliminary position of its clamping points. The second step, called Refinement, uses a conventional optimization method to move the position of the clamps and refine the preliminary harness definition aiming for the minimum cost and the satisfaction of all the design constraints. This approach has been implemented into a KBE application connected with a commercial optimization package and tested on several routing cases. The results demonstrate that the proposed method is capable of handling cases of representative geometric complexity and design constraints and delivering proper 3D harness models in full automation.

show abstract

Smart hill climbing for agile dynamic mapping in many-core systems

Cited by 80 publications

References 21 publications

Improving the performance of packet-switched networks-on-chip by SDM-based adaptive shortcut paths

Improving the performance of packet-switched networks-on-chip by SDM-based adaptive shortcut paths

Bubble Budgeting: Throughput Optimization for Dynamic Workloads by Exploiting Dark Cores in Many Core Systems

A methodology to enable automatic 3D routing of aircraft Electrical Wiring Interconnection System

Contact Info

Product

Resources

About