Power and performance optimization in FPGA‐accelerated clouds

Abstract

Energy management has become increasingly necessary in data centers to address all energy-related costs, including capital costs, operating expenses, and environmental impacts.Heterogeneous systems with mixed hardware architectures provide both throughput and processing efficiency for different specialized application types and thus have a potential for significant energy savings. However, the presence of multiple and different processing elements increases the complexity of resource assignment. In this paper, we propose a system for efficient resource management in heterogeneous clouds. The proposed approach maps applications' requirement to different resources reducing power usage with minimum impact on performance. A technique that combines the scheduling of custom hardware accelerators, in our case, Field-Programmable Gate Arrays (FPGAs) and optimized resource allocation technique for commodity servers, is proposed. We consider an energy-aware scheduling technique that uses both the applications' performance and their deadlines to control the assignment of FPGAs to applications that would consume the most energy. Once the scheduler has performed the mapping between a VM and an FPGA, an optimizer handles the remaining VMs in the server, using vertical scaling and CPU frequency adaptation to reduce energy consumption while maintaining the required performance. Our evaluation using interactive and data-intensive applications compare the effectiveness of the proposed solution in energy savings as well as maintaining applications performance, obtaining up to a 32% improvement in the performance-energy ratio on a mix of multimedia and e-commerce applications.

Energy-Aware Real-Time Tasks Processing for FPGA-Based Heterogeneous Cloud

“…With the increasing complexity of cloud architecture, energy consumption also increases. Hence, the authors in [27], proposed an energyaware scheduling technique for CPU-FPGA based heterogeneous cloud platform. Proposed strategy first maps the VM applications to the custom hardware accelerators through a fine-grained hardware aware scheduler, which reduces the power consumption by optimal usages of FPGAs.…”

Power-Optimal Mapping of CNN Applications to Cloud-Based Multi-FPGA Platforms

“…Resource allocation is a well-studied problem for highperformance data centers with heterogeneous hardware (CPUs with Graphical Processing Unit (GPU) or FPGA accelerators). Tesfatsion et al [2] provide a resource management framework with a hardware scheduler and an optimizer for FPGAaccelerated clouds. Similar to our work, they split workloads into "chunks" run by Virtual Machines on CPUs and sharing FPGA accelerators.…”

Numerical investigation of supersonic transverse jet interaction on CPU/GPU system

“…As a result, the computation scale and computing efficiency are unsatisfactory. Fortunately, high-performance computing platforms such as graphic processing unit (GPU) [5], many integrated core (MIC) [6], and field programming gate array (FPGA) [7,8] have more efficient performance in huge data processing than CPU. The comparison of these platforms' advantages and disadvantages is beyond the scope of this work.…”