Special Issue on Heterogeneous Computation in Specific Domain Accelerations (HC-SDA)
摘要截稿:
全文截稿: 2019-05-15
影响因子: 6.125
期刊难度:
CCF分类: C类
中科院JCR分区:
• 大类 : 计算机科学 - 2区
• 小类 : 计算机:理论方法 - 1区
Overview
Data are being generated at an unprecedented rate in the IoT era across various applications. How to process the Big Data in a timely manner is a major obstacle we are facing nowadays. The growing diversity and heterogeneity of the hardware platform just added another layer of difficulty on top of a challenging problem. Even though heterogeneous platform such a GPUs, Xeon Phis, and FPGAs, has been widely adopted. However, how to effectively and efficiently utilize different hardware accelerators together to serve one single application remains a challenge for heterogeneous computing researchers.
In this special issue, we seek original unpublished research on algorithms, models, applications and tools for heterogeneous computing to accelerate the performance, to improve energy efficiency, and to enhance reliability of heterogeneous platforms from edge to cloud and in between. We are particularly interested in heterogeneous computing research employing two or more different types of hardware accelerators.
Topics of interest include, but are not limited to, the following areas:
Microarchitecture design on heterogeneous processor/system combined with emerging memory/storage system (PCM, SSD, etc.)
Heterogeneous parallel programming paradigms and models.
Energy efficient parallel accelerating models for heterogeneous platforms.
Parallel algorithms for heterogeneous and/or hierarchical systems, including many-cores and hardware accelerators (FPGAs, GPUs, Xeon Phis, etc.)
Heterogeneous computation supports for autonomous vehicle driving, or other applications using artificial intelligent algorithms (e.g. images processing, features recognition, etc.)
Heterogeneous computing in large-scale datacenter.
Software engineering implementation on heterogeneous computing.
Multiple objectives optimization on heterogeneous platforms.
Task scheduling algorithms on heterogeneous computation, cloud and datacenter platforms.
Experience of porting parallel software from supercomputers to heterogeneous platforms.
Fault tolerance of parallel computations on heterogeneous platform.