Next Generation Heterogeneous Compute Device Architectures
Next Generation Heterogeneous Compute Device Architectures
批准号:
238664-2012
负责人:
Moshovos, Andreas
金额:
$2.4万
依托单位:
依托单位国家:
加拿大
项目类别:
Discovery Grants Program - Individual
财政年份:
2016
资助国家:
加拿大
项目状态:
已结题
起止时间:
2016-01-01 至 2017-12-31
中文摘要
在几乎所有现代电子设备的核心,如计算机、智能手机、平板电脑、高清电视、游戏机和网络路由器,都至少有一个由存储系统支持的处理器。处理器是一种能够处理以数字表示的信息的设备,而存储系统则存储这些数字。处理器很有用,因为它们可以通过传感器(例如麦克风)和执行器(例如显示器)与环境交互。处理器在科学探索中也很有用,它们被用来评估复杂的物理现象模型,如化学和行星间的相互作用,这些模型过于昂贵或不切实际,无法进行其他研究。今天,许多日常活动和许多科学探索都依赖于优化的廉价计算设备的可用性。从历史上看,在降低成本的同时不断改进处理器和存储器的特性(例如,处理速度和容量)是可能的。在80年代早期,一台典型的计算机要花费几万美元,而今天只需几百美元,却比它强大1000多倍。通信和计算行业的大部分都依赖于这一趋势。改进处理器和存储器的一个关键机制是计算机体系结构,它研究如何在给定现有制造技术的情况下构建这些设备,同时考虑到这些设备将用于的应用。计算机体系结构面临着持续的挑战,因为底层制造技术和应用程序的属性随着时间的推移发生了重大变化。由于某些潜在制造技术趋势的急剧放缓,处理器和内存设计今天处于关键的交叉点。本研究回顾了我们如何构建处理器和存储系统,以减轻这些挑战。它利用了新应用程序的特性,这些新应用程序是随着处理器嵌入到除传统计算机之外的许多设备中以及多媒体和面向互联网的应用的出现而出现的。从智能手机到超级计算机,各种各样的设备都将从这项研究中受益。
英文摘要
At the core of virtually all modern electronic devices such as computers, smartphones, tablets, high-definition TVs, game consoles, and network routers there is at least one processor supported by a memory system. A processor is a device that can manipulate information represented as numbers and memory systems store such numbers. Processors are useful because they can interact with the environment via sensors (e.g., a microphone) and actuators (e.g., a display). Processors are also useful in scientific exploration where they are used to evaluate complex physical phenomena models such as chemical, and inter-planetary interactions that are too expensive or impractical to study otherwise. Today many everyday activities and much of scientific exploration rely on the availability of optimized, inexpensive computing devices. Historically, it was possible to continuously improve the characteristics of processors and memories (e.g., processing speed and capacity) while reducing cost. A typical computer in the early 80's cost several tens of thousands, while today just a few hundred dollars while being more than 1000 times as powerful. Much of the communications and computing industry relies on this trend. A key mechanism for improving processors and memories is computer architecture which studies how to build such devices given the available manufacturing technologies while taking into consideration the applications that these devices will be used for. Computer architecture faces continuous challenges since the properties of the underlying manufacturing technology and the applications change significantly over time. Processor and memory design is today at a critical junction due to a drastic slowdown in certain underlying manufacturing technology trends. This research revisits how we build processors and memory systems aiming at mitigating these challenges. It takes advantage of the properties of new applications that have emerged as processors are embedded in many devices besides conventional computers and with the emergence of multimedia and internet oriented uses. A variety of devices ranging from smart-phones to supercomputers will benefit from this research.
期刊论文(0)
专著(0)
科研奖励(0)
会议论文
Deep Learning Hardware: Enabling the next wave of applications and innovation
-
批准号:RGPIN-2017-06064
-
项目类别:Discovery Grants Program - Individual
-
资助金额:$6.85万
-
财政年份:2021
-
负责人:Moshovos, Andreas
-
依托单位:
NSERC COHESA: Computing Hardware for Emerging Intelligent Sensory Applications
-
批准号:485577-2015
-
项目类别:Strategic Network Grants Program
-
资助金额:$81.96万
-
财政年份:2021
-
负责人:Moshovos, Andreas
-
依托单位:
A Business / Market Opportunity Assessment for Hardware Concepts & Device Designs for Brain-Machine Interfacing
-
批准号:571002-2022
-
项目类别:Idea to Innovation
-
资助金额:$1.08万
-
财政年份:2021
-
负责人:Moshovos, Andreas
-
依托单位:
NSERC COHESA: Computing Hardware for Emerging Intelligent Sensory Applications
-
批准号:485577-2015
-
项目类别:Strategic Network Grants Program
-
资助金额:$81.96万
-
财政年份:2020
-
负责人:Moshovos, Andreas
-
依托单位:
Deep Learning Hardware: Enabling the next wave of applications and innovation
-
批准号:RGPIN-2017-06064
-
项目类别:Discovery Grants Program - Individual
-
资助金额:$3.42万
-
财政年份:2020
-
负责人:Moshovos, Andreas
-
依托单位:
Deep Learning Hardware: Enabling the next wave of applications and innovation
-
批准号:DGDND-2017-00010
-
项目类别:DND/NSERC Discovery Grant Supplement
-
资助金额:$2.91万
-
财政年份:2019
-
负责人:Moshovos, Andreas
-
依托单位:
NSERC COHESA: Computing Hardware for Emerging Intelligent Sensory Applications
-
批准号:485577-2015
-
项目类别:Strategic Network Grants Program
-
资助金额:$81.96万
-
财政年份:2019
-
负责人:Moshovos, Andreas
-
依托单位:
Deep Learning Hardware: Enabling the next wave of applications and innovation
-
批准号:RGPIN-2017-06064
-
项目类别:Discovery Grants Program - Individual
-
资助金额:$3.42万
-
财政年份:2019
-
负责人:Moshovos, Andreas
-
依托单位:
A novel system architecture for online operational analytics
-
批准号:463355-2014
-
项目类别:Strategic Projects - Group
-
资助金额:$12.69万
-
财政年份:2018
-
负责人:Moshovos, Andreas
-
依托单位:
Deep Learning Hardware: Enabling the next wave of applications and innovation
-
批准号:RGPIN-2017-06064
-
项目类别:Discovery Grants Program - Individual
-
资助金额:$3.42万
-
财政年份:2018
-
负责人:Moshovos, Andreas
-
依托单位:
NSERC COHESA: Computing Hardware for Emerging Intelligent Sensory Applications
-
批准号:485577-2015
-
项目类别:Strategic Network Grants Program
-
资助金额:$68.02万
-
财政年份:2018
-
负责人:Moshovos, Andreas
-
依托单位:
Deep Learning Hardware: Enabling the next wave of applications and innovation
-
批准号:DGDND-2017-00010
-
项目类别:DND/NSERC Discovery Grant Supplement
-
资助金额:$2.91万
-
财政年份:2018
-
负责人:Moshovos, Andreas
-
依托单位:
Phase I: laconic: a deep learning accelerator for energy efficiency and high-performance for edge devices
-
批准号:530383-2018
-
项目类别:Idea to Innovation
-
资助金额:$8.2万
-
财政年份:2018
-
负责人:Moshovos, Andreas
-
依托单位:
Deep Learning Hardware: Enabling the next wave of applications and innovation
-
批准号:RGPIN-2017-06064
-
项目类别:Discovery Grants Program - Individual
-
资助金额:$3.42万
-
财政年份:2017
-
负责人:Moshovos, Andreas
-
依托单位:
A business / market opportunity assessment for hardware concepts & device designs for deep learning (DL)
-
批准号:508578-2017
-
项目类别:Idea to Innovation
-
资助金额:$1.08万
-
财政年份:2017
-
负责人:Moshovos, Andreas
-
依托单位:
Deep Learning Hardware: Enabling the next wave of applications and innovation
-
批准号:DGDND-2017-00010
-
项目类别:DND/NSERC Discovery Grant Supplement
-
资助金额:$2.91万
-
财政年份:2017
-
负责人:Moshovos, Andreas
-
依托单位:
NSERC COHESA: Computing Hardware for Emerging Intelligent Sensory Applications
-
批准号:485577-2015
-
项目类别:Strategic Network Grants Program
-
资助金额:$72.85万
-
财政年份:2016
-
负责人:Moshovos, Andreas
-
依托单位:
An infrastructure for analysing the performance of computational imaging algorithms on modern hardware
-
批准号:500612-2016
-
项目类别:Engage Grants Program
-
资助金额:$1.82万
-
财政年份:2016
-
负责人:Moshovos, Andreas
-
依托单位:
A novel system architecture for online operational analytics
-
批准号:463355-2014
-
项目类别:Strategic Projects - Group
-
资助金额:$12.69万
-
财政年份:2015
-
负责人:Moshovos, Andreas
-
依托单位:
Funding for full-application preparation - COHESA: Computing Hardware for Emerging "Intelligent" Sensory Applications
-
批准号:490560-2015
-
项目类别:Strategic Network Grants Program
-
资助金额:$1.82万
-
财政年份:2015
-
负责人:Moshovos, Andreas
-
依托单位:
国内基金
海外基金
Next Generation Majorana Nanowire Hybrids
-
批准号:--
-
项目类别:--
-
资助金额:20万元
-
批准年份:2020
-
负责人:Panagiotis Kotetes
-
依托单位: