基本信息
浏览量:0
职业迁徙
个人简介
Research interests
One of the advantages of today’s personal computers is that they’re able to operate a wide range of different software programs. The flipside of this flexibility, however, is that any given software program is unlikely to make use of all the available hardware. Dr David Boland’s research focuses on this issue, with the aim of achieving faster and more energy-efficient computing.
“I study the computer algorithms used to solve problems across many domains, including scientific computing, machine learning and optical communications.
“I then consider how to design fully customised hardware solutions to these problems. In doing this I focus on how to perform as little computation as necessary, while still obtaining results as accurate as those produced by the original algorithm used in the software. This helps to maximise energy efficiency. I also look at how to do as much parallel computation as possible, in order to maximise performance.
“The technology to support this exists: we can already create application-specific integrated circuits, and we can use field-programmable gate arrays on which custom hardware designs can be programmed after manufacturing. However, designing customised hardware architectures is challenging, so is currently restricted to experts. I’m interested in making this performance more widely accessible.
“My research to date has largely focused on accelerating common computational kernels and basic arithmetic structures. I’m currently working on creating custom hardware accelerators for machine learning within the context of analysing wireless communication systems.
“The ability to perform fast analysis and prediction using this information could be critical for effective military operations, for example. High-performance and energy-efficient computation is also critical for the development and widespread use of novel technologies such as artificial intelligence, which could transform society. The challenge is to develop design techniques that are of benefit both now and in the future.
“I’ve been working in this field since 2007, and I joined the University of Sydney in 2017. The University has provided me with more opportunities to engage with industry and to work alongside like-minded colleagues to maximise the impact of my research.”
One of the advantages of today’s personal computers is that they’re able to operate a wide range of different software programs. The flipside of this flexibility, however, is that any given software program is unlikely to make use of all the available hardware. Dr David Boland’s research focuses on this issue, with the aim of achieving faster and more energy-efficient computing.
“I study the computer algorithms used to solve problems across many domains, including scientific computing, machine learning and optical communications.
“I then consider how to design fully customised hardware solutions to these problems. In doing this I focus on how to perform as little computation as necessary, while still obtaining results as accurate as those produced by the original algorithm used in the software. This helps to maximise energy efficiency. I also look at how to do as much parallel computation as possible, in order to maximise performance.
“The technology to support this exists: we can already create application-specific integrated circuits, and we can use field-programmable gate arrays on which custom hardware designs can be programmed after manufacturing. However, designing customised hardware architectures is challenging, so is currently restricted to experts. I’m interested in making this performance more widely accessible.
“My research to date has largely focused on accelerating common computational kernels and basic arithmetic structures. I’m currently working on creating custom hardware accelerators for machine learning within the context of analysing wireless communication systems.
“The ability to perform fast analysis and prediction using this information could be critical for effective military operations, for example. High-performance and energy-efficient computation is also critical for the development and widespread use of novel technologies such as artificial intelligence, which could transform society. The challenge is to develop design techniques that are of benefit both now and in the future.
“I’ve been working in this field since 2007, and I joined the University of Sydney in 2017. The University has provided me with more opportunities to engage with industry and to work alongside like-minded colleagues to maximise the impact of my research.”
研究兴趣
论文共 60 篇作者统计合作学者相似作者
按年份排序按引用量排序主题筛选期刊级别筛选合作者筛选合作机构筛选
时间
引用量
主题
期刊级别
合作者
合作机构
IEEE SIGNAL PROCESSING LETTERS (2024): 646-650
IEEE Signal Process. Lett. (2024): 646-650
IEEE Signal Processing Letters (2024): 646-650
PROCEEDINGS OF THE 2023 ACM/SIGDA INTERNATIONAL SYMPOSIUM ON FIELD PROGRAMMABLE GATE ARRAYS, FPGA 2023pp.209-219, (2023)
2023 IEEE/ACM INTERNATIONAL CONFERENCE ON COMPUTER AIDED DESIGN, ICCADpp.1-9, (2023)
Mohammad Reza Jabbarpour,Bahman Javadi,Philip H. W. Leong, Rodrigo N. Calheiros,David Boland, Chris Butler
ACM Transactions on Reconfigurable Technology and Systemsno. 3 (2023): 1-27
Mohammad Reza Jabbarpour,Bahman Javadi,Philip Leong, Rodrigo N. Calheiros,David Boland, Chris Butler
Proceedings of the IEEE/ACM 16th International Conference on Utility and Cloud Computing (2023)
加载更多
作者统计
合作学者
合作机构
D-Core
- 合作者
- 学生
- 导师
数据免责声明
页面数据均来自互联网公开来源、合作出版商和通过AI技术自动分析结果,我们不对页面数据的有效性、准确性、正确性、可靠性、完整性和及时性做出任何承诺和保证。若有疑问,可以通过电子邮件方式联系我们:report@aminer.cn