DAAC Workshops

Data centers are becoming complex, highly automated computing dynamos that are essential for all forms of large-scale processing including cloud, grid, parallel, business, scientific, and distributed computing. Recent advances in both the hardware and software aspects of data centers have drawn considerable attention to the needs for corresponding advances in development of high-performance analytics, automation, and control methods specifically aimed for the needs of large-scale data centers, and to the needs for continual improvement in support for HPC/cloud software stacks, grids, hyper-converged and container-based infrastructures and automation of associated methods.

Meanwhile, government/national laboratories, industrial and academic institutions have already begun to pay close attention to how to manage computational, storage, networking, and supporting infrastructure resources efficiently and to control these resources in real-time using automated technologies. Meeting these challenges increasingly requires customized and highly automated data center designs that are aware of their workloads for power, monitoring, error recovery, closely coupled with the needs of the applications to be supported. Data center designs can no longer be decoupled from the applications, frameworks, and management techniques for user workloads, and instead these must be closely coupled. Further advances of the design and operation of data center equipment to optimize components for operation at large/extreme-scale are also desired.

The Industry/University Joint International Workshop on Data-center Automation, Analytics, and Control (DAAC) series provides a forum to discuss fundamental issues on real-time operation of highly automated data centers, including methods to provision, debug, analyze and control data center equipment and to improve how machines are operated, monitored, and used. The objective of this workshop series is to promote and stimulate community's interactions to address some of most critical challenges in automation, analytics, and control specifically aimed for the needs of large-scale data centers in high-performance/high-end computing.

Topics covered include how software stacks are deployed and provisioned, how data is processed and transferred in real time, methods for error detection and automated workload transition, power measurement and management, as well as challenges in design and implementation of novel automated data center architectures and systems. New methods, techniques, hardware, software, and standards for data center automation and control will also be discussed. The workshop expects to attract attendees from academia, industry, and government national labs interested in forefront topics related to data center automated management, operation, optimization, monitoring, and maintenance.

• The fifth in the series, DAAC 2022, will be held in parallel with the 15th IEEE/ACM International Conference on Utility and Cloud Computing (UCC 2022) hosted in December 2022 in Vancouver, Washington, USA. This event is planed to be a working session coordinated with pre-conference activities with the goal to gather community input on an action plan with a special focus on Reducing HPC's Carbon Footprint. This event follows on from previous DAAC Workshops as well as related BoFs and workshops at the SuperComputing conferences, the most recent of which is the Birds of a Feather session on Addressing HPC's Carbon Footprint at SC'22.

• The fourth in the series was DAAC 2021 also with a special focus on Data Center Control for Renewable Energy Applications was held at the UCC 2021 conference December 6-9, 2021 in Leicester, UK. More information is available here.

• The third in the series was DAAC 2019 held Friday, November 22 at the Supercomputing 2019 conference in Denver, Colorado. It featured a very strong set of papers covering state-of-the-art topics in monitoring and visualization and a lively panel discussion with representatives from industry, university, and national lab participants. Further information is available here.

• The second in the series was DAAC 2018 held Monday, November 12 at the SuperComputing 2018 conference in Dallas, Texas. Further information is available here. An additional related tutorial was held at the UCC/BDCAT 2018 conferences Dec. 17-20, 2018 in Zürich, Switzerland.

• The 1st International Industry/University Workshop on Data-center Automation, Analytics, and Control (DAAC 2017) was held December 5, 2017 at the 10th IEEE/ACM International Conference on Utility and Cloud Computing (UCC 2017) and 4th International Conference on Big Data Computing, Applications and Technologies (BDCAT 2017) in Austin, Texas, USA.

If you are interested in hosting a workshop on data center automation, analytics, and control topics to be part of this series, please contact the organizers.