As the data center workload exhibits plenty of diversities and totally different behaviors in comparison with traditional HPC and desktop workloads, more and more researchers are realizing the significant mismatch between existing data center hardware platform and their workloads. The designers of data center server need to understand the behaviors of such dynamically changing workloads, identify their common requirements and predict their trends. On the other hand, data center system is facing more and more pressures from the power consumption, cost control, as well as the area limitation. The goal of this workshop is to bring together academic researchers and industry practitioners to share their insights from data center workloads, exchange their understanding on the common behaviors and potential trends of such real workloads, and discuss the possible explorations on architectural innovations for future data center system. So the topic of this round workshop is about efficient data center system, the scope of which covers the power efficiency, cost efficiency and space efficiency etc.
This workshop in a series is about data center system. This is the 3rd one, which is focus on the workload analysis and architecture exploration for data center server. The 1st round held in conjunction with HPCA 2012 specialized its topic on the Micro-Server system. And the 2nd round held in conjunction with HPCA 2013 is focus on data center infrastructure, cloud deployment and DC application analysis.
Topics of interest include but are not limited to:
- Big data related topics, e.g., workload analysis, big data system design etc.
- Virtualization platform in data center system
- Power efficient data center system architecture
- High density data center server
- Cost effective data center architecture
- Security architecture for data center server
- Efficient processor architecture for data center server
- Hybrid system architecture and related software issues (scheduling, programming etc)
- New architecture exploration inspired by emerging techniques, such as storage class memory, and on-chip optics etc.
We encourage both industry and academic researchers to submit their research. Preliminary results of interesting ideas and work-in-progress are welcome. Submissions that are likely to generate vigorous discussion will be favored!