The 1st International Workshop on Architecture for Graph Processing
In the age of “big data”, modern applications are increasingly data-intensive. Large-scale graph processing has been a resurgence of interest to gain new solutions and insights. This renewed interest spurred the software community to develop more efficient graph analytics processing frameworks at different scales, from distributed graph processing to single machine out-of-core systems to in-memory graph processing. However, the unique characteristics of graph processing and the need of processing bigger graphs pose significant challenges to architectures and memory/storage systems. The goal of the workshop is to bring together academic researchers and industry practitioners to share their insights from different perspectives (system, runtime/programming model, architecture, technology) on graph processing, with the hope of proposing the innovative and cross-layer approaches with the cutting-edge technologies. We will make AGP workshop an inclusive and interactive event that features invited keynotes and presentations.
The first AGP Workshop will cover a wide range of topics in graph processing architecture, such as GPUs, memory/storage, accelerators, nonconventional architecture and software/hardware interaction, etc. We specifically encourage the submissions that look forward to future and emerging graph applications, interactions of multiple system stacks, and the disruptive architectures with novel technology. However, all topics related to efficient graph processing architectures are welcome:
- Parallel graph processing on GPUs, clusters and multicores
- Out-of-core graph processing systems
- Efficient data movement and memory hierarchy
- In-memory/-storage graph processing
- Interactions between programming model, runtime system and architecture
- Novel architecture inspired by emerging technology
- Architectural supports for irregular/asynchronous applications
- Resilience in distributed graph processing
- Interactions between algorithms and programming model
- Graph processing accelerators
- Graph processing in IoT devices
- The paper must be formatted according to ACM Proceeding Double Column format available here.
- Page limit: 6 (minimum) to 8 (maximum) pages, including references and relevant artworks.
- Please prepare for double-blind review.
- Abstract Deadline: Monday, May 1st, 2017 (11:59pm PST)
- Full paper Deadline: Monday, May 1st, 2017 (11:59pm PST)
- Notification: Friday, May 26th, 2017
- Workshop: Saturday, June 24th, 2017