Context and Scope

As the sun sets on the first wave of cloud computing, characterized largely by homogeneity and large scale consolidated resources, it is time to consider the exciting challenges that will arrive with the next wave. With the tenability of Moore's law and Dennard scaling in the near to medium term being widely questioned, a new cloud comprised of a more heterogeneous mix of interconnected compute units must be considered as a possibility. Add to this advances in Open Source Hardware which will significantly reduce the cost of developing bespoke processors and it becomes a credible proposition rather than an academic curiosity. Then consider the proliferation of compute workloads more suited to today's GPUs or FPGAs, many of which will be driven by data collected from the edge of the network and it becomes clear that the cloud must evolve to become more Heterogeneous and Distributed. Indeed, the established players have taken initial steps in this direction but there is still much work to do to fully realize the potential of these new architectures which are qualitatively different from what has come before.

The goals of this workshop are to understand how today's cloud systems will evolve in this direction, to understand what they will ultimately become and to understand what this means for both cloud users and applications and more generally to the industry itself.

Topics of interest include:

• Novel cloud based applications that can leverage heterogeneity and distributed computing capability
• Application deployment techniques and mechanisms for more distributed, heterogeneous contexts
• Distributed storage solutions particularly solutions which understand data storage and access requirements in the distributed context
• Resource sharing between a complex mix of disparate compute and storage resources
• Management systems for aggregates of distributed, heterogeneous compute and storage resources
• Characterization of different compute, network and storage entities including large scale data centres, regional data centres, micro data centres
• Ownership models in heterogeneous distributed clouds
• Security solutions which support trusted application deployment over heterogeneous distributed resources
• Charging & Pricing models in heterogeneous distributed clouds
• Auto-scaling and scheduling over distributed and heterogeneous clouds
• Novel virtualization solutions for heterogeneous compute contexts

Intended Audience

The intended audience for this workshop can be broadly divided in two: (i) the classical cloud computing R&D community which is looking at new hardware compute platforms and (ii) those coming from the world of innovative hardware computing platforms and wanting to understand how they will impact the cloud computing sector. Given the practical and timely nature of the topic, the workshop will be of interest to both academia and industry.

Important Dates

• Paper submission deadline: 18th August 2017
• Acceptance notification: 10th September 2017
• Camera ready paper due: 22nd September 2017
• Workshop: 5-8th December 2017