A key barrier to cloud adoption is system resilience with regards to functionality, performance and security. Users sign a contract that holds providers accountable to the negotiated and desired level of service. We are investigating technologies spanning from static analysis and verification to online software bug detection and removal in cloud systems to lower the barrier for cloud adoption.
The combination of dwindling efficiencies in conventional digital platform technologies and higher demand have led to unsustainable electrical consumptions in data centers worldwide. We are investigating a three-pronged approach to holistic energy optimization in system stack layers and infrastructure to provide: strategic vertical integration, with transparency to energy utilization; specialization, to minimize energy for critical services; approximation, to trade off service fidelity and quality for less energy.
We are investigating open-source cloud applications and services to facilitate robust software and service development, to monitor sensor data, to benchmark platform efficiency, and to enforce privacy in the clouds.
Data is growing at a much faster rate than conventional improvements from technology requiring out-of-the-box solutions to data management. We are investigating technologies to maximize efficiency with in-memory data services, in-situ query processing on streamed data, and on-demand query engine customization using multi-objective compiler optimization.