The main objective of the OPERA project is to improve energy efficiency in the IT sector by means of heterogeneous hardware and software architectures, capable of communicating through programmable boards (FPGA) to distribute the workload according to applications and server technology.
OPERA will develop an efficient workload distribution system to automatically and dynamically perform tasks at the most suitable time and on the most appropriate hardware architecture.
CSI-Piemonte is responsible for verifying the validity of the hardware and software equipment produced by OPERA's industrial partners in two real contexts:
- Datacenter in a Box (Truck)
- Virtual Desktop
Datacenter in a Box is the truck set up for the Civil Protection of the Metropolitan City of Turin that intervenes in critical situations (earthquakes, floods...) by providing a mobile satellite connection to ensure VoIP voice transmission and Internet and, moreover, using a drone to take pictures and shoot videos of the terrain. To date, there is a considerable amount of energy required to properly aggregate all the photos taken by the drone to obtain an orthophoto, a geometrically correct aerial photography with a high resolution that makes it possible to navigate within a 3D image. Thanks to OPERA, it will be possible to reduce the processing time of ortho-photos, thus reducing the consumption of electricity produced by the van. This lower fuel consumption will allow to work for a longer period of time or cover a longer distance with the vehicle.
As far as Virtual Desktop is concerned, OPERA will look for a solution for server-side virtualization that can host an equal number of users with reduced energy consumption.
All this will be realized through the development of a specific software that will enable to distribute the workload among the available hardware architectures based on the statistical data collected.
Furthermore, the research of the OPERA project is aimed at finding the right energy efficiency metrics, derived from the model sought and translated into the algorithms better suited to the circumstances.
Energy saving: efficiency in the relationship between workload and processing resources.
Space saving: it will be possible to concentrate the same potentialities in a smaller space.
Potential re-use of the three use-cases for public administrations in Piedmont.
Acquisition of specialized skills in the field of hardware and software energy consumption optimization, OpenStack and evaluation metrics.
Development of a protocol for the acquisition of requirements within a heterogeneous virtual working group, replicable in the future.