assistantship positions available in SeeLab
Energy consumption is a critical constraint in the design of modern computer systems. Research in SEE lab addresses energy efficiency in systems of all sizes, from sensor nodes to data centers, transport networks and power grids.
Portable systems, such as mobile embedded systems and wireless sensor networks, typically operate with a limited energy source such as batteries. The design process for portable systems is characterized by a trade-off between high performance and low power consumption, emphasizing the need to provide a good user experience while maximizing the efficient use of a limited energy source such as batteries. Power dissipation and temperature management are critical factors in lowering packaging and cooling costs, especially as heterogeneous systems and many-core processors become more prevalent.
Moreover, beyond 65-nm dimension process variations impacting the delay and the power of the circuits have become a critical issue in the design of very large scale integrated (VLSI) circuits using advanced CMOS technologies. Process variations cause performance degradation and energy unpredictability. In additions to variations, many aging and wear-out phenomena effects rise up causing reliability degradation and compromising the system lifetime.
On the other end, datacenters also require energy efficiency due the operating costs and environmental concerns related to commercial data centers. The annual energy cost of a large data center can be in the range of millions of dollars. At a server level, we study how the most pervasive air-cooling solutions with fans can generate vibrations, dramatically degrading the performance of data-intensive workloads and inflating total energy consumption of servers. At the cluster and datacenter level, our research explores green energy aware workload scheduling and the use of energy efficient optical switching technology to satisfy the increasing traffic demands of data centers.
The power industry is undergoing a transformation to a smarter energy distribution grid. Energy efficiency problems take different forms depending on the scale, from a single house to a microgrid with hundreds of buildings, to a large scale grid serving millions. Within a residential house, we investigate controlling and predicting appliances, HVAC, lighting and other loads, which play a critical role in energy efficiency of the house and its relation to the grid. On a larger scale, as part of one of the best and most unique microgrids in the US, we research the energy consumption of home automation; distributed generation with high variance; energy storage selection; electric vehicles; and load-flow control problems on the microgrid scale and beyond.
energy efficiency lab is part of Embedded
Systems and Software group at UCSD.
Old photos :