10,000 publications from this institution
Serving deep neural networks in latency critical interactive settings often requires GPU acceleration. However, the small batch sizes typical in online inference results in poor GPU utilization, a potential performance gap which GPU resource sharing can address. In this paper, we explore several techniques to leverage both temporal and spatial multiplexing to improve GPU utilization for deep learning inference workloads. We evaluate the performance trade-offs of each approach with respect to resource-efficiency, latency predictability, and isolation when compared with conventional batched inference. Our experimental analysis suggests up to a 5x potential for improved utilization through the exploration of more advanced spatial and temporal multiplexing strategies. Our preliminary prototype of a dynamic space-time scheduler demonstrates a 3.23x floating-point throughput increase over space-only multiplexing and a 7.73x increase over time-only multiplexing for convolutions, while also providing better isolation and latency predictability.
Abstract For Abstract see ChemInform Abstract in Full Text.
As technologies continue to evolve at exponential rates, online platforms are becoming an increasingly salient social context for adolescents. Adolescents are often early adopters, savvy users, and innovators of technology use. This not only creates new vulnerabilities but also presents new opportunities for positive impact-particularly, the use of technology to promote healthy learning and adaptation during developmental windows of opportunity. For example, early adolescence appears to represent a developmental inflection point in health trajectories and in technology use in ways that may be strategically targeted for prevention and intervention. The field of adolescent health can capitalize on technology use during developmental windows of opportunity to promote well-being and behavior change in the following ways: (1) through a deeper understanding of the specific ways that developmental changes create new opportunities for motivation and engagement with technologies; (2) by leveraging these insights for more effective use of technology in intervention and prevention efforts; and (3) by combining developmental science-informed targeting with broader-reach technologic approaches to health behavior change at the population level (e.g., leveraging and changing social norms). Collaboration across disciplines-including developmental science, medicine, psychology, public health, and computer science-can create compelling innovations to use digital technology to promote health in adolescents.
The \nwork at the University of California, Berkeley, was supported \nby the Miller Institute for Basic Research in Science, by NSF grant AST 94-17213, and by grant GO-7505 from \nthe Space Telescope Science Institute, which is operated by \nthe Association of Universities for Research in Astronomy, \nInc., under NASA contract NAS 5-26555.
Some of the geometric problems of interest to molecular biologists have macroscopic analogues in the field of robotics. Two examples of such analogies are those between protein docking and model-based perception, and between ring closure and inverse kinematics. Molecular dynamics simulation, too, has much in common with the study of robot dynamics. In this paper we give a brief survey of recent work on these and related problems.
Abstract ChemInform is a weekly Abstracting Service, delivering concise information at a glance that was extracted from about 200 leading journals. To access a ChemInform Abstract of an article which was published elsewhere, please select a “Full Text” option. The original article is trackable via the “References” option.
Rural health workers in India do not always have the training, credibility or motivation to effectively convince clients to adopt healthy practices. To help build their efficacy, we provided them with messages on mobile phones to present to clients. We present a study which compared three presentations of persuasive health messages by health workers using a phone-based lecture-style message, a phone-based dialogic message that elicits user responses, or no additional aids. We found that dialogic messages significantly improve the quality of counseling sessions and increase discussion between health workers and clients; however, we did not statistically measure an effect of either phone-based message on health behavioral outcomes. We analyze these results in light of the challenges we faced and compromises we made through the research process due to the interplay of social, cultural and environmental realities, and discuss how these factors affect ICTD projects at large.
This memo describes the network element behavior required to deliver a guaranteed service (guaranteed delay and bandwidth) in the Internet. Guaranteed service provides firm (mathematically provable) bounds on end-to-end datagram queueing delays. This service makes it possible to provide a service that guarantees both delay and bandwidth. This specification follows the service specification template described in [1].
