Search

Predictors essentially predicts the most recent events based on the record of past events, history. It is obvious that prediction performance largely relies on regularity–randomness level of the history. This paper concentrates on extracting effective information from branch history, and discusses expected performance of branch predictors. For this purpose, this paper introduces entropy point-of-views for quantitative characterization of both program behavior and prediction mechanism. This paper defines four new entropies from different viewpoints; two of them are independent of prediction methods and the others are dependent on predictor organization. These new entropies are useful tools for analyzing upper-bound of prediction performance. This paper shows some evaluation results of typical predictors. [ABSTRACT FROM AUTHOR]

An ordinal interconnection network is composed of many independent routers that can cooperate as a communication subsystem in a massively parallel system. Many routing algorithms are proposed in the past, however, they do scarcely utilize global network information. In this paper, we propose a new adaptive routing method, Cross-Line, that makes efficient use of global information over the network. The algorithm achieves global routing control efficiently by collecting just one-bit information of each virtual channel in the x- and y-directions. Analytical and simulation results reveal the effectiveness of the algorithm. [ABSTRACT FROM AUTHOR]

The behavior of wild and hatchery-reared red tilefish Branchiostegus japonicus was analyzed using two different methods: video observation and acoustic telemetry tracking. In the laboratory, digging and swimming activity of seven wild and five hatchery-reared fish were recorded for 2-4 days in an experimental aquarium and related to changes in light intensity. The activity of wild individuals increased with light intensity, while hatchery-reared individuals were active during both day and night. In the field, 18 wild and 9 hatchery-reared fish were released and tracked using an acoustic telemetry system during the winter and the summer in Maizuru Bay, Kyoto (Japan). Seven wild and three hatchery-reared fish settled within 2 km of the release point for 21-200 days whilst 11 wild and 6 hatchery-reared fish were detected in the central part of the bay within 13 days. The settled fish demonstrated a diel pattern of behavior; most of them probably stayed outside their burrows in the daytime but inside their burrows during the nighttime. During the winter wild fish tended to stay near the release site for longer periods than the hatchery-reared fish. However, in summer the hatchery-reared fish tended to remain close to the release site for longer than the wild fish. Based on these two experiments, we suggest that the activity of wild red tilefish respond to changes in low levels of light intensity (0-0.1 μmol m−2 s−1) at dawn and display a diel behavior. Hatchery-reared fish may differ from wild fish with regard to their behavior, response to water temperature and light conditions. [ABSTRACT FROM AUTHOR]

A large-scale direct interconnection network usually consists of enormous number of simple routers. However, its behavior is sometimes very complicated. Such a complicated behavior prevents us from accurate understanding and efficient control of the network. Among serious problems in interconnection networks, congestion control is of extreme importance since network performance is drastically degraded by a congested situation. We focus our discussion on throttling, injection limitation in other words, as one of the most hopeful solutions to the congestion problem. Our approach is inspired from physics. We define entropy as a desirable metric for representing the network's congestion level. We also define packet mobility ratio as a proper approximation of entropy. Thus we reach a new throttling method called ‘Entropy Throttling' that is based on theoretical discussion on congestion. Evaluation results by our simulator reveal effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]