Abstruct-It has historically been ditficult to distribute a well-aligned hardware clock throughout the physical extent of a synchronous processor. 'haditionally, this task has been accomplished by distributing the output of a central oscillator over a tree-like network, with repeaters at necessary intervals. While straightforward in concept, this method suffers from poor reliability, poor scalability and high skew.In this paper, we present an alternative approach-Distributed Synchronous Clocking-that maintains the simplicity of synchronous operation without suffering the drawbacks of centralized clocking. A network of independent oscillators takes the place of the centralized clock source, providing separate clock signals to the physically distant parts of a computing system. A distributed error correction algorithm effects global phase alignment by utilizing local comparisons of neighboring oscillator phase.In contrast to centralized clock distribution, distributed clocking has the inherent potential for complete scalability and graceful degradation. However, because oscillator phase is a modular quantity, a naive implementation of distributed synchronous clocking can suffer from mode-lock-the trapping of local oscillator phase in undesirable stable equilibria where global phase is not aligned [%I. We present a simple method for eliminating this problem in k-ary Cartesian meshes and give a proof of its correctness for two-dimensional networks. An electronic implementation is also presented and several engineering issues relating to error tolerance are discussed.

Background:Patellar tendon tears impart potentially debilitating sequelae among professional basketball athletes.Hypothesis:Professional basketball athletes with patellar tendon tears have decreased return-to-play performance in seasons after injury compared with preinjury statistics.Study Design:Case series; Level of evidence, 4.Methods:Patellar tendon tears among National Basketball Association (NBA) athletes from the 1999-2000 to 2014-2015 seasons were identified. Player performance statistics for players who underwent operative patellar tendon repair were compared from 1 season before injury to 1 season after injury and 2 seasons before injury to 2 seasons after injury using the primary outcome of player efficiency rating (PER). Secondary performance outcomes were also analyzed.Results:A total of 13 patellar tendon tears (10 complete, 3 partial) were identified among 12 NBA athletes. Three players (25%) did not return to play in the NBA. No significant differences were found in PER in comparisons of 1 season before and after injury (16.6 ± 1.5 vs 14.3 ± 1.7; P = .20) or in comparisons of 2 seasons before and after injury (15.8 ± 0.8 vs 6.3 ± 2.3; P = .49). Diminished performance outcomes were noted for total minutes played (2598 ± 100 vs 1695 ± 78; P = .01), games played (74.8 ± 1.9 vs 60.5 ± 1.4; P = .04), and minutes per game (34.8 ± 1.5 vs 28.2 ± 1.8; P = .02) in comparisons of 1 season before and after injury. Total minutes played per season (2491 ± 190 vs 799 ± 280; P = .045) decreased in comparisons of 2 seasons before and after injury.Conclusion:Patellar tendon tears were not associated with diminished efficiency-adjusted performance, as measured by PER, games played, minutes per game played, points per 36 minutes, and rebounds per 36 minutes. However, decreases in total minutes played were observed following patellar tendon tear. Orthopaedic surgeons may be better prepared to counsel basketball athlete patients with patellar tendon tear given these findings.

No abstract

It is dificult to distribute a well aligned hardware clock throughout the physical extent of a synchronous processor. In this paper, we present an alternative approach -Distributed Synchronous Clocking -that maintains the simplicity of synchronous operation without sufering the drawbacks of centralized clocking. A network of independent oscillators takes the place of the centralized clock source, providing separate clock signals to the physically distant parts of a computing system. A distributed error correction algorithm &ects global phase alignment by utilizing local comparisons of neighboring oscillator phase.In contrast to centralized clock distribution, distributed synchronous clocking has the inherent potential for complete scalability and graceful degradation. However, because oscillator phase is a modular quantity, a naive implementation of distributed synchronous clocking can slrfferfrom mode-lockthe trapping of local oscillator phase in undesiruble stable equilibria where global phase is not aligned.1211 We present a simple method for eliminating this problem in k-ary Cartesian meshes and show an electronic implementation with good performance. In another paper[l4] we give a proof of this method's correctness for 2 dimensional networks and discuss tolerance issues in the electronic implementation.

s Abstract The interactivity on the Internet is Faster networks, faster processors, and standardized protocols have enabled the emergence of interactive applications running over commercial networks such as the Internet. In such applications, multiple users interact with one another by exchanging real-time information such as user position and orientation in a virtual world, live and recorded audio, video, and text. These applications include interactive shopping, team training, virtual meeting room, and multi-player games. However, to date, these interactive system have supported a limited number of information types, offered limited scalability, and have failed to account for a heterogeneous network and processor environment. In this paper, we describe the design and implementation of In Verse, an infrastructure that supports real-time interactive applications on the Internet, Inverse provides a common backplane for disseminating and managing multiple real-time data streams. Within this general-purpose structure, the Inverse system maximizes scalability by implementing a hybrid communications architecture that adapts itself to available network bandwidth, observed network latency, installed network security measures, and available services such as multicast.