Search for: All records

Computer systems research heavily relies on simulation tools like gem5 to effectively prototype and validate new ideas. However, publicly available simulators struggle to accurately model systems as architectures evolve rapidly. This is a major issue because incorrect simulator models may lead researchers to draw misleading or even incorrect conclusions about their research prototypes from these simulators. Although this challenge pertains to many open source simulators, we focus on the widely used, open source gem5 simulator. In GAP we showed that gem5’s GPGPU models have significant correlation issues versus real hardware. GAP also improved the fidelity of gem5’s AMDGPU model, particularly for cache access latencies and bandwidths. However, one critical issue remains: our microbenchmarks reveal 88% error in memory bandwidth between gem5’s current model and corresponding real AMD GPUs. To narrow this gap, we examined recent patents and gem5’s memory system bottlenecks, then made several improvements including: utilizing a redesigned HBM memory controller, enhancing TLB request coalescing, adding support for multiple page sizes, adding a page walk cache, and improving network bandwidth modeling. Collectively, these optimizations significantly improve gem5’s GPU memory bandwidth by 3.8x: from 153 GB/s to 583 GB/s. Moreover, our address translation enhancements can be ported to other ISAs where similar support is also needed, improving gem5’s MMU support. 

Traditional text classification typically categorizes texts into pre-defined coarse-grained classes, from which the produced models cannot handle the real-world scenario where finer categories emerge periodically for accurate services. In this work, we investigate the setting where fine-grained classification is done only using the annotation of coarse-grained categories and the coarse-to-fine mapping. We propose a lightweight contrastive clustering-based bootstrapping method to iteratively refine the labels of passages. During clustering, it pulls away negative passage-prototype pairs under the guidance of the mapping from both global and local perspectives. Experiments on NYT and 20News show that our method outperforms the state-of-the-art methods by a large margin. 

ABSTRACT STS scholarship has produced important insights about relationships between the roles of peer review and the social construction of knowledge. Yet, barriers related to access have been a continual challenge for such work. This article overcomes some past access challenges and explores peer review normativities operating in the new discipline of Engineering Education. In doing so, it contributes new insights about disciplinary development, interdisciplinarity, and peer review as a site of knowledge construction. In particular, it draws attention to an aspect of peer review not previously discussed – how peer review normativities are shaped by disciplinary origins. A content analysis of peer review documentation revealed that a hyperfocus on methods, which can be traced back to disciplinary origins, continues to be a guiding normativity. However, interviews with editors revealed that they do not acknowledge that normativity. Implications of those findings and their misalignment are discussed, as are contrasts with the history of other disciplines. 

ABSTRACT CONTEXT The peer review process plays a critical role in ensuring the quality of work published within a field and advancing the knowledge within the research community. However, for many members of the community, the process of peer review largely remains a black box to many scholars, especially those with less experience within the community. Therefore, there is a need to illuminate the peer review process for the research community. PURPOSE OR GOAL To more transparently reveal the contents of the black box around the peer review process, we interviewed editors (associate and deputy editors) for the Journal of Engineering Education (JEE) to provide editor perspectives on the overall peer review process. The goal of this paper is to clearly articulate the behind-the-scenes processes of peer review as well as the expectations and perceptions of the editors with respect to publishing within JEE. By bringing these processes to light, we hope that more members of the field will be aware of the overall process and the associated expectations for contributing to the field. APPROACH OR METHODOLOGY/METHODS To meet the goals of this study, we conducted semi-structured interviews with six editors of JEE who worked in the field of engineering education research (EER), as a part of a larger project exploring the boundaries of the field as expressed within the peer reviews process. The interviewer from the research team followed a protocol but also asked additional questions to elicit more details in some cases. The interviews were recorded, transcribed, and thematically coded using an open-coding process. ACTUAL OR ANTICIPATED OUTCOMES Based on the analysis of the editor interviews, we present three critical aspects of the peer review process: the types of editors, the process that editors typically conduct to identify reviewers, and the types of decisions through the process. Additionally, we highlight considerations and advice from the editors to help members of the EER community develop. CONCLUSIONS/RECOMMENDATIONS/SUMMARY The current study makes the editors’ perspectives and decision-making processes more explicit to readers. These decision-making processes are full of careful considerations and also challenges. By doing so, we hope to help the members of the EER community gain a better understanding of what is going on backstage of the peer review process.