Browsing by Author "Wang, Yun-Ting"

Now showing 1 - 4 of 4

Aging-aware Chip Health Prediction Adopting an Innovative Monitoring Strategy
(2019-01-01)
Wang, Yun-Ting
;
Wu, Kai-Chiang
;
Chou, Chung-Han
;
Chang, Shih-Chieh
;
資訊工程學系
Concerns exist that the reliability of chips is worsening because of downscaling technology. Among various reliability challenges, device aging is a dominant concern because it degrades circuit performance over time. Traditionally, runtime monitoring approaches are proposed to estimate aging effects. However, such techniques tend to predict and monitor delay degradation status for circuit mitigation measures rather than the health condition of the chip. In this paper, we propose an aging-aware chip health prediction methodology that adapts to workload conditions and process, supply voltage, and temperature variations. Our prediction methodology adopts an innovative on-chip delay monitoring strategy by tracing representative aging-aware delay behavior. The delay behavior is then fed into a machine learning engine to predict the age of the tested chips. Experimental results indicate that our strategy can obtain 97.40% accuracy with 4.14% area overhead on average. To the authors' knowledge, this is the first method that accurately predicts current chip age and provides information regarding future chip health.
Computation and Communication Aware Task Graph Scheduling on Multi-GPU Systems
(2015)
Wang, Yun-Ting
;
Lee, Jia-Ying
;
Lai, Bo-Cheng Charles
;
電子工程學系及電子研究所
;
Department of Electronics Engineering and Institute of Electronics
GPUs have emerged as popular throughput computing platforms due to the massively parallel computing capability and low cost. To attain further performance enhancement beyond single GPU, there is a growing interest in exploiting systems with multiple GPUs. Attaining superior performance in a multi-GPU system involves three main design challenges, namely load balance, memory utilization, and data transfer. Imbalanced loading across a system could cause idling of GPUs while poor data reuse would trigger excessive memory accesses. The inefficient data transfer between a host and a device becomes a considerable performance overhead during high throughput computing. This paper aims at addressing the above design issues by proposing a Computation and Communication Aware task graph Scheduling (CCAS) for multi-GPU systems. The proposed scheduling approach (CCAS) adopts an effective heuristic algorithm that considers both data reuse and load balance in a multi-GPU system. The data transfer overhead is hidden by extensively overlapping computation and data communication. The experimental results of the proposed CCAS have demonstrated an average of 22.15% performance enhancement when compared with a previous work.
Hypoxia-responsive miRNAs target argonaute 1 to promote angiogenesis
(2013-03-01)
Chen, Zhen
;
Lai, Tsung-Ching
;
Jan, Yi-Hua
;
Lin, Feng-Mao
;
Wang, Wei-Chi
Despite a general repression of translation under hypoxia, cells selectively upregulate a set of hypoxia-inducible genes. Results from deep sequencing revealed that Let-7 and miR-103/107 are hypoxia-responsive microRNAs (HRMs) that are strongly induced in vascular endothelial cells. In silico bioinformatics and in vitro validation showed that these HRMs are induced by HIF1 alpha and target argonaute 1 (AGO1), which anchors the microRNA-induced silencing complex (miRISC). HRM targeting of AGO1 resulted in the translational desuppression of VEGF mRNA. Inhibition of HRM or overexpression of AGO1 without the 3' untranslated region decreased hypoxia-induced angiogenesis. Conversely, AGO1 knockdown increased angiogenesis under normoxia in vivo. In addition, data from tumor xenografts and human cancer specimens indicate that AGO1-mediated translational desuppression of VEGF may be associated with tumor angiogenesis and poor prognosis. These fmdings provide evidence for an angiogenic pathway involving HRMs that target AGO1 and suggest that this pathway may be a suitable target for anti- or proangiogenesis strategies.
MicroRNA-92a Mediates Endothelial Dysfunction in CKD
(2017-11-01)
Shang, Fenqing
;
Wang, Shen-Chih
;
Hsu, Chien-Yi
;
Miao, Yifei
;
Martin, Marcy
CKD is an independent risk factor for cardiovascular disease (CVD). The accumulation of uremic toxins in CKD induces oxidative stress and endothelial dysfunction. MicroRNA-92a (miR-92a) is induced by oxidative stress in endothelial cells (ECs) and involved in angiogenesis and atherosclerosis. We investigated a role for oxidative stress responsive miR-92a in CKD. Our study of patients at three clinical sites showed increased serum miR-92a level with decreased kidney function. In cultured ECs, human CKD serum or uremic toxins (such as indoxyl sulfate), compared with non-CKD serum, induced the levels of miR-92a and suppressed the expression of miR-92a targets, including key endothelial-protective molecules. The antioxidant N-acetylcysteine inhibited these vasculopathic properties. In rats, adenine-induced CKD associated with increased levels of miR-92a in aortas, serum, and CD144(+) endothelial microparticles. Furthermore, CD144(+) microparticles from human uremic serum contained more miR-92a than those from control serum. Additional analysis showed a positive correlation between serum levels of miR-92a and indoxyl sulfate in a cohort of patients with ESRD undergoing hemodialysis. Collectively, our findings suggest that the uremic toxins accumulated in CKD can upregulate miR-92a in ECs, which impairs EC function and predisposes patients to CVD.