News

Congratulations to William for his Paper “Lifetime Reliability Characterization and Management of Many-Core Processors” Selected as Best in Session at SRC TECHCON 2014

Posted by on Nov 2, 2014 in News

William Song took top honors in the System Design: Architectures and Thermal/Reliability Management Session at SRC TECHCON 2014. His paper, entitled “Lifetime Reliability Characterization and Management of Many-Core Processors,” represented the work of the Computer Architecture and Systems Laboratory, led by his Ph.D. advisor and ECE Professor Sudhakar Yalamanchili. Yalamanchili and ECE Associate Professor Saibal Mukhopadhyay were Song’s coauthors on the paper.

Song’s paper presents a characterization method for understanding how parallel applications create reliability variation in a multicore processor and how such variability is projected onto processor-level lifetime reliability. The key observation in Song’s paper leads to proposals for dynamic reliability management techniques that minimize the variance of reliability distribution on a multicore die and balance long term reliability-performance tradeoffs.

Congratulations to all the authors!

Sudhakar Yalamanchili gave a keynote address “The Era of Heterogeneity: Oppporunities and Challenges” at the Annual China National Computer Congress (CNCC) in Zhengzhou, China in October.

Posted by on Nov 2, 2014 in News

CNCC (the China National Computer Congress) is the largest and highest academic and technical conference in the area of computer science and technology in China. It was founded in 2003 and is annually held by CCF. This year, CNCC2014 was be held in Zhengzhou, Henan province on 23th-25th, October. The theme of CNCC2014 was ‘Information Security, Data being all important’; the topic is about Information Security. ACM and IEEE Computer Society have close participation in CNCC2014. Professor Yalamanchili was invited as a keynote speaker as a representative of the IEEE Computer Society.

Jin Wang presented her paper (Best Paper Nominee) “Characterization and Analysis of Dynamic Parallelism in Unstructured GPU Application” at IISWC 2014.

Posted by on Nov 2, 2014 in News

Jin Wang presented the paper “Characterization and Analysis of Dynamic Parallelism in Unstructured GPU Application” at IISWC-2014. This paper analyzes and characterizes the dynamically formed pockets of parallelism that exists in unstructured GPU applications. Specifically the analysis targets a comprehensive understanding of the new CUDA Dynamic Parallelism feature. The paper was nominated for the best paper.

Minhaj Hassan presented his paper “Bubble Sharing: Area and Energy Efficient Adaptive Routers using Centralized Buffers” at NOCS 2014 in September

Posted by on Nov 2, 2014 in News

Minhaj Hassan presented the paper “Bubble Sharing: Area and Energy Efficient Adaptive Routers using Centralized Buffers” at NOCS 2014 in September. The paper proposes a flow control mechanism for on-chip networks that provides deadlock freedom guarantees with the use of small, shared central buffers, eliminating the need of edge buffers in the routers. The result is an improved buffer space utilization with lower power and higher throughput. The technique effectively reconciles the trade-off between high radix and buffer space, encouraging the use of low hop count, high-radix topologies, with both deterministic and adaptive routing. See paper.

Congratulations to Jin Wang for having her paper “Characterization and Analysis of Dynamic Parallelism in Unstructured GPU Applications” nominated for best paper at the 2014 IEEE International Symposium on Workload Characterization

Posted by on Sep 4, 2014 in News

Congratulations to Jin Wang for having her paper “Characterization and Analysis of Dynamic Parallelism in Unstructured GPU Applications” nominated for best paper at the 2014 IEEE International Symposium on Workload Characterization. The Symposium will be held in October 2014 in Raleigh NC. The paper characterizes and evaluates the CUDA Dynamic Parallelism implementations for irregular applications with dynamically formed structured parallelism. See paper.

Congratulations to Jin Wang on being selected as a 2014-2015 NVIDIA Graduate Fellow

Posted by on Aug 22, 2014 in News

Congratulations to Jin Wang on being selected as a 2014-2015 NVIDIA Graduate Fellow. This highly select group of 5 Fellows were selected from more than a hundred applicants in 21 countries. Their projects involve a variety of technical challenges, including computer architecture, programming models, character animation, computer graphics, and computational methods for simulating chemical events. – See more here.

Congratulations to Andrew, Ifrah and Se Hoon for successfully submitting their Master’s Thesis

Posted by on Apr 10, 2014 in News

Andrew Vanderhayden, Ifrah Saeed and Se Hoon Shon have successfully submitted their Master’s Thesis. Congratulations!

Congratulations to Haicheng for his successful PhD Thesis Proposal!

Posted by on Apr 4, 2014 in News

Congratulations to Haicheng Wu for his successful PhD Thesis proposal on April 4, 2014!

Haicheng Wu presented Red Fox Execution Environment at GTC 2014

Posted by on Mar 26, 2014 in News

Haicheng Wu presented the Red Fox project at the GTC 2014 on March 26, 2016 in San Jose, CA. See the details.

William Song presented poster and short paper “Energy Introspector: A Parallel, Composable Framework for Integrated Power-Reliability-Thermal Modeling for Multicore Architectures” at ISPASS 2014

Posted by on Mar 24, 2014 in News

William Song presented poster and short paper “Energy Introspector: A Parallel, Composable Framework for Integrated Power-Reliability-Thermal Modeling for Multicore Architectures” at 2014 IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS-2014) on March 24, 2014 at Monterey, CA.