Y
Yichen Yang
Researcher at University of Michigan
Publications - 9
Citations - 52
Yichen Yang is an academic researcher from University of Michigan. The author has contributed to research in topics: Computer science & Engineering. The author has an hindex of 2, co-authored 3 publications receiving 9 citations.
Papers
More filters
Proceedings ArticleDOI
Prodigy: Improving the Memory Latency of Data-Indirect Irregular Workloads Using Hardware-Software Co-Design
Nishil Talati,Kyle May,Armand Behroozi,Yichen Yang,Kuba Kaszyk,Christos Vasiladiotis,Tarunesh Verma,Lu Li,Brandon Nguyen,Jiawen Sun,John Magnus Morton,Agreen Ahmadi,Todd Austin,Michael O'Boyle,Scott Mahlke,Trevor Mudge,Ronald G. Dreslinski +16 more
TL;DR: Prodigy as discussed by the authors is a low-cost hardware-software codesign solution for intelligent prefetching to improve the memory latency of several important irregular workloads, including graph analytics, sparse linear algebra, and fluid mechanics that exhibit two specific types of datadependent memory access patterns.
Proceedings ArticleDOI
NDMiner: accelerating graph pattern mining using near data processing
TL;DR: NDMiner is presented, a Near Data Processing (NDP) architecture that improves the performance of GPM workloads and proposes a new graph remapping scheme in memory and a hardware-based set operation reordering technique to best optimize bank, rank, and channel-level parallelism in DRAM.
Journal ArticleDOI
A data-driven CUF-based beam model based on the tree-search algorithm
Yan Hu,Xiaowei Bai,Yichen Yang,Jie Yang,Q. Huang,Xiao Jun Liu,Wei Huang,Gaetano Giunta,Heng Hu +8 more
TL;DR: In this paper , two strategies are employed in the framework of DDCM to improve computational efficiency: Carrera Unified Formula (CUF) is used to build the numerical model, which reduces the computational cost (amount of gauss points) and the number of iterations.
Proceedings ArticleDOI
Parallelism Analysis of Prominent Desktop Applications: An 18- Year Perspective
TL;DR: The analyses show that the harnessed parallelism has improved and emerging workloads show good utilization of hardware resources, and the effectiveness of software in utilizing the underlying hardware has improved, but still has scope for optimizations.
Proceedings ArticleDOI
NDMiner
Nishil Talati,Haojie Ye,Yichen Yang,Leul Belayneh,Kuan-Yu Chen,David Blaauw,Trevor Mudge,Ronald G. Dreslinski +7 more
TL;DR: Graph Pattern Mining (GPM) algorithms mine structural patterns in graphs as mentioned in this paper , but the performance of GPM workloads is bottlenecked by control flow and memory stalls, this is because of data-dependent branches used in set intersection and difference operations that dominate the execution time.