P
Prerna Budhkar
Researcher at University of California, Riverside
Publications - 9
Citations - 99
Prerna Budhkar is an academic researcher from University of California, Riverside. The author has contributed to research in topics: Multithreading & Cache. The author has an hindex of 4, co-authored 9 publications receiving 86 citations. Previous affiliations of Prerna Budhkar include Eaton Corporation & Intel.
Papers
More filters
Journal ArticleDOI
High-Level Language Tools for Reconfigurable Computing
Skyler Windh,Xiaoyin Ma,Robert J. Halstead,Prerna Budhkar,Zabdiel Luna,Omar Hussaini,Walid Najjar +6 more
TL;DR: The history of using FPGAs as hardware accelerators is reviewed, the challenges facing the raising of the programming abstraction layers are summarized and five High-Level Language tools for the development of FPGA programs are surveyed.
Proceedings ArticleDOI
FPGA-accelerated group-by aggregation using synchronizing caches
Ildar Absalyamov,Prerna Budhkar,Skyler Windh,Robert J. Halstead,Walid Najjar,Vassilis J. Tsotras +5 more
TL;DR: This work relies on hardware multithreading to efficiently mask long memory access latency by implementing a custom operation datapath on FPGA, and proposes using CAMs (Content Addressable Memories) as a mechanism of synchronization and local pre-aggregation.
Proceedings ArticleDOI
CAMs as Synchronizing Caches for Multithreaded Irregular Applications on FPGAs
TL;DR: This paper describes the use of CAMs (Content Addressable Memories) as synchronizing caches for hardware multithreading and demonstrates and evaluates this mechanism using graph breadth-first search (BFS).
Proceedings ArticleDOI
Scalable Low-Latency Persistent Neural Machine Translation on CPU Server with Multiple FPGAs
Eriko Nurvitadhi,Mishali Naik,Andrew Boutros,Prerna Budhkar,Ali Jafari,Dongup Kwon,Sheffield David B,Abirami Prabhakaran,Karthik Gururaj,Pranavi Appana +9 more
TL;DR: The first multi-FPGA evaluation of a complex NMT with bi-directional LSTMs, attention, and beam search is presented, offering 110× better latency than the only prior NMT work on FPGAs, which uses a high-end FPGA and stores the model off-chip.
CAMs as Synchronizing Caches for Multithreaded Irregular Applications on FPGAs Invited Paper
TL;DR: In this article, the authors describe the use of CAMs (Content Addressable Memories) as synchronizing caches for hardware multithreading and demonstrate and evaluate this mechanism using graph breadth-first search (BFS).