BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Unlocking the Power of Data-Centric Acceleration for Modern Applic
 ations - Haiyu Mao\, ETH Zurich
DTSTART:20231110T150000Z
DTEND:20231110T160000Z
UID:TALK207565@talks.cam.ac.uk
CONTACT:Timothy Jones
DESCRIPTION:In today's digital landscape\, the exponential growth of data 
 has become the driving force behind cutting-edge applications\, such as ge
 nome analysis and machine learning applications\, revolutionizing our appr
 oach to healthcare and overall living quality. However\, this unprecedente
 d deluge of data poses a formidable challenge to traditional von Neumann c
 omputer architectures. The inefficiencies arising from the constant data m
 ovement between processors and memory consume a substantial portion of bot
 h execution time and energy when running modern applications on convention
 al von Neumann computers. To reduce this significant data movement\, data-
 centric architectures\, particularly processing-in-memory accelerators\, e
 merge as a promising solution by enabling the processing of data directly 
 where it resides. Nonetheless\, most existing data-centric architectures p
 rimarily focus on accelerating specific arithmetic operations\, inadverten
 tly leaving a substantial gap between the architectural enhancements and t
 he holistic needs of modern applications. Concurrently\, conventional soft
 ware optimizations often treat the architecture as a black box\, which inh
 erently limits the potential acceleration of applications.\n\nThis talk se
 eks to bridge the gaps between modern applications and data-centric archit
 ectures and revolutionize the landscape of data-centric acceleration for t
 wo vital categories of modern applications: genome analysis and machine le
 arning. Firstly\, this talk offers a comprehensive analysis of the pressin
 g challenges within state-of-the-art genome analysis pipelines and introdu
 ces an innovative end-to-end data-centric acceleration approach achieved t
 hrough seamless software-and-hardware co-design. Secondly\, this talk illu
 minates the path to closing the gap between data-centric accelerators and 
 the execution of real-world applications by presenting a compelling case s
 tudy centered on a crucial machine learning application based on generativ
 e adversarial networks (GANs). Furthermore\, this talk delves into the int
 ricate challenges of data-centric acceleration for modern applications and
  explores potential solutions to surmount these obstacles\, paving the way
  for a future where data-centric acceleration seamlessly integrates with t
 he ever-evolving landscape of advanced applications.
LOCATION:SS03\, Computer Laboratory\, William Gates Building
END:VEVENT
END:VCALENDAR