Abstract
The Internet of Things (IoT) is a rapidly approaching
technological change that envisions ubiquitous and
network-accessible digital instrumentation and actuation of
literally every "thing" we encounter in everyday life. Like the
World Wide Web (now simply called The Internet) before it, IoT
will likely represent another societal sea change as objects in
the physical world become network-enabled so that they can
communicate and interact with people and, autonomously, with each
other.
This technological vision also carries with it significant new challenges. With estimates of between 50 billion and 1 trillion network-connected IoT devices in the next 20 years, the energy efficiency of these devices and the network technologies that interconnect them is paramount to their utility. Moreover, the current Internet architecture, which is evolving to accommodate cloud computing, will require substantial additional innovation and augmentation before IoT will come to complete fruition.
In this talk, we will discuss some of the computer science research questions that have grown from early experiences in architecting and deploying working IoT systems and infrastructure. In particular, the talk will focus on potential new approach to software infrastructure that is designed to meet many of the current and future IoT challenges.
To save power, reduce network latency, and easy network congestion, devices export data and actuation services that are accessed by applications running in the cloud. "Flipping" the current Internet architecture in this way, with services at the extreme edge of the network and applications running at the core (i.e., in the cloud), requires a new technological approach that creates a Software Platform of Things -- SPOT -- spanning devices, computing elements located at the edge (e.g. edge clouds), and traditional cloud data centers.
We will outline our experiences in building and deploying IoT systems using this new "flipped" approach to cloud computing discuss the myriad of new research opportunities that arise as a result.
Bio
Dr. Rich Wolski is a Professor of Computer Science at the
University of California, Santa Barbara (UCSB) where he holds the
Duval Presidential Chair in Energy Efficiency. Having received
his M.S. and Ph.D. degrees from the University of California at
Davis (while a research scientist at Lawrence Livermore National
Laboratory) he has also held positions at the University of
California, San Diego, and the University of Tennessee, the the
San Diego Supercomputer Center and Lawrence Berkeley National
Laboratory. Rich has led several national scale research efforts
in the area of distributed systems and is the progenitor of the
Eucalyptus open source cloud project.
Abstract
Exscalate (EXaSCale smArt pLatform Against paThogEns) is a
drug-discovery platform that includes a "chemical library" of
several hundred billions of molecules and a processing capacity
in the order of millions of molecules per second. The platform
represents a powerful tool to accelerate the in-silico
computational phase of the development of new therapies, called
virtual screening process, which needs to exploit at the best the
underlying supercomputing resources. Exscalate platform is owned
by Dompé Farmaceutici and developed thanks to a collaboration
between Dompé, Cineca, and Politecnico di Milano. The Exscalate
platform has already been used in 2019 in the context of the
Antarex European project for the study of the Zika virus. This
talk will describe the Exscalate platform currently used in the
EXscalate4CoV Project funded by the European Commission’s Horizon
2020 programme for research and innovation to accelerate the
search for drugs against the coronavirus and to fight future
pandemics. In the race against the coronavirus, the platform can
now count on the two most powerful supercomputers in Europe: the
new HPC5 system installed in Italy at the energy company Eni and
the Marconi-100 system installed at the Cineca Italian
supercomputing center.
Exscalate4CoV European Project
Bio
Cristina Silvano is a Full Professor of Computer Engineering at
the Politecnico di Milano. She is currently the Chair of the
research area on Computer Science and Engineering and Vice-Chair
of the PhD programme in Information Technology for Computer
Science and Engineering at Politecnico di Milano. Her research
focuses on computer architectures and electronic design
automation focusing on design space exploration for
energy-efficient computer architectures and application
autotuning for manycores and HPC systems. She has been Project
Coordinator of the European research projects ANTAREX, 2PARMA and
MULTICUBE. She is currently responsible of the task force on High
Performance Computing for the Exscalate4CoV H2020 European
project on accelerating the virtual screening in the drug
discovery process to fight pandemics. She is also responsible for
the POLIMI research unit of the H2020 AI4DI European project on
artificial intelligence for digitizing industry in collaboration
with STMicroelectronics. She is an active member of the
scientific community and she regularly serves in several
international program committees. She is Associate Editor of the
IEEE Trans. on Computers and the ACM Trans. on Architecture and
Code Optimization. She served as Independent Expert Reviewer for
the European Commission and for several science foundations. She
is an IEEE Fellow and a member of the HiPEAC network.
Abstract
Sustainability has become a key driver of business decisions for
many corporations. All major information technology companies
have committed to reducing their environmental impact and moving
towards a more sustainable future. Data centers themselves are an
integral part of the modern IT infrastructure, and their rapid
growth now and into the future make them a significant
contributor to carbon, water and waste issues. In addition, the
demands that the data centers place on the global ecosystem
impact the entire planet.
We will examine the sustainability benefits and challenges of modern data centers from the building infrastructure to the architecture of the servers, and from the carbon emissions resulting from the manufacturing and use of the systems to their eventual waste disposal. Although the talk focuses on data centers, many of the topics and challenges are applicable to other computer products and designs. This talk will provide inspiration for the academic community to incorporate sustainability as a primary objective (first class citizen?) in how we design and use current and future systems.
Bio
Srilatha (Bobbie) Manne has worked in the computer industry for
over two decades in both industrial labs and product teams at
Compaq, Intel, AMD and Cavium. She is currently a Principal
Hardware Engineer in the Azure Hardware Systems and
Infrastructure group at Microsoft. Her work has focused on power
and performance analysis from processor microarchitecture to data
centers. Srilatha has continued to publish while in industry, and
has over 20 patents granted. She served as the General Chair for
ISCA 2019 and has served on numerous program committees
throughout the years. Her latest passion is collaborating with,
and learning from, an interdisciplinary group of engineers and
technical experts on sustainability topics centered around data
centers. She currently lives in Seattle with her husband and two
children.
Best Paper Award
Winner: Bit-Exact ECC Recovery (BEER): Determining DRAM On-Die ECC Functions by Exploiting DRAM Data Retention Characteristics
Runner-Up: Virtualized Logical Qubits: A 2.5D Architecture for Error-Corrected Quantum Computing
Student Research Competition Winners
Undergraduate: Guyue Huang
Graduate: Peng Gu
MICRO Hall of Fame
Nael Abu-Ghazaleh, Jorge Albericio, Jason Mars, Lingjia Tang, Thomas F. Wenisch
B. Ramakrishna Rau Award
André Seznec