Sensors to Cloud Architectures Workshop

Event Dates

Feb 08, 2015 - Feb 08, 2015

Location

San Francisco Bay Area, California, USA

Submission Deadline

Dec 15, 2014

*************************************************************************

Sensors to Cloud Architectures Workshop (SCAW-2015)

February 8th 2015, Bay Area, California, USA

Held in conjunction with HPCA-21

http://darksilicon.org/hpca/SCAW-2015/

*************************************************************************

Organizing Chairs:

——————

Ramesh Illikkal Intel ramesh.g.illikkal@intel.com

Ravi Iyer Intel ravishankar.iyer@intel.com

Pei Zhang Carnegie Mellon University peizhang@cmu.edu

Christina Delimitrou Stanford University cdel@stanford.edu

Overview

——–

The computer industry is witnessing a major inflection point – Internet

of Things – that has implications from end (sensor devices) to end

(cloud architectures). Many technologies contribute to this inflection

point: Computing platforms are getting smaller (e.g., handheld devices,

wearables), richer (e.g., image and language understanding) and broader

(i.e., reaching the masses via Internet of Things). Sensors operating

in constrained environments connected through intelligent gateways and

cloud backend create a very complex environment for the operators, system

integrators, and developers of this emerging technology. Discovering and

managing sensor devices; collecting, cleaning and storing discoverable data;

normalizing, aggregating and analyzing the data for insights and actions;

managing the security and privacy of the data, enforcing the access

privileges and trusted execution environments – all these are required to

make this revolution happen.

The research challenges in IoT platforms are multi-fold: (a) providing

rich functionality and wider power/performance range for sensor devices,

(b) covering a broad range of applications that can be migrated from

cloud to gateways and sensor devices, (c) enabling a scalable and modular

cloud architecture that provides the required real-time and uptime

capabilities, and (d) providing a rich software programming environment

that facilitates developing applications on end-to-end platforms consisting

of elements ranging from sensors to gateways to cloud. The goal of this

workshop is to bring together academic researchers and industry practitioners

to discuss future IoT sensor-to-cloud architectures including sensors,

gateways and cloud architectures.

Below is the proposed list of topics for the workshop. Topics include,

but are not restricted to, the following:

• Sensors, Actuators, Gateway & Controllers Architectures

– Architectures for wearable and IoT devices

– Heterogeneity in cores, frequency, cache, memory

– Power, performance, energy optimizations

– SoCs, CPU/GPU, CPU/GPGPU architectures

– Ultra-Low power core micro-architectures

– Fabrics/Network-on-chip, cache/memory hierarchies

– HW support for heterogeneity, programmability, modularity

– Simulation/Emulation methodologies

– Protocols and abstraction layers (MQTT, CoAP, REST, …)

• Cloud Architecture

– Data Center architectures for IoT; customization/specialization

– Edge/Fog computing – Dynamic cloud-gateway-device offloads

– Design patterns and application programming frameworks

– Service creation and orchestration

– IaaS, PaaS, SaaS and DaaS

– Deployment models – Cloud, Private, Hybrid

• Emerging Workloads and Use Cases

– New wearable and IoT use cases and workloads

– Speech/Image recognition and understanding,

cognitive computing

– Personal assistants, predictive/prescriptive analytics

– Machine learning algorithms & applications,

graph processing, deep neural networks

– Descriptive, predictive and prescriptive analytics

– Workload analysis for power/performance/energy

optimization and acceleration

– Batch, streaming and distributed analytics

– Workload/Algorithm partitioning between

heterogeneous cores and accelerators

– Performance monitoring and simulation

• Novel Accelerator Designs

– Specialized accelerator architectures and designs

– Machine learning, neural network and graph processing accelerators

– Domain-Specific programmable/configurable accelerators

– Accelerator interfaces for programmability

– Development environments for accelerator design

Submission Guidelines: Interested authors are encouraged to submit

extended abstracts (1 – 2 pages) or short papers (6 pages) by email to

the organizing chairs. The deadline for submission is December 15th,

2014. Final (short) papers will be due on Jan 23rd, 2015 and will be

printed in a workshop proceedings made available to the workshop

attendees.

Important Dates:

—————-

Abstract/Paper submission: December 15 2014 23:59 PST

Author Notification: January 02 2015

Final Paper Submission: January 23 2015

Workshop: February 08 2015