Linley Data Center Conference 2016
Covers system design for data-center servers and networking
Held on February 9 - 10, 2016
Hyatt Regency Hotel, Santa Clara, CA

» Events  |  Event Info  |  Agenda Day One  |  Proceedings

Agenda for Day Two: February 10, 2016
View Day One

Click arrow below to view presentation abstract or click here -> to view all abstracts
9:00am-10:00amKeynote

How ARM Servers Can Take over the World
Jon Masters, Computer Architect, Red Hat

Industry standard ARM servers, built using flexible System-on-Chip technology, and featuring workload-optimized acceleration, promise to bring innovation and competitive disruption to traditional datacenters over the coming years. Jon Masters explains how an industry is coming together to build compatible, standardized hardware, which nevertheless contains many differentiating features. This session touches upon work being done to port Infrastructure and Operating System software to the ARMv8 Architecture, as well as solution stacks that encompass a diverse range from Telco (OPNFV) to Cloud (OpenStack), to Storage (Ceph), and beyond. Finally, consideration is given to emerging industry standardization efforts being applied to workload accelerators, offload engines, and reconfigurable logic (FPGAs), which will form an important component in ARM powered servers of the future.

There will be time for Q&A following the presentation.

10:00am-10:20amBREAK – Sponsored by NXP
10:20am-11:25amSession 4: Network Fabrics for Cluster Computing

This session, moderated by Bob Wheeler, Principal Analyst at The Linley Group, will feature low-latency fabrics that maximize performance in compute clusters based on a variety of processor architectures.

The Intel Omni-Path Architecture (OPA): The Next Generation High Performance Fabric
Barry Davis, General Manager, HPC Compute & Networking, Data Center Group, Intel

The Intel Omni-Path Fabric addresses the needs of High Performance Computing and workloads demanding high bandwidth and low latency. This presentation will discuss the OPA differentiation along two vectors: raw performance required for HPC workloads and price/performance. Further, it will illustrate how the OPA changes the economics for scale out clusters, allowing end users to win in the marketplace by gaining new insights into their businesses.

Edge + Data Center Analytics for Mobile Network with RapidIO
Devashish Paul, Director Strategic Marketing-Systems Solutions, IDT

New services require analytics with ultra-low latency close to end users. Network providers are looking at deploying analytics close to users to support applications that run in data centers today. The latency requirements of 5G require computing at the network edge, coordinated with applications in the data center. This session will discuss RapidIO-based heterogeneous computing with x86/ARM/GPU/FPGA being used for edge and data-center real-time analytics. The discussion will cover top of rack switches and converged switching + server appliances as well as use case examples of this distributed computing paradigm.

There will be Q&A and a panel discussion featuring above speakers.

11:25am-12:30pmSession 5: Custom Silicon for the Cloud

This session, moderated by Loring Wirbel, Senior Analyst at The Linley Group, will explore leading-edge semiconductor technologies and associated intellectual-property cores that enable high efficiency and density in data-center silicon.

High-performance ASIC Solutions for Data Center
Nobuaki Nonaka, Manager, High Performance SoC Business Unit, Socionext

Socionext develops and delivers innovative System-on-Chip products for customers worldwide. The company's leading-edge technologies enable today's dynamic data-center applications. This presentation will illustrate how Socionext's advanced technologies can help customers design differentiated products and reduce development time. The presentation will detail leading edge SerDeS and building blocks designed to expedite ASIC development in 20/16nm and finer technologies.

Accelerating HyperCloud Data Center Solutions with Breakthrough Integration Technologies
Sanjay Charagulla, Senior Director, Segment Marketing, GLOBALFOUNDRIES

HyperCloud data centers are creating demand for various compute-intensive workloads, storing the data over the cloud, and switching applications running centrally. This is driving the need for larger computing clusters for processing that will require large amounts of memory bandwidth and high-speed interconnectivity between servers, storage, and switches. This presentation will explore various integration technologies and solutions including hundreds of processors, 2.5D/3D stack memories, and optical and networking links (up to 56Gbps) that will enable better utilization and efficiency in the data center.

There will be Q&A and a panel discussion featuring above speakers.

12:30pm-1:30pmLUNCH
1:30pm-2:30pmSession 6: Cloud-Server Processors

This session, moderated by Jag Bolaria, Principal Analyst at The Linley Group, will discuss new processor options for data-center servers.

Open Innovation to Accelerate the Data Center
Randy Swanberg, Distinguished Engineer, IBM

Big data, internet of things, and the need to extract value from all data sources are driving new workloads. The open software ecosystem is revolutionizing the full stack of software, from hypervisors, operating systems, middleware, and tools to applications, helping to fuel advancements in data science, machine learning, deep learning, and cognitive computing. This session will discuss how hardware innovation from the OpenPOWER Foundation is playing a role in meeting current and future workload demands.

Optimizing the Cloud Datacenter with Scale-Out Processors
Rishi Chugh, Director, Product Marketing, Cavium

Today's cloud computing is driving new workloads, requiring power efficient servers. Designed for enterprise, cloud, and HPC applications, conventional server processors fall short in maximizing the performance and cost of ownership for these new workloads. This presentation will discuss how the innovations in Cavium's ThunderX processors are enabling a new class of throughput and power efficiency. It will also detail how ThunderX and VirtSoc meet the disruptive needs of cloud computing applications such as SDN and NFV.

There will be Q&A and a panel discussion featuring above speakers.

2:30pm-2:40pmBREAK – Sponsored by NXP
2:40pm-3:30pmSpecial Panel: Future Directions and Trends in Cloud Computing

In this session, Jag Bolaria will host a panel of experts to discuss the latest trends and evolution of cloud computing. Thought leaders and researchers from various companies working on hyper scale data centers, carrier networks, cloud software, and system architecture will share their insight and perspectives.

Panelist
Yuval Bachar, Principal Engineer, Global Infrastructure and Strategy, LinkedIn

Panelist.

Panelist
Christos Kolias, Sr. Research Scientist, Orange Silicon Valley

Panelist.

Panelist
Jon Masters, Computer Architect, Red Hat

Panelist.

Panelist
Peter Magnusson, SVP of Oracle Public Cloud Development, Oracle

Panelist.

 

Platinum Sponsor

Gold Sponsor

Ihphi

IDT

EZchip

Cavium Networks

Industry Sponsor

EEMBC