Session Descriptions
Track: Saving Energy in Servers and Datacenters
Track Description:
Servers and data centers now consume a measurable part of the world’s energy, and their consumption is growing rapidly. The US Environmental Protection Agency believes that consumption could reach 100 billion kWh by 2011, costing over $7 billion. Huge server farms, such as those used by Google, Facebook, Microsoft, Salesforce.com, Amazon.com, and Yahoo, are being located specifically in areas with low power costs. This Workshop will deal with ways to monitor energy usage and improve server efficiency. It will include presentations on reducing power consumption for processors, memory, networking components, and storage. It will also have a section on the efficient design of power systems. The special issues of data centers will be discussed, including the utilization of low-power modes, power and performance tradeoffs, power monitoring, and startup issues.
Intended Audience:
Hardware designers, engineering managers, server designers, systems engineers, systems analysts and integrators, marketing and product engineers and managers, technology managers, test engineers, telco and enterprise end users, consultants, design and communications specialists, and design service providers.
Track: Increasing Server Performance
Track Description:
Servers are a $40 billion market today. Hence, data centers are trying to get the most out of the huge expenditures being made on them. This Workshop will describe a variety of methods for measuring and increasing performance. It will pay particular attention to virtualization, which data centers are employing heavily to increase server utilization. Analyzing the performance of virtual servers requires special monitoring tools and methods. It will also discuss ways to increase performance at the component, system, and peripheral level, as well as ways to accelerate servers in general and for running particularly common and demanding applications such as Oracle and SAP.
Intended Audience:
Server designers; hardware designers and managers; CIOs and network, computer, telecom, and data center managers; facilities planners; performance analysts; system designers and engineers; CTOs and engineering managers; product planners, managers, and marketing engineers.
Track: Designing Cloud Servers
Track Description:
Cloud services are an exploding market with expectations (from Gartner) of $150 billion in revenue by 2013. Servers will play a huge role in cloud computing, since clouds themselves consist largely of servers capable of handling a wide variety of applications. Furthermore, users will obviously access clouds over server links. The issue then is what kind of servers are best-suited for building and accessing clouds. Which architectures are most suitable, what special support and capabilities are necessary, and what are the best cost-performance tradeoffs for individual cloud servers? What are the best designs for servers that must provide low-cost, high-capability, and highly flexible computing services over network links? This Workshop will discuss how to design servers for cloud computing. It will pay particular emphasis to virtualization, high-speed links, caching, latency, and performance measurement and optimization.
Intended Audience:
Server and cloud designers; hardware designers and managers; CIOs and network, computer, hosting center, cloud, and data center managers; facilities planners; cloud services specialists; system designers and engineers; CTOs and engineering managers; product planners, product and marketing managers, and marketing engineers.
| Tuesday, November 29th | ||||
|
8:30-9:45am
|
Chairperson: Steve Ohr, Gartner
Paper Presenters:
Session Description:
Using the Power Utilization Effective (PUE) index developed by The Green Grid, data center managers have been trying to reduce energy usage in data centers. One approach is to develop more efficient cooling techniques. Another is to substitute more energy-efficient semiconductor devices. This session offers perspectives from equipment and semiconductor makers on the best ways of promoting energy efficiency.
Intended Audience:
Hardware designers, engineering managers, server designers, systems engineers, systems analysts and integrators, marketing and product engineers and managers, technology managers, test engineers, telco and enterprise end users, facilities managers, consultants, design and communications specialists, and design service providers.
About the Chairperson:
Steve Ohr is Gartner’s research director for analog and power management. His current research focuses on data center power and cooling techniques, especially the means of measuring and controlling electrical power consumption.
Before joining Gartner, Ohr served as an editor and website manager for EE Times. His editorial career includes assignments with Computer Design magazine, Asian Electronic Engineer and Electronic Design. He has received a Jesse Neil Award for his reporting on microprocessor development. He holds an MS in sociology from Rutgers and a BS in industrial engineering from New Jersey Institute of Technology. |
|||
|
8:30-9:45am
|
Chairperson: Brian Berg, Berg Software Design
Paper Presenters:
Session Description:
Processor technology has advanced rapidly in recent years with multicores and smaller process dimensions. However, other parts of the server have not kept up. Advances in memory and storage are essential to avoid having the latest processors spend their time waiting for these other facilities to respond. New DRAM technologies, flash memory, and solid state drives (SSDs) are among the latest approaches to this problem. SSDs can use high-speed buses such as PCI Express that are widely used and already present in most architectures.
Intended Audience:
Hardware designers, engineering managers, server designers, systems engineers, systems analysts and integrators, marketing and product engineers and managers, technology managers, test engineers, telco and enterprise end users, server managers, consultants, design and communications specialists, and design service providers.
About the Chairperson:
Brian Berg, President of Berg Software Design, has been a consultant for 30 years. He has extensive experience with storage devices and interfaces, including flash memory, disk, DVD and CD, as well as USB, Fibre Channel, SATA, SAS, SCSI and storage area networks. He has been a project leader, software developer, industry analyst, technical marketer and author, as well as conference chair, session chair, and speaker. He has also designed and implemented storage subsystems, been a technical marketer, seminar leader, and tutor, and reviewed patent and trade secret issues as an expert witness. He received his B.S. in Mathematics and did CS/EE graduate work at Stanford. |
|||
|
8:30-9:45am
|
Chairperson: Nabil Damouny, Netronome
Paper Presenters:
Session Description:
A cloud is basically just a collection of IT assets that can serve a wide variety of purposes as a general utility. Servers clearly form an important part of those assets, since they perform both networking and computation tasks. What kinds of servers are best suited to provide the best cloud functionality? How can we achieve the optimal combination of cost, performance, flexibility, and availability required to fulfill the promise of cloud computing?
Intended Audience:
Hardware designers, engineering managers, server designers, systems engineers, systems analysts and integrators, marketing and product engineers and managers, technology managers, test engineers, telco and enterprise end users, consultants, design and communications specialists, and design service providers.
About the Chairperson:
Nabil G. Damouny is Senior Director of Strategic Marketing at Netronome. He has over 25 years of Silicon Valley marketing and engineering experience in communications and networking for both semiconductors and systems. Nabil was a founder and the Vice President of Marketing and Business Development at Basis Communications, where he positioned the company as a leader in network processors, leading to its acquisition by Intel.
During his 5 years at Intel, Nabil was most recently a member of the CTO Office in the Mobility Group, responsible for driving VoIP strategy. For the prior 3 years, he was responsible for strategic planning at Intel’s Communications Group focusing on security, management, and wireless communications. He has managed his own consulting business for the past 2 years when he can spare time from his primary position as “Retired Gentleman”. Earlier he held engineering and marketing positions at Fairchild, Philips Semiconductor, Nortel, and NEC America. He earned a BSEE from IIT in Chicago, and a MSECE from UC Santa Barbara. He holds 3 patents in computer architecture and remote networking. |
|||
|
10:00-11:20am
|
Chairperson: Mark Szalkus, GE Digital Energy
Paper Presenters:
Session Description:
One way to reduce data center power consumption is by upgrading the power and cooling systems. The new ASHRAE standards allow for wider variations in data center conditions, thus reducing the need for power and cooling apparatus within servers. One can also manage power from a server application point-of-view so systems are active only when the application needs them. On-chip liquid cooling can also reduce the needs for power-consuming fans and other equipment.
Data centers help drive today’s global economy, but they are increasingly more expensive to power and cool. Currently, they consume a little over 1% of the power in the United States annually per the Department of Energy. This figure will rise as global compute needs grow, density increases, and power and cooling demands increase. Power issues mean thaqt data centers will reach their designed full capacity limitations earlier than planned. We have time to develop and implement corrective measures. There are many approaches that can improve energy Intended Audience:
Hardware designers, engineering managers, server designers, systems engineers, systems analysts and integrators, marketing and product engineers and managers, technology managers, test engineers, telco and enterprise end users, consultants, design and communications specialists, and design service providers.
About the Chairperson:
Mark Szalkus is a Director of Technical Application Engineering at GE Energy’s Digital Energy business, responsible for the application engineering of Digital Energy products in North America. He has over 20 years experience in electrical distribution and power quality businesses, with leadership roles in field service, application engineering, service engineering and marketing.
Mark began his GE career with GE Zenith Controls in 1996. He has held roles of increasing responsibility in service and engineering including Field Service Coordinator, Senior Application Engineer, and Application Engineering Manager. Most recently, he was Marketing Manager for the Power Equipment P&L’s Power Quality business. Before joining GE, he was a Field Service Engineer and NETA certified test engineer in the Utility industry and Director of Engineering for a leading industrial equipment manufacturer. Mr. Szalkus has a BS degree in Electrical Engineering Technology. He is a member of The Green Grid, EGSA, NFPA, and Uptime Institute. He is a published author of several technical articles and papers in the power quality industry. | |||
|
10:00-11:20am |
Chairperson: Jim Porter, DiskTrend
Paper Presenters:
Session Description:
With increased computing demands and more data to handle, data center managers face a constant need for higher performance at minimal cost. One popular current approach is to use solid state drives (SSDs). These devices can supplement traditional hard disk drives, acting as another storage tier, a high-speed non-volatile memory, or a cache. Such drives can fit virtually anywhere in a system, since they have no mechanical parts. They are much faster than hard disk drives, but significantly more expensive on a per-byte basis.
Intended Audience:
Hardware designers, engineering managers, server designers, storage managers and specialists, systems engineers, systems analysts and integrators, marketing and product engineers and managers, technology managers, test engineers, telco and enterprise end users, consultants, design and communications specialists, and design service providers.
About the Chairperson:
Jim Porter first became involved with data storage products in 1968, when he joined the industry pioneer Memorex, serving in a variety of marketing management positions through 1971. After experience with early hand-held calculators at Rockwell International, with the first consumer video tape recorder at Cartridge Television, and with the original computer controlled video editing systems at CMX Systems, a CBS-Memorex joint venture, he began his own management consulting business in 1974. In 1977 he founded DISK/TREND, publisher of market studies of the worldwide disk drive and data storage industries, and has frequently acted as a consultant to the management of data storage manufacturers. He is a member of the Advisory Board of the Computer History Museum, and for several years chaired the museum’s Storage Special Interest Group. He is also a founder of IDEMA, the disk drive industry's trade association, and an active participant since its early days. |
|||
|
10:00-11:20am |
Chairperson: Frank Chang, Vitesse Semiconductor
Paper Presenters:
Session Description:
Clouds require high-speed connections both internally and externally to customers. Low latency is essential to avoid annoying delays at the user end. New versions of Ethernet can provide these characteristics, as can other interfaces. However, extremely high-speed connections always require extra components to preserve signal integrity over practical cable lengths.
Intended Audience:
Hardware designers, engineering managers, server designers, systems engineers, systems analysts and integrators, marketing and product engineers and managers, technology managers, test engineers, telco and enterprise end users, consultants, design and communications specialists, and design service providers.
About the Chairperson:
Frank Chang is a principal engineer for Vitesse Semiconductor. He specializes in optical system engineering, IC product specifications, and application issues for telecoSERVER DESIGN SUMMIT KEYNOTE DESCRIPTION - SAPm, datacom, and PON access markets. He is also knowledgeable in electronic dispersion compensation (EDC), PMD, PHY, and FEC chipsets and their application in optical networking. He has authored or co-authored over 55 peer-reviewed journal and conference articles. Frank is active in standard-setting bodies including OIF/ITU, IEEE 802.3, and FSAN for the definition of various optical specifications. He is currently serving in the OFC/NFOEC Program Committee under Technical Subcommittee B: Network Technologies, IEEE LEOS Technical Committee on “Mitigating Channel Degrading Effects,” the IEEE P802.3ba (formerly HSSG) 40G, and the 100G Ethernet Task Force organization. |
|||
|
11:20am-Noon |
Speaker: Jason Waxman, General Manager - High Density Computing - Data Center Group, Intel
Introducer: Swapna Yasarapu, STEC
Abstract:
Market researchers predict that the number of servers used in cloud computing will increase 25% annually during the next few years. As this number rises, it will affect server design, particularly forcing more focus on efficiency, performance, and systems management. Vendors will also have to emphasize the best overall combination of servers, storage, and networking. Cloud computing and mega-datacenters (ones with over 100,000 servers) will thus force key new trends in server design and manufacturing.
About the Speaker:
Jason Waxman is General Manager in Intel’s Data Center and Connected Systems Group, leading the Cloud Infrastructure Group (CIG). He is responsible for Intel’s products and technologies for large datacenters, including silicon components, optimized system design, software, management, and data center facility optimization. Jason’s business responsibility focuses on working with systems and software vendors to address the needs of the high-growth cloud service provider industry, including internet datacenters, telcos, infrastructure/ software-as-a-service, and hosting. Jason holds executive positions in several industry design efforts including the boards of the Open Compute Foundation and the Server System Infrastructure Forum and technical advisor to the Open Data Center Alliance. Jason has spent the last 14 years of his Intel career in enterprise computing focused on server products and technologies. He has helped introduce over 12 new platforms. Before joining Intel, Jason worked in strategic planning for an industrial components company and as a management consultant. He holds Bachelor and Masters Degrees in engineering and a Masters of Business Administration from Cornell University. He can be followed on twitter @jpwaxman.
About Intel
Intel (NASDAQ: INTC), the world leader in silicon innovation, develops technologies, products and initiatives to continually advance how people work and live. Additional information about Intel is available at www.intel.com/pressroom and blogs.intel.com.
|
|||
|
1:30-2:00pm |
Speaker: Robert Hormuth, Dell
Introducer: Sergis Mushell, Gartner
Abstract:
Today’s enterprises face many new information challenges. There is much more data and computing power available, and many new ways to use them. Mobile devices and social media are everywhere, and the cloud offers on-demand access to extra resources at reasonable costs. Meanwhile, budgets and staff levels are flat at best. New server designs must provide great flexibility as well as the ability to apply computing power everywhere in the enterprise. Server companies must offer tools that allow enterprises to handle the latest standards efficiently and automate common tasks easily and quickly. At the same time, they must keep innovating in high-frequency board design, power and cooling techniques, and cost controls. They will be key partners in helping enterprises survive and prosper in a highly uncertain, fast-changing, and difficult economic world.
About the Speaker:
Robert Hormuth is Director of Server Architecture in the Office of CTO at Dell. Robert has 24 years in the computer industry. Robert joined Dell in 2007 after 8 years wit
h Intel and 12 years at National Instruments. Past design/architecture activities include: IO peripheral designs, x86 (386, 486, Pentium, Pentium Pro, Pentium II) system design, BIOS, firmware, software, and FPGA/ASIC design. Additionally Robert served as a Field Application Engineer supporting multiple fortune 100 companies.
Robert has participated in creation of multiple industry standards (VME, VXI, PCI, PXI). Robert has a B.S. in Electrical and Computer Engineering from the University of Texas at Austin. Robert holds 3 patents with 4 pending. About Dell:
Dell (NASDAQ: DELL) listens to its customers and uses that insight to make technology simpler and create innovative solutions that deliver reliable, long-term value. The company delivers innovative technology and services which customers trust and value. As a leading technology company, we offer a broad range of product categories, including mobility products, desktop PCs, software and peripherals, servers and networking, and storage. Our services include a broad range of configurable IT and business related services, including infrastructure technology, consulting and applications, and business process services. We are committed to managing and operating our business in a responsible and sustainable manner around the globe. This includes our commitment to environmental responsibility in all areas of our business. Learn more at www.dell.com. |
|||
|
2:00-2:30pm |
Speaker: Sarwar Raza, Hewlett-Packard
Chairperson: Jean Bozman, IDC
Abstract:
Advances in server and storage virtualization have enabled a new class of cloud based applications and services. These new applications and services have transformed how enterprises view and build their IT infrastructures. The network glue that binds the end users of these cloud services to applications is also evolving to meet a new set of scalability, security and management challenges. To meet new requirements – ask yourself, how are networks, and the management principles governing their operations evolving to meet the needs of server administrators, network managers and end users in the cloud today? Simple, it must be done at the server-network edge.
About the Speaker:
Sarwar Raza is a distinguished architect in the Office of the Chief Technology Officer for Networking at HP. He provides technical and marketing leadership spanning the areas of cloud networking and orchestration, network and systems management, and unified communications and collaboration to engineering, marketing and sales teams across HP’s product and service businesses.
Raza joined HP in 2010 via the acquisition of 3Com Corporation, where for 10 years he held a variety of senior roles in product management and engineering. Most recently, he led the Product Management team for 3Com’s Unified Communications and IP Surveillance business lines. Prior to this, he served as a software architect on the Voice Engineering team for the VCX product line. His network management expertise includes work on carrier-class management systems for RAS, VOIP and 3G systems that have been deployed by some of the world’s largest service providers. Raza also was lead architect for 3Com’s flagship NMS product, Enterprise Management Suite. Raza has a bachelor’s degree in economics and computer science (Phi Beta Kappa) from Clark University, a master’s degree in computer science with a concentration in computer and communications networks from Worcester Polytechnic Institute, and an executive certificate in management and strategy from the Sloan School of Management at MIT. About HP
HP creates new possibilities for technology to have a meaningful impact on people, businesses, governments and society. The world’s largest technology company, HP brings together a portfolio that spans printing, personal computing, software, services and IT infrastructure at the convergence of the cloud and connectivity, creating seamless, secure, context-aware experiences for a connected world. More information about HP (NYSE: HPQ) is available at http://www.hp.com.
|
|||
|
2:30-3:45pm
|
Chairperson: Jim Cantore, JLC Associates
Paper Presenters:
Session Description:
Data center strategies can help reduce server energy consumption. For example, liquid cooling can replace air cooling to reduce energy lost to fans and other air conditioning equipment. Higher temperatures and humidity levelsw can be tolerated as well under specific conditions. Legislation and regulations can affect strategies, as centers may be required to meet external mandates and promote sustainability.
Intended Audience:
Hardware designers, engineering managers, server designers, systems engineers, systems analysts and integrators, marketing and product engineers and managers, technology managers, test engineers, telco and enterprise end users, data center and server managers, consultants, design and communications specialists, and design service providers.
About the Chairperson:
Jim Cantore is the President and Chief Analyst at JLC Associates, a high technology consulting firm since September 2003. Mr. Cantore consults in technology, business development, strategic marketing and market intelligence to financial, high technology and semiconductor companies. His specialties include processors, memory, graphics devices, and DSPs. He has previous experience with iSuppli, IDC, Micron, Mosel Vitelic, Natiional Semiconductor, and IDC. |
|||
|
2:30-3:45pm
|
Chairperson: Stan McClellan, ZNYX Networks
Paper Presenters:
Session Description:
Today’s servers can be extended in many ways. PCI Express is an easy interface to use to expand servers to meet higher performance needs or provide additional flexibility or capabilities. Servers can also process packets and provide switching functions either in software or with the aid of additional hardware such as DSPs, FPGAs, and network processors.
Intended Audience:
Hardware designers, engineering managers, server designers, systems engineers, systems analysts and integrators, marketing and product engineers and managers, technology managers, test engineers, telco and enterprise end users, consultants, design and communications specialists, and design service providers.
About the Chairperson:
Stan McClellan, PhD is the Associate Director of the Ingram School of Engineering at Texas State University and Chief Architect of Systems & Solutions at ZNYX Networks. Dr. McClellan has held positions in enterprise, military/aerospace, and academic industries, and served as chief technologist, chief architect, or lead engineer for several distributed real-time systems. He has also served as a technology and business consultant for such organizations as, Cisco, 3Com, Alcatel, N.E.T. Motorola, Nortel, BellSouth, MCI/Worldcom, and others. Dr. McClellan is the author of many peer-reviewed technical publications and US/international patents, and has made invited contributions to well-known references including Advances in Computers, The IEEE/CRC Electrical Engineering Handbook, and The Encyclopedias of Electrical & Electronics Engineering. |
|||
|
2:30-3:45pm
|
Chairperson: Tom Coughlin, Coughlin Associates
Paper Presenters:
Session Description:
Virtualization is a key element enabling the creation of cost effective and flexible clouds to handle a variety of applications. All elements of the data center used in a cloud must be virtualized, including servers, I/O, storage, and networking. The cloud also requires an operating system that can take advantage of its features and provide services to applications. This combination provides the rapid reconfiguration and efficient assignment of resources required for multiuser public and private clouds.
Intended Audience:
hardware designers, engineering managers, server designers, systems engineers, systems analysts and integrators, marketing and product engineers and managers, technology managers, test engineers, telco and enterprise end users, data center managers, cloud specialists, server managers, consultants, design and communications specialists, and design service providers.
About the Chairperson:
Tom Coughlin is President of Coughlin Associates, a data storage consulting firm specializing in data storage components, systems, and software. He has over 20 years of industrial experience working at such companies as 3M, Polaroid, Seagate, Maxtor, Ampex, and SyQuest. He has over 60 articles, reports, and technical presentations to his credit and 6 patents. Tom is the author of the book “Digital Storage in Consumer Electronics: The Essential Guide”, published by a division of Elsevier. He is also the organizer of the annual Storage Visions Conference and the Creative Storage Conference. |
|||
|
4:00--5:15pm
|
Chairperson: Sergis Mushell, Gartner
Panelists:
Session Description
What do the processor makers have in store for servers over the next few years? We’ll hear from Intel, AMD, and Calxeda (a maker of ARM-based servers) as to what we can expect. Their roadmaps will give us guidelines as to what must be accomplished in the rest of the server systems to maximize performance and minimize energy consumption. Processor advances will also govern general design schedules and performance estimates and analyses.
Intended Audience:
Hardware designers, engineering managers, server designers, systems engineers, systems analysts and integrators, marketing and product engineers and managers, technology managers, test engineers, telco and enterprise end users, server and data center managers, consultants, design and communications specialists, and design service providers.
About the Chairperson:
Sergis Mushell is a principal research analyst with Gartner's Technology and Service Provider Research group. His primary focus is on processor technology, NICs, and storage semiconductors. He has previous experience with Phaselink, Altera, and IDT. He has worked within standard bodies, such as JEDEC and IEEE, on new standards and initiatives. |
|||
|
7:30-9:00pm |
Organizers: Frank Chang, Vitesse and Jonathan Hinkle, Viking Technology
Table Leaders:
Session Description:
The Beer and Pizza Session is an opportunity for attendees to meet top experts in many crucial areas and ask questions in an informal setting. Each table has a different subject, and attendees are welcome to move from table to table. Table subjects will include cooling methods, memory, storage performance, high-speed links, virtualization, security, caching methods, application acceleration, server accelerators, cloud servers, and distributed computing. Beer, wine, soft drinks, and pizza will be served to promote the informal atmosphere and encourage networking. Emphasis will be on frequently asked questions, best practices, hints and warnings, major issues, and key products and standards.
Intended Audience:
Hardware designers, project engineers, software designers, software engineers, technology managers, communications equipment designers, systems analysts and integrators, engineering managers, consultants, design specialists, design service providers, product managers, marketing communications specialists, marketing engineers and managers, and test engineers.
About the Organizers:
Jonathan Hinkle is the Memory Systems Architect at Viking Technology, a Sanmina-SCI company. He leads the division's efforts to define and drive future technology and products with a current emphasis on hybrid memory and advanced SSDs. Most recently, he developed the SATADIMM SSD, an SSD built to occupy a DIMM slot on a circuit board. Previously, he worked at IBM developing x86 systems from high-end rackmount boxes to blade servers. At IBM, he also invented and drove standardization of the Very Low Profile (VLP) DIMM. He is actively involved in setting JEDEC industry standards for memory products. Mr. Hinkle is a Senior Member of the IEEE.
Frank Chang is a principal engineer for Vitesse Semiconductor. He specializes in optical system engineering, IC product specifications, and application issues for telecom, datacom, and PON access markets. He is also knowledgeable in electronic dispersion compensation (EDC), PMD, PHY, and FEC chipsets and their application in optical networking. He has authored or co-authored over 55 peer-reviewed journal and conference articles.
Frank is active in standard-setting bodies including OIF/ITU, IEEE 802.3, and FSAN for the definition of various optical specifications. He is currently serving in the OFC/NFOEC Program Committee under Technical Subcommittee B: Network Technologies, IEEE LEOS Technical Committee on “Mitigating Channel Degrading Effects,” the IEEE P802.3ba (formerly HSSG) 40G, and the 100G Ethernet Task Force organization. |
|||
| Wednesday, November 30th | ||||
|
8:30-9:45am |
Chairperson: Steve Ohr, Gartner
Paper Presenters:
Session Description:
There are many approaches to managing power for servers and data centers. One can focus on the power grid or on the uninterruptible power supplies (UPSes) that provide backup power — but often use a surprising amount of energy themselves. One can eliminate transformers to save energy, but that may decrease reliability. And despite the advantages of aggregating computing resources, an argument can be made for distributing tasks as a means of minimizing power usage. This session will offer insights into fresh approaches to data center power conservation.
Intended Audience:
Hardware designers, engineering managers, server designers, systems engineers, systems analysts and integrators, marketing and product engineers and managers, technology managers, test engineers, telco and enterprise end users, consultants, design and communications specialists, and design service providers.
About the Chairperson:
Steve Ohr is Gartner’s research director for analog and power management. His current research focuses on data center power and cooling techniques, especially the means of measuring and controlling electrical power consumption. Before joining Gartner, Ohr served as an editor and website manager for EE Times. His editorial career includes assignments with Computer Design magazine, Asian Electronic Engineer and Electronic Design. He has received a Jesse Neil Award for his reporting on microprocessor development. He holds an MS in sociology from Rutgers and a BS in industrial engineering from New Jersey Institute of Technology. |
|||
|
8:30-9:45am |
Chairperson: Tom Burniece, Burniece Consulting Services
Paper Presenters:
|
|||
|
8:30-9:45am |
Chairperson: Andras Boross, SunEdison
Paper Presenters:
|
|||
|
10:00-11:00am
|
Chairperson: Dileep Bhandarkar, Microsoft
Speaker: Amir Michael, Facebook
Course Description:
This talk will review the Open Compute Project, an initiative to share the custom-engineered technology in Facebook¹s first dedicated data center. The technology delivered a 38 percent increase in energy efficiency at 24 percent lower cost. Inspired by the success of open source software, and aiming to encourage industry-wide collaboration around best practices for data center and server technology, Facebook has published technical specifications and mechanical CAD files for its data center¹s servers, power supplies, server racks, battery backup systems, and building design.
The data center achieved an initial power usage effectiveness (PUE) ratio of 1.07, compared with 1.5 for our existing facilities, which already fell into the “best practices” category defined by the U.S.EPA. Established by the Green Grid in 2007, PUE measures data center energy efficiency - the lower the number, the better. The session will describe the motivations behind the project, the design of the data center and servers, and our open strategy. It will also cover our plans to drive both technological and environmental efficiency at Facebook and in the broader industry, Intended Audience:
Hardware designers, engineering managers, server designers, systems engineers, systems analysts and integrators, marketing and product engineers and managers, technology managers, test engineers, telco and enterprise end users, consultants, design and communications specialists, and design service providers.
About the Presenter:
Amir Michael is Manager of Hardware Design at Facebook. With Facebook since 2009, he leads the team responsible for the design and implementation of the servers that power one of the world’s most heavily trafficked websites.
Amir has focused on building an extremely efficient computing infrastructure at the lowest possible cost. He plays an active role in the Open Compute Project, Facebook’s initiative to share the custom-engineered technology in its first dedicated data center. Before joining Facebook, Amir worked at Google where he developed server and data center infrastructure. He holds a Bachelor of Science in Electrical and Computer Engineering from the University of California, Santa Barbara. About the Chairperson:
Dr. Dileep Bhandarkar joined Microsoft as a Distinguished Engineer responsible for Server Hardware Architecture and Standards for Global Foundation Services (GFS) in May 2007. He is currently Chief Architect responsible for the technology roadmap for the compute infrastructure for Microsoft’s online services. He was elected an IEEE Fellow in 1997 for contributions and technical leadership in the design of complex and reduced instruction set architecture and in computer system performance analysis. He has previously held senior technical and management positions at Intel, Digital Equipment, and Texas Instruments. Dr. Bhandarkar holds 16 U.S. Patents and has published more than 30 technical papers in various journals and conference proceedings. He is also the author of a book titled Alpha Implementations and Architecture. In 1998, he was recognized as a Distinguished Alumnus of the Indian Institute of Technology, Bombay, where he received his B. Tech in Electrical Engineering in 1970. He also has a M.S. and Ph.D. in Electrical Engineering from Carnegie Mellon University, and has done graduate work in Business Administration at the University of Dallas. |
|||
|
11:00-11:30am |
Speaker: Keith Klemba, Vice Persident, SAP Research
Introducer: Rich Fetik, Data Confidential
Abstract:
Today's highly competitive economy requires enterprises to get the most from all the data they have. They need to analyze it thoroughly and quickly - and be able to act on it right away. If you can't keep up with today's flood of information at top speed, your competitors will simply pass you by. Symmetric multiprocessing, large memories, in-memory databases, and other technical advances provide the basis for handling huge amounts of data. New business application architectures allow you to take advantage of them to make key decisions quickly.
About the Speaker:
Keith Klemba is Vice President, SAP Research, Palo Alto, California. In his current role, he is architecting and implementing next generation computing systems for in-memory computing. His experience brings together a keen sense of Systems Engineering with grounded Business awareness. He is considered an expert in Network Communications, Security, Multimedia, and Systems Engineering.
Prior to joining SAP, Keith served as Co-Founder and Chief Architect for Firetide Inc. a wireless mesh networking company specializing in city-wide video surveillance systems. Before starting at Firetide, Keith had a long career with Hewlett-Packard as a Chief Technical Adviser. While at HP he served as the Security Technology Coordinator for the Computer Organization and as Principal Architect for Hewlett-Packard's award winning network management framework "OpenView". Prior to Hewlett-Packard, Keith worked for VitaLink Communication Corp. where he was the product development manager for their worldwide commercial satellite-earth station monitoring and control system. Before Vitalink, Keith was a Principal Investigator in Telecommunications Engineering at SRI International. While there he specialized in Packet Switched Radio Communications, Secure Communications, Internet Protocols, and Global Weather Analysis. Keith holds over a dozen patents in the areas of Security, Communications, and Networking. He is considered an expert in the area of Wireless Networking and Networking Infrastructures. About SAP:
As the market leader in enterprise application software, SAP (NYSE: SAP) helps companies of all sizes and industries run better. From back office to boardroom, warehouse to storefront, desktop to mobile device – SAP empowers people and organizations to work together more efficiently and use business insight more effectively to stay ahead of the competition. SAP applications and services enable more than 170,000 customers to operate profitably, adapt continuously, and grow sustainably. For more information, visit www.sap.com.
|
|||
|
11:030am-Noon |
Speaker: Adam Roberts, Solid State Memory Architect and Senior Engineer - Storage, IBM
Introducer: Anthony Anwari, VP Sales, STEC
Abstract:
Coming soon.
About the Speaker:
Adam Roberts is the Solid State Memory Architect and Senior Engineer for the IBM System x Storage Development Group. He is responsible for developing the strategic direction of Solid State based architectures and solutions for the IBM System x development team based in Research Triangle Park, North Carolina. Roberts has held a variety of positions in his 20 years of Engineering Experience including over 10 years with IBM in both Server Development and Server Storage Development. Prior to IBM, Roberts held the position of Pharmaceutical Engineer for Wyeth Ayerst. He holds a Bachelor's Degree in Computer and Electrical Engineering at North Carolina State University. |
|||
|
2:00-3:15pm |
Chairperson: Samba Murthy, Independent Investor
Panelists:
|
|||
|
2:00-3:15pm |
Chairperson: Jim Harrison, Electronics Products Magazine
Panelists:
|
|||
|
3:30-5:00pm |
Chairperson: Jean Bozman, IDC
Panelists:
Course Description:
Current issues in the server area include:
This session will discuss all these issues, including distributed approaches that involve many small, low-cost devices, the fabrics to connect them, and the software to coordinate their processing.
Intended Audience:
Hardware designers, engineering managers, server designers, systems engineers, systems analysts and integrators, marketing and product engineers and managers, technology managers, test engineers, telco and enterprise end users, consultants, design and communications specialists, and design service providers.
About the Presenter:
Jean Bozman is Research Vice-President in IDC’s Enterprise Server Group, where she analyzes the worldwide server market and manages Clustering and Availability Software (CLAS) market research. She has over 20 years experience covering markets for operating environments, servers, and server workloads. Ms. Bozman is widely quoted in business publications, including BusinessWeek and Investors Business Daily; in daily newspapers, including the San Jose Mercury News and Los Angeles Times; and in online publications, such as CNET (news.com), Bloomberg, and Reuters. Before joining IDC, Ms. Bozman was with Computerworld where she started the Chicago Bureau and later served as Senior West Coast Editor. She holds a B.S. from SUNY Stony Brook and a master's degree from Stanford. |
|||
