Data Center Knowledge | News and analysis for the data center industry - Industr's Journal
 
[Most Recent Entries] [Calendar View]

Wednesday, August 16th, 2017

    Time Event
    12:00p
    Equinix Makes Big Bet on Fuel Cell-Powered Data Centers

    Two years after it did a pilot fuel-cell installation at one of its Silicon Valley data centers, Equinix is making a big bet on the technology, which uses a chemical process to convert natural gas to electricity.

    The Redwood City, California-based data center provider today announced a deal that will see fuel cells installed at 12 additional US data centers. The deal will be financed by the utility Southern Company, with whom Equinix signed a power purchase agreement for a total of about 37MW of generation capacity, which will be the largest single deployment of fuel cells for data centers to date.

    The fuel cells will be supplied by Bloom Energy, a leader in the market whose energy servers have been deployed on corporate campuses, at data centers by eBay, Apple, NTT, CenturyLink, and at two Verizon data centers Equinix took over when it acquired the large data center portfolio from the telco earlier this year.

    Fuel cells produce significantly fewer carbon emissions than traditional gas-fueled power plants; they also don’t require water, unlike power plants, which consume massive amounts of water to generate energy. Apple has made fuel cells part of its corporate sustainability program, using Bloom’s technology to provide 10MW of capacity for its Maiden, North Carolina, data center campus.

    The technology, which produces energy on-site, is also considered more reliable than America’s old, outage-prone electrical grid. eBay is using Bloom fuel cells as a primary energy source for its data center in Utah, relying for backup on the electrical grid instead of diesel generators one sees in most data centers.

    The most commonly cited drawback of fuel cells is their high upfront cost, which has been addressed with various financing schemes, such as Equinix’s agreement with Southern. In 2013 Bloom created a leasing program for its energy servers, giving customers the option of paying for the product over time, financed by Bank of America.

    The deal will bring Equinix’s total fuel-cell capacity to more than 40MW, including its existing Silicon Valley installation and the two Verizon sites, in Los Angeles and New York.

    New fuel cells will be installed at the following Equinix data centers:

    • Silicon Valley: SV1 through SV6, plus SV10;
    • New York metro: NY2, NY4, and NY5;
    • Los Angeles metro: LA3 and LA4

    Customer appetite for colocation data center services powered by renewable energy has been on the rise in recent years, and providers like Equinix have been responding by making public commitments to renewable energy goals and by striking the types of utility-scale power purchase agreements that Google pioneered to clean up its data center energy supply.

    Equinix, the world’s largest colocation and interconnection company, has a long-term goal of powering its data centers with 100 percent renewable energy. According to its most recent sustainability report, it was at 56 percent renewable as of last year.

    3:00p
    Embedded FPGA: Coming to a Data Center Near You

    Geoff Tate is CEO and co-founder of Flex Logix, Inc.

    FPGA chips have increasingly emerged as a valuable tool in creating a reconfigurable data center: reconfigurable networking adapter cards, reconfigurable networking systems products, and reconfigurable data center projects such as Microsoft’s Project Catapult, Microsoft’s Azure SmartNIC and Amazon EC2 F1.

    However, it has recently been proven that embedding this technology on chip can speed performance even more. That is why we are now seeing a host of embedded FPGA companies emerge. This was also the main driver behind Intel paying $16 billion for Altera. Through this strategic acquisition, Intel can first integrate Xeon processors and FPGA chips in multi-chip packages and then eventually integrate FPGA onto the same die as the CPU.

    With all this activity, it is clear that embedded FPGA is making its way into the data center in a rapid way, and not soon enough considering the bandwidth demands that are now being placed on them. Every data center manager today wants to build their facility to be more reconfigurable  and programmable as opposed to the old PC-era age of hardware that needs to be replaced to keep up with the changing standards and protocols.  Embedded FPGA allows this to happen for the first time. This is a win-win for everyone since the data center gets the flexibility that they need and the chip designers can extend the life of their chips and target more markets with those same chips.

    Embedded FPGA vs Traditional FPGA

    While standard FPGA has delivered many benefits to the data center, it does not solve the performance bottlenecks in chip-level data center usages in the PCIe bus. However, when the FPGA is integrated in a multi-chip package, the bus widths and speeds possible through a interposer are 10x higher than over the PCIe bus. This performance can go up yet again if the FPGA is integrated onto the same die because the buses possible on silicon are much wider and faster than interconnects on PCB boards/connectors.

    Also, there are applications such as packet parsers where FPGA is useful in many, distributed small chunks integrated in the networking chip.

    What is Embedded FPGA?

    Embedded FPGA is now available and silicon has been proven in multiple, popular process nodes. This technology is quickly finding a home in micro-controllers, IoT, deep learning, SoCs/ASICs, wireless base stations as well as in networking/data centers. However, many people still don’t know the difference between this technology and mainstream FPGAs offered by companies such as Xilinx. In the following few paragraphs, we’ll discuss some of the differences.

    The basic building block of an FPGA is a LUT, or look-up-table, which when combined with an optional, by-passable flip flop enables programmable logic of any function. The programmable logic is available in a fabric of programmable interconnect which allows any logic block to connect with any other logic block. Verilog or VHDL are used to code FPGA: software maps the code onto the FPGA fabric and produces a bitstream which when loaded in the FPGA causes it to execute the RTL as desired.

    Networking/data center applications require 28nm or 16/14nm performance and only three suppliers have embedded FPGA in these advanced processes. The big performance benefit of embedding FPGA is the bandwidth the FPGA can have with the rest of the chip. Even a small embedded FPGA, such as 5K LUTs, can have 1000 inputs and 1000 outputs with data rates of >500MHz in 28nm or more in 16/14nm.

    In an FPGA chip, typically there is a mix of programmable logic blocks (logic cells = LUT + flip-flop), DSP/MAC (16×16 or larger multiplier-accumulators, useful for signal processing or deep learning), and Block RAM (a reconfigurable dual-port memory).

    With embedded FPGA, there is also more choices on the ratio of logic/MAC/memory. This is very important because in high volume chips, customers don’t want to have functionality they don’t need.

    In some cases, chip companies may prefer initially to use an FPGA in a multi-chip module, such as in Intel’s first use of FPGA with Xeon. There are also suppliers that offer support to implement “FPGA Chiplets” for 2.5D multi-chip packaging.  Chip-to-chip performance is less than on-chip buses, but much faster than the PCIe bus.

    Example Applications of Embedded FPGA in Data Centers

    While embedded FPGA is new, we expect volume applications to emerge in the next few years, including:

    • Xeon with integrated FPGA
    • SmartNICs (a la Microsoft)
    • Reconfigurable switches with programmable packet parsers
    • Deep learning

    Harvard University recently announced it is integrating embedded FPGA into its deep learning chip in 16nm, which is now in fabrication. The reason is deep learning algorithms are evolving rapidly; with embedded FPGA, the algorithms can evolve in real time, not once a year when a new chip tapes out.

    The Future

    Every data center in the world will benefit from being more programmable and reconfigurable in order to keep up with the constantly changing standards and protocols. In the past, data centers typically had to wait for new hardware and chips to meet these new requirements. This is all changing with embedded FPGA and data centers will soon find themselves able to quickly and seamlessly stay ahead of the curve, and ahead of the world’s bandwidth demands.

    Opinions expressed in the article above do not necessarily reflect the opinions of Data Center Knowledge and Informa.

    Industry Perspectives is a content channel at Data Center Knowledge highlighting thought leadership in the data center arena. See our guidelines and submission process for information on participating. View previously published Industry Perspectives in our Knowledge Library.

    3:29p
    London Leads in Private Data Center Interconnection Ahead of Brexit

    Jeremy Kahn (Bloomberg) — London is set to retain its position as the city with the greatest capacity in  private data links, a lead that may help stem any erosion in the city’s position as Europe’s premier financial center, even after the U.K. leaves the European Union.The U.K. capital currently has an installed bandwidth of 159 terabytes per second, a figure that exceeds any other European city and is greater than the capacities in either New York or Silicon Valley, according to a new report from Equinix Inc., a U.S. company specializing in data centers that host many of these private data connections.

    The report, released Wednesday, forecasts that London will retain this lead, with its overall private bandwidth tripling to 486 terabytes per second by 2020.

    Businesses are turning to private data networks due to the huge amounts of data they are now generating, where public internet connections may struggle, said Russell Poole, managing director of Equinix’s U.K. business.

    The rapid transfer of data has long been of importance to the financial sector. Exchanges have been letting traders place their computer servers next to exchange systems -– a practice known as colocation -– in order to facilitate high frequency trading.

    After telecommunications companies and firms that sell cloud and IT services, the banking and insurance sector is the largest user of private networks, Equinix said. And by 2020, the sector will exceed even those network-centric industries in its use of private data infrastructure.

    Connections between businesses that have co-located their data in the same or adjacent data centers will “dwarf global internet traffic in three years,” Poole said.

    Frankfurt, which has been trying to lure financial firms away from London since the Brexit vote, is also rapidly adding data centers and co-location facilities. The city will see its private network capacity expand almost 50 percent per year through 2020, the report said. But, at 252 terabytes per second, Frankfurt will continue to lag London by a large margin.

    “The U.K. is already the largest collocation market in Europe and the third largest in the world due to its well-established interconnectivity,” said Darren Watkins, managing director for Virtus Data Centres Ltd., adding that since the Brexit vote, he had not seen any decline in business in the U.K.

    Amsterdam, which is currently ranked second in Europe after London in the Equinix assessment, will see its capacity grow 42 percent per year to 242 terabytes per second by 2020. Paris, which has also tried to position itself as a rival post-Brexit financial hub to London, will only have 120 terabytes per second of installed capacity at that point.

    5:20p
    Equinix: Private Data Exchange is Outpacing Internet’s Growth

    Total bandwidth on the private links companies use to interconnect their networks directly is growing faster than total bandwidth of the public internet and three years from now will outpace growth of the internet’s total bandwidth by almost two times, reaching about six times the volume of global IP traffic, according to a study by Equinix, the world’s largest data center colocation and interconnection provider.

    In the first study of its kind, Equinix has attempted to quantify private interconnection and create a forecast model for growth of this type of network links. For the company, which charges customers monthly fees for every such link provisioned inside its data centers, interconnection has been the fastest-growing source of revenue; and it projects this business will only continue to accelerate, as digital business transformation marches on, spurring companies to seek more and more partners and service providers to link to without involving the internet, which is considered slower, less reliable, and less secure than private connections.

    “The worldwide reliance on interconnection bandwidth is growing rapidly as businesses competing within this evolving landscape mature from exchanging traffic over a few connections through a single network carrier as the intermediary, to activating multiple direct interconnection exchange points among a global ecosystem of providers and counterparties,” the analysts wrote.

    What kind of bandwidth are we talking about here? The report projects private interconnection to reach north of 5,000 Tbps by 2020. That’s enough to transmit all printed content in the US Library of Congress three times within one second, according to the analysts; or process more than half a million electronic payments per minute:

    “Assuming the average value is $50, that’s $27.5M per minute or $1.6B per hour.”

    So who’s linking to whom exactly? At the highest level of abstraction, Equinix splits the participants into enterprises and service providers.

    Enterprises are described as organizations that aren’t at the core in the business of providing infrastructure-oriented digital services; these are your banks, insurers, manufacturers, traders, professional services firms, energy companies, government agencies, etc.

    Service providers by Equinix’s definition are companies whose primary business is telecommunications, cloud and IT, as well as content and media.

    Here’s how both segments’ interconnection bandwidth needs will change between 2016 and 2020, according to Equinix (click chart to enlarge):

    (Source: Equinix)

    Both categories, however, have use cases for private interconnection with network providers, cloud and IT providers, financial services, supply chain partners, and content companies. Not surprisingly, the amount of bandwidth for interconnecting with network providers far outstrips all other use cases today, and in the analysts’ 2020 projections (click chart to enlarge):

    (Source: Equinix)

    Interconnection bandwidth demand by enterprises — considered the largest growth area for all manner of service providers, from data center companies like Equinix to network, IT, and cloud providers – is projected to skyrocket over the next few years, going from 12 Tbps of installed capacity in 2016 to 547 Tbps in 2020, growing 160 percent annually (click chart to enlarge):

    (Source: Equinix)

    US has the largest portion of total interconnection bandwidth today and is expected to retail that lead over the next three years, while Latin America will have the highest annual bandwidth growth rate within that timeframe: 62 percent per year, reaching 620 Tbps of installed capacity by 2020.

    Download the full report here.

    7:56p
    Choose Your Own Adventure: Data Center World Workshops

    As our Data Center World Global Content Advisory Team is working hard to build the best agenda yet, we’re leaving some of the most important choices up to you! This year we’re crowdsourcing more than ever. We need your help to select the workshops at Data Center World Global. The three* with the most votes will be offered at Data Center World Global on Monday, March 12.

    Data Center World workshops are 3.5 hour interactive, educational sessions that take a deep-dive into a topic that is of current importance to the data center industry. Workshops include in-depth discussions and hands-on activities for an engaging and immersive experience. Workshop attendees will leave with practical experience as well as tools, plans, and checklists to implement their new knowledge into their organization immediately.

    Attendance in a workshop is included in the All Access Pass along with other VIP benefits. All workshops will take place on Monday, March 12, 2018.

    Cast your vote today! (Scroll down to read the descriptions)

    Take Our Poll

    WORKSHOP DESCRIPTIONS

    Data Center Disaster Recovery – Not Just for the BC/DR Team!

    Everyone’s job is disaster preparation and recovery. Not just those who have it in their job title. Through this interactive workshop, participants will learn about best practices in data center disaster planning and recovery through solving real-life scenarios. Small groups made up of varying roles in a data center will work together to develop optimal solutions to challenges they may face in their data center. Participants will walk away with practical experience and a disaster prevention and recovery plan template to implement in their data center.

    Assessing Data Center Models

    There are more data center delivery model options than there ever have been before and with that comes more fear, uncertainty, and doubt. Take a hard look at the pros/cons, considerations, and questions to ask to determine the right hybrid architecture for your data center: on-premises, colocation, cloud (public, private), or all of the above! Through small interactive workgroups, participants will utilize a check list of key questions to be answered in evaluating strategy alternatives to determine the best model option for any given business scenario.

    Preventing and Addressing Data Center Security Threats

    Are you ready to respond to a cyber-attack? Find out through a series of hypothetical (or real!) data center security breach scenarios. Workshop attendees will participate in hands-on “war games” to evaluate standard data center security response plans and where improvements need to be made in their own data centers. Through this interactive workshop, small groups will work together to learn about their company’s ability to react to a security attack by exploring best practices for identifying, containing, communicating, and adjusting business strategies in the wake of a security breach.

    What does Hyperconvergence Really Mean for Data Centers?

    According to Gartner, by 2019, approximately 30% of the global storage array capacity installed in enterprise data centers will be deployed on software-defined storage (SDS) or Hyperconverged integrated system (HCIS) architectures based on x86 hardware systems, up from less than 5% today. Through this interactive workshop, participants will explore what data center managers need to know NOW to be proactive and prepared.

    Data Center Capacity Optimization – for IT People Only!

    Based on the popularity of last year’s workshop, “Data Center Construction – for IT People Only,” this year we will offer training specific for data center IT staff on capacity optimization. Have you been in a situation where IT says the data center is full but Facilities says otherwise? This hands-on workshop will explore capacity optimization through various real-life data center scenarios. After learning the key considerations and tools for capacity planning, participants will work in small groups to solve problems, and then compare their results with 3D models of the optimized solutions. Participants will walk away with a checklist to evaluate data center capacity which can also be used as a scorecard for comparing and benchmarking their own data centers.

    Effective Communication for Non-Communicators

    Everyone communicates. Few connect. Good communication is crucial to business success; however workplace communication is rarely adequate. The good news is that you don’t have to be an extrovert or even particularly nurturing to communicate effectively. In this workshop, participants will learn five principles and five practices to connection one-on-one, in a group, or with an audience. You’ll see all of your relationships, from your boss to your kids, from your teammates, to your neighbors, take a quantum leap.

    HOW ELSE CAN YOU HAVE YOUR VOICE HEARD?

    Submit a session proposal! The call for proposals deadline is October 1, 2017.

    *Final selection of workshops is subject to change and dependent upon workshop instructor availability.

    10:53p
    IBM CEO Says Trump Advisory Forum Could No Longer Serve Purpose

    (Bloomberg) — IBM Chief Executive Officer Ginni Rometty, who had been a member of U.S. President Donald Trump’s strategy and policy advisory group, said the panel was dissolved Wednesday after members determined that it could “no longer serve the purpose for which it was formed.”

    “In the past week, we have seen and heard of public events and statements that run counter to our values as a country and a company,” she wrote in an email to International Business Machines Corp. employees that was shared with Bloomberg. “Earlier today I spoke with other members of the Forum and we agreed to disband the group.”

    Despite facing some criticism in the past from employees, Rometty had been steadfast in her stance that it was better to have a seat at the table so that she could help influence policies in a way that would be beneficial for the computer giant. That changed after a number of business leaders began exiting Trump’s corporate advisory groups this week, following his remarks about violence at a white supremacist gathering and counter-protest in Charlottesville, Virginia.

    See also: Energy Department Awards $258M to Develop Exascale Supercomputers

    As those executive departures began to pile up, Rometty was among the remaining CEOs who were facing questions about their positions and whether they would also quit the policy group. An hour after reports surfaced that the forum was planning to disband, Trump announced that two panels, the Manufacturing Council and the Strategy and Policy Forum, would be abolished.

    “We have always believed that dialogue is critical to progress; that is why I joined the President’s Forum earlier this year,” Rometty wrote in the email. “IBM will continue to work with all parts of the government for policies that support job growth, vocational education and global trade, as well as fair and informed policies on immigration and taxation.”

    << Previous Day 2017/08/16
    [Calendar]
    Next Day >>

Data Center Knowledge | News and analysis for the data center industry - Industry News and Analysis About Data Centers   About LJ.Rossia.org