Category Archives for "Network World Data Center"

Nvidia launches new hardware and software for on-prem and cloud providers

Nvidia used its GPU Technology Conference in San Jose to introduce new blade servers for on-premises use and announce new cloud AI acceleration.The RTX Blade Server packs up to 40 Turing-generation GPUs into an 8U enclosure, and multiple enclosures can be combined into a "pod" with up to 1,280 GPUs working as a single system and using Mellanox technology as the storage and networking interconnect. Which likely explains why Nvidia is paying close to $7 billion for Mellanox.Instead of AI, where Nvidia has become a leader, the RTX Blade Server is positioned for 3D rendering, ray tracing and cloud gaming. The company said this setup will enable the rendering of realistic-looking 3D images in real time for VR and AR.To read this article in full, please click here

Data center fiber to jump to 800 gigabits in 2019

The upper limits on fiber capacity haven't been reached just yet. Two announcements made around an optical-fiber conference and trade show in San Diego recently indicate continued progress in squeezing more data into fiber.In the first announcement, researchers say they’ve obtained 26.2 terabits per second over the roughly 4,000 mile-long trans-Atlantic MAREA cable, in an experiment; and in the second, networking company Ciena says it will start deliveries of an 800 gigabit-per-second, single wavelength light throughput system in Q3 2019.High-speed laser MAREA, translated as “tide” in Spanish, is the Telefónica-operated cable running between Virginia Beach, Va., and Bilbao in Spain. The fiber cable, initiated a year ago, is designed to handle 160 terabits of data per second through its eight 20-terabit pairs. Each one of those pairs is thus big enough to carry 4 million high-definition videos at the same time, network-provider Infinera explains in an Optical Fiber Conference and Exhibition published press release.To read this article in full, please click here

Startups introduce new liquid cooling designs

With the increase in compute density making air cooling less and less feasible, liquid cooling is going mainstream. For data centers. Overclockers have been doing it for years.For the most part, liquid cooling involves piping in cooled water to a heat sink attached to the CPU. The water then cools the heat sink, and is pumped away to be circulated and cooled down.But there are some cases where immersion is used. That is where the entire motherboard is submerged in a nonvolatile liquid. Immersion is used in only the most extreme of cases, with the highest compute density and performance. For a variety of reasons, it isn’t widely used.One startup that hopes to change that showed its wares at the Open Compute Project Summit 2019, which ran last week in San Jose. The OCP has a special project called Advanced Cooling Solutions to promote liquid cooling and other advanced cooling approaches.To read this article in full, please click here

Cisco spreads AI across Webex meetings

Cisco Webex has rolled out a package of AI-based features that brings together recently acquired technologies it says will make business meetings more efficient and intuitive.The Webex conferencing tool enhancements, which include faster meeting startup, a better way to know the people attending a meeting and facial-recognition improvements will help customers more effectively collaborate from any location, the company said. Read about SD-WAN How to buy SD-WAN technology: Key questions to consider when selecting a supplier How to pick an off-site data-backup method SD-Branch: What it is and why you’ll need it What are the options for security SD-WAN? Cisco bought Webex in 2007 for about $3.2 billion with an eye toward competing more effectively with Microsoft and other collaboration software vendors. Today Webex conferencing tools are used by over 130 million customers a month, Cisco says.To read this article in full, please click here

DoE plans world’s fastest supercomputer

The U.S. Department of Energy says it is working on a supercomputer that will break the target of exaFLOP computation – a quintillion (1018)  floating-point computations per second – in order to handle high-performance computing and artificial intelligence.Being built in conjunction with Intel and Cray Computing,  the Aurora supercomputer will  cost more than half a billion dollars and be turned over to Argonne National Laboratory in Chicago in 2021, according to a statement by the DoE. [Click here to see the current top 10 fastest supercomputers.]To read this article in full, please click here

Exec: How SDN, SD-WAN, security fit in VMware’s strategy

It has been just 10 months since Tom Gillis became VMware's senior vice president and general manager of its networking and security business, and in that time he has overseen some major changes in the company’s core products.Most recent is a milestone release of the company’s NSX-T Data Center software, making it VMware’s primary networking platform for organizations looking to support multivendor cloud-native applications, bare-metal workloads as well as the growing hybrid and multi-cloud worlds.To read this article in full, please click here

How did Facebook go down despite multiple data centers?

The Mercury retrograde kicked in big time on Wednesday as Facebook suffered an eight hour-outage that also affected Instagram and Facebook Messenger.No one was believed to be harmed; a few might have even had offline interactions with other human beings. Learn about backup and recovery: Backup vs. archive: Why it’s important to know the difference How to pick an off-site data-backup method Tape vs. disk storage: Why isn’t tape dead yet? The correct levels of backup save time, bandwidth, space Facebook said it wasn’t an attack, like a Denial of Service attack, and has since issued a statement attributing it to a configuration error.To read this article in full, please click here

The noise in fiber could be used to increase data capacity

Increasing the capacity of fiber-optic cables might one day be possible through the exploitation of a part of the signal commonly thought of as substandard. That imperfect element in a carrier, called “noise” is usually something one tries to avoid—it can muddy the accurate reading of the data.However, scientists now suggest that one could, in fact, embrace the rubbishy, and thus far unusable, part of the signal to hold data and allow it to be decoded. The ordinarily data-obscuring hubbub could potentially be harnessed and used to increase data capacity in light waves.“Information is encoded in the correlated noise between spatially separated light waves,” writes Oliver Morsch in an article on the website of ETH Zurich, a technical and scientific university. “The new coding technology, developed by ETH researchers, makes it possible to make better use of the transmission capacity of optical fibers.”To read this article in full, please click here

Data center giants announce new high-speed interconnect

A group of big names in the data center space have linked arms to develop yet another high-speed interconnect, this one designed to connect processor chips.It's called Compute Express Link, or CXL, which is aimed at plugging data-center CPUs into accelerator chips. Members of the alliance that developed the spec are Intel, Microsoft, Google, Facebook, HPE, Cisco, and Dell-EMC, plus Huawei and Alibaba.[ Learn who's developing quantum computers. ] Where are IBM, AMD, Nvidia, Xilinx, or any of the ARM server vendors such as Marvell/Cavium? They have their own PCIe-nased spec, called CCIX. The group consists of AMD, Arm, Mellanox, Qualcomm, Xilinx, and Huawei.To read this article in full, please click here

Nvidia grabs Mellanox out from under Intel’s nose

After months of speculation, Mellanox found a suitor -- and it was a surprise, to say the least. GPU leader Nvidia snatched up the networking vendor for $6.9 billion, topping a rumored previous offer of $6 billion from Nvidia’s nemesis, Intel.The acquisition ends months of rumors of a suitor for Mellanox. Intel, Microsoft, and Xilinix were all reportedly bidding for the Israeli company, which specializes in high-speed networking. [ Read also: How to plan a software-defined data-center network ] Mellanox Technology was formed in 1999 by a former Intel executive and was a pioneer in the early adoption of InfiniBand interconnect technology, which along with its high-speed Ethernet products is now used in over half of the world’s fastest supercomputers and in many leading hyperscale data centers.To read this article in full, please click here

Facebook gets into the fiber-optic connectivity business

When you think of Facebook services, high-speed connectivity is not the first thing that comes to mind. But the social media giant is doing just that, offering high-capacity fiber-optic routes to sell unused capacity between its data centers for third parties.Facebook has created a subsidiary called Middle Mile Infrastructure to sell excess capacity on its fiber, starting with new fiber routes between its data center campuses in Virginia, Ohio, and North Carolina. The company made the announcement in a blog post by Kevin Salvadori, director of network investments.To read this article in full, please click here

4 Tips for Easier Edge Deployments

There are plenty of reasons for organizations to embrace edge computing. By moving applications, data, and computing services to the edge of a network, as opposed to a large data center or cloud, organizations can lower operating costs, improve application performance, reduce network traffic, and achieve real-time data analytics.As more organizations come to know the advantages of edge of network devices, many are eyeing deployments. In fact, according to the IDG 2018 State of the Network, 56% of networking professionals have plans for edge computing in their organizations.To read this article in full, please click here

The data center is being reimagined, not disappearing

I have documented more than once that the data center is not going away; it’s being reimagined. And now comes a report with greater details on that change.Spiceworks has released its 2019 State of Servers report that examines on-premises server infrastructure in the workplace, including purchase plans, brand prevalence, and perceptions. The results of the survey, which was conducted in February and included 530 IT buyers from organizations across North America and Europe, show that 98 percent of businesses currently run on-premises servers, and 72 percent of businesses plan to purchase new server hardware within the next three years. To read this article in full, please click here

The Open Compute Project is quickly gaining ground

Eight years ago, Facebook launched the Open Compute Project (OCP), an open-source hardware initiative to design the most energy-efficient server gear for massive, hyperscale data centers. The promise was flexibility of hardware and software and designs for greater power efficiency.Very quickly, Intel, Rackspace, Goldman Sachs and Sun Microsystems' co-founder Andy Bechtolsheim joined with Facebook to launch the OCP project, with Microsoft joining in 2014.The project has hummed along quietly with no sales figures until now, thanks to supply chain market research specialists IHS Markit. It surveyed both Facebook, Microsoft, and Rackspace, as founding partners, and looked at sales to customers beyond those three.To read this article in full, please click here

New chemistry-based data storage would blow Moore’s Law out of the water

Molecular electronics, where charges move through tiny, sole molecules, could be the future of computing and, in particular, storage, some scientists say.Researchers at Arizona State University (ASU) point out that a molecule-level computing technique, if its development succeeds, would slam Gordon Moore’s 1965 prophesy — Moore's Law — that the number of transistors on a chip will double every year, and thus allow electronics to get proportionally smaller. In this case, hardware, including transistors, will conceivably fit on individual molecules, reducing chip sizes much more significantly than Moore ever envisaged.[ Now read: What is quantum computing (and why enterprises should care) ] “The intersection of physical and chemical properties occurring at the molecular scale” is now being explored, and shows promise, an ASU article says. The researchers think Moore’s miniaturization projections will be blown out of the water.To read this article in full, please click here

VMware preps milestone NSX release for enterprise-cloud push

Looking to ease deployments of software-defined networks while reinforcing automation and security for hybrid and multicloud customers, VMware has taken the wraps off of a major release of its NSX-T Data Center software.While the NSX-T 2.4 announcement includes over 100 upgrades, VMware said the release anoints NSX-T as the company’s go-to platform for future software-defined cloud developments.[ Also see How to plan a software-defined data-center network and Efficient container use requires data-center software networking.] “This is NSX-T’s coming out party—it is now our primary platform and includes all the tools, services, security and support for future growth,” said Tom McCafferty, VMware’s senior director of product marketing for NSX.To read this article in full, please click here

How to adopt a disruptive network technology with minimal disruption

Disruptive network technologies are great—at least until they threaten to disrupt essential everyday network services and activities. That's when it's time to consider how innovations such as SDN, SD-WAN, intent-based networking (IBN) and network functions virtualization (NFV) can be transitioned into place without losing a beat.To read this article in full, please click here

How to move to a disruptive network technology with minimal disruption

Disruptive network technologies are great—at least until they threaten to disrupt essential everyday network services and activities. That's when it's time to consider how innovations such as SDN, SD-WAN, intent-based networking (IBN) and network functions virtualization (NFV) can be transitioned into place without losing a beat.To read this article in full, please click here

1 2 3 112