Five Reasons Data Center Liquid Cooling Is on the Rise

Traditionally reserved for mainframes and academic supercomputers, liquid cooling may soon be seeping into more enterprise data centers. New, more demanding enterprise workloads are pushing up power densities, leaving data center managers looking for more efficient alternatives to air-based cooling systems. We’ve asked a number of data center opearators and vendors about the applications that are driving liquid cooling into the mainstream. Some of them didn't want to disclose specific applications, saying they viewed those workloads and the way they’re cooled as a competitive advantage.  #Hyperscale #cloud operators, including @Microsoft, @Alphabet’s @Google, @Facebook, and @Baidu, have formed a group working on an open specification for #liquidcooled server racks without saying that exactly they would use them for. At least one category of workloads in the hyperscalers’ arsenal, however, clearly calls for liquid cooling: #machinelearning systems accelerated by #GPUs, or, in Google’s case, also #TPUs, which the company has said publicly are now cooled using a direct-to-chip liquid cooling design. Despite operators’ caginess around this subject, some usage trends are starting to emerge. If you're supporting any of the following workloads in your data center, liquid cooling may be in your future too:

http://www.datacenterknowledge.com/power-and-cooling/five-reasons-data-center-liquid-cooling-rise

Google, Facebook, Microsoft, and Twitter partner for ambitious new data project

Today, @Google, @Facebook, @Microsoft, and @Twitter joined to announce a new standards initiative called the Data Transfer Project, designed as a new way to move data between platforms. In a blog post, Google described the project as letting users “transfer data directly from one service to another, without needing to download and re-upload it.” The current version of the system supports data transfer for photos, mail, contacts, calendars, and tasks, drawing from publicly available APIs from Google, Microsoft, Twitter, Flickr, Instagram, Remember the Milk, and SmugMug. Many of those transfers could already be accomplished through other means, but participants hope the project will grow into a more robust and flexible alternative to conventional APIs. In its own blog post, Microsoft called for more companies to sign onto the effort, adding that “portability and interoperability are central to cloud innovation and competition.” “THE FUTURE OF PORTABILITY WILL NEED TO BE MORE INCLUSIVE, FLEXIBLE, AND OPEN.” The existing code for the project is available open-source on GitHub, along with a white paper describing its scope. Much of the codebase consists of “adapters” that can translate proprietary APIs into an interoperable transfer, making Instagram data workable for Flickr and vice versa. Between those adapters, engineers have also built a system to encrypt the data in transit, issuing forward-secret keys for each transaction. Notably, that system is focused on one-time transfers rather than the continuous interoperability enabled by many APIs. “The future of portability will need to be more inclusive, flexible, and open,” reads the white paper. “Our hope for this project is that it will enable a connection between any two public-facing product interfaces for importing and exporting data directly.” The bulk of the coding so far has been done by Google and Microsoft engineers who have long been tinkering with the idea of a more robust data transfer system. According to Greg Fair, product manager for Google Takeout, the idea arose from a frustration with the available options for managing data after it’s downloaded. Without a clear way to import that same data to a different service, tools like Takeout were only solving half the problem. “When people have data, they want to be able to move it from one product to another, and they can’t,” says Fair. “It’s a problem that we can’t really solve alone.” Most platforms already offer some kind of data-download tool, but those tools rarely connect with other services. Europe’s new GDPR legislation requires tools to provide all available data on a given user, which means it’s far more comprehensive than what you’d get from an API. Along with emails or photos, you’ll find thornier data like location history and facial recognition profiles that many users don’t even realize are being collected. There are a few projects trying to make use of that data — most notably Digi.me, which is building an entire app ecosystem around it — but for the most part, it ends up sitting on users’ hard drives. Download tools are presented as proof that users really do own their data, but owning your data and using it have turned into completely different things. “WE ALWAYS WANT TO THINK ABOUT USER DATA PROTECTION FIRST.” The project was envisioned as an open-source standard, and many of the engineers involved say a broader shift in governance will be necessary if the standard is successful. “In the long term, we want there to be a consortium of industry leaders, consumer groups, government groups,” says Fair. “But until we have a reasonable critical mass, it’s not an interesting conversation.” This is a delicate time for a data-sharing project. Facebook’s API was at the center of the Cambridge Analytica scandal, and the industry is still feeling out exactly how much users should be trusted with their own data. Google has struggled with its own API scandal, facing outcry over third-party email apps mishandling Gmail users’ data. In some ways, the proposed consortium would be a way to manage that risk, spreading the responsibility out among more groups.

https://www.theverge.com/2018/7/20/17589246/data-transfer-project-google-facebook-microsoft-twitter

IBM demands $167 million from Groupon for using its patents

@IBM and @Groupon have been engaged in a legal battle over patents since 2016, with the tech giant accusing the deals website of violating four of its e-commerce IPs without permission. Now, IBM is seeking $167 million from Groupon, telling jurors that other companies like Amazon and Facebook bought licenses to use its technologies for between $20 million to $50 million, while the deals marketplace refuses to do so. "The new kid on the block refuses to take responsibility for the technology it's using," IBM lawyer John Desmarais said in a statement. "IBM spends literally billions of dollars every year on research and development to make our lives easier." Two of the patents involved in the lawsuit are old intellectual properties that came out of the pre-internet Prodigy online service from 80's. The newest of the bunch is a "single sign on" technology allowing users to log into a website with their Facebook or Google accounts. Groupon lawyer David Hadden argued that the case isn't valid, because the patents don't cover the world wide web, because IBM did not invent it.

https://www.google.com/amp/s/www.engadget.com/amp/2018/07/17/ibm-groupon-patent-lawsuit/

Amazon May Be Gunning for Cisco

A report from The Information on Friday sent shares of networking hardware companies like @Cisco Systems (NASDAQ: #CSCO), @Juniper Networks, and @Arista Networks tumbling. @Amazon.com (NASDAQ: #AMZN), the e-commerce and cloud computing giant, is reportedly toying with the idea of selling cheap white-box networking switches to enterprise customers. A gateway to @AWS Amazon is already using white-box switches inside its data centers, according to The Information's sources. Other large internet companies have taken that path as well, including @Facebook and @Alphabet's @Google. Google has gone even further, designing its own custom chips for artificial intelligence. IMAGE SOURCE: GETTY IMAGES. But selling custom hardware to third-parties is a new frontier. Some customers are already reportedly testing Amazon's switches, with a launch planned within the next 18 months. Amazon's strategy will be focused on price, with plans to undercut the switches from Cisco by 70% to 80%. Amazon's switches will be unbranded and run open-source software. They'll have built-in connections to Amazon Web Services (AWS), enabling customers to more easily operate in a hybrid-cloud configuration, where on-premise hardware is mixed with public cloud computing. That could help make the process of moving to the cloud simpler and less painful for large enterprises. Given the reported pricing, Amazon probably won't make much money from selling the hardware itself. Instead, the switches could drive more business to AWS, which is by far the most profitable part of Amazon. Speaking to The Information, a Cisco spokeswoman downplayed the report: "Only Cisco offers an end-to-end intent-based networking portfolio that delivers assurance, industry-leading security, policy-based automation, and segmentation. We look forward to continue leading the industry on innovation and helping our customers simplify and manage their networks." Should Cisco be worried? One thing to remember: Cheap white-box networking hardware is not new. Cheap alternatives to Cisco's pricey hardware are also not new. If price were the only factor, Cisco wouldn't dominate the market for networking switches. But it does, with a market share of 53.4% in the first quarter, according to IDC. Other cloud computing companies have every incentive to work with networking hardware companies to fight back against the threat from Amazon. Cisco and Google are already collaborating, working on developing software that will connect data centers with Cisco hardware to Google's cloud services. Cisco is also a partner for Azure Stack, Microsoft's hybrid cloud solution. Amazon certainly shouldn't be underestimated, but selling enterprise hardware for use in on-premise data centers is a very different business than renting public cloud computing resources. Swapping out hardware from Cisco or other providers isn't trivial -- there are meaningful switching costs. And AWS certainly isn't the only game in town when it comes to hybrid cloud. Whether companies would want to lock themselves into AWS with Amazon's switching hardware is an open question. Amazon may ultimately abandon this effort, instead sticking with using the hardware in its own data centers. Even with low prices, winning market share will be a slog. Cisco has revamped its business model in recent years, moving to a subscription model with its latest Catalyst 9000 line of switches. Those products have proven popular, and the fact that they're bundled with subscription software may make it even more difficult for customers to move away from Cisco. Cisco has increasingly become a seller of networking solutions, and that makes its products and services even stickier. Cisco has been under attack from cheap commodity hardware for years. The company's dominance has so far held up. Just like Amazon shouldn't be underestimated, neither should Cisco.

https://www.fool.com/investing/2018/07/13/amazon-may-be-gunning-for-cisco.aspx

USING OPEN SOURCE DESIGNS TO CREATE MORE SPECIALIZED CHIPS

THE OPEN SOURCE movement changed how companies build software. @Facebook, @Twitter, and @Yahoo employees pitched in during the early days of the data-crunching software #Hadoop. Even after the relationship between @Apple and @Google soured, the companies' coders kept working together on an obscure but important piece of software called #LLVM. @Microsoft now uses and contributes to the @Linux operating system, even though it competes with Windows. The embrace of open source isn't about altruism. Facebook started using Hadoop because there was no commercial off-the-shelf software that met the company's needs as it grew. Because Hadoop is open source, Facebook could customize and extend it to solve its specific problems; sharing its changes allowed others to innovate further, making the software better for Facebook and all other users. Collaborating on freely available code enables companies and programmers to pool resources to solve common problems and avoid reinventing the wheel. Companies build competing products and services from these open source foundations that they might never have been able to build otherwise. But the open source revolution has been slow to come to the hardware world. A number of open source gadgets and circuit boards have been released in recent years, but while it's possible to run a laptop or server on nothing but open source software, the inner workings of our gadgets remain proprietary. An open source chip architecture called RISC-V could soon help change that. Chip maker Nvidia and storage company Western Digital have both announced plans to use RISC-V chips in their core products, and dozens of other companies have joined the RISC-V Foundation, including Google, Tesla, and chip giants like IBM, Samsung, and Qualcomm. RISC-V isn’t the first open source chip architecture, but it’s unusual for such a project to attract much attention outside of academia.
https://www.wired.com/story/using-open-source-designs-to-create-more-specialized-chips/

Port shipments of 25- and 50-Gbps network adapters expected to continue growth in 2018

According to the Dell'Oro Group, port shipments of next-generation 25-Gbps and 50-Gbps network adapters rose in the fourth quarter of 2017 as a result of strong demand from cloud service providers. Shipment growth is expected to continue in 2018, with 25- and 50-Gbps shipments more than doubling, the market research firm says in its newly released "Controller and Adapter Quarterly Report" on the fourth quarter of 2017. 10GBase-T port shipments reached 1 million in the fourth quarter of 2017, which Dell'Oro states is the highest level ever reached. 10GBase-T will make up a higher portion of 10-Gbps shipments over the next five years, the firm predicts. Shipments of 40-Gbps ports on enterprise servers only rose at a double-digit growth rate quarter-over-quarter. Led by shipments from Intel to a cloud service provider, adoption of 100-Gbps server connectivity is growing, Dell'Oro says in the report. "Shipment of 25/50-Gbps network adapters grew at a healthy double-digit rate to over 1 million from the prior quarter," said Barong Fung, senior analyst at Dell'Oro Group. "Three of the largest service providers – @Google, @Amazon, #Facebook – drove the market in the fourth quarter of 2017. We expect @Microsoft to begin their adoption of 50 Gbps in 2018. We also expect strong 25-Gbps adoption by the broader market of small cloud service providers and large enterprises in 2018." Dell'Oro's "Controller and Adapter Quarterly Report" provides tables that cover manufacturers' revenue, average selling prices, and unit and port shipments (by speed: 1 Gbps, 10 Gbps, 25 Gbps, 40 Gbps, 50 Gbps, and 100 Gbps) for Ethernet and Fibre Channel over Ethernet (FCoE) controllers and adapters. The report covers InfiniBand controllers and adapters as well.

http://www.lightwaveonline.com/articles/2018/04/port-shipments-of-25-and-50-gbps-network-adapters-expected-to-continue-growth-in-2018-dell-oro.html

Oculus Rift VR Gaming PC

I'm constantly looking for technical projects my son and I can enjoy working on together. As a father that spends most of his days time immersed in #datacenter technologies, and having a son that spends the majority of his free time playing video games, I thought it only fitting that with our forces combined we embark on the VR PC building journey. I spent time researching the differences between the @HTCVive and @Facebooks Oculus Rift #VR platforms. The choice for me to go with Oculus Rift was simple once reading about the Rift Core 2.0 update. Below is a list of hardware components purchased for my VR platform. I'd ask you to wish me luck but who needs luck when your prepared? 

“Luck Is What Happens When Preparation Meets  Opportunity". @Seneca

@AMD #Ryzen 7 1700
@MSI #X370 Gaming Pro Carbon
@Nvidia #GTX 1080Ti
@OculusRift
@Geil 16gb DDR4
@Nvidia 750 watt Gold PS
@Microsoft Windows 10 Home 64bit
@Samsung 850 #VNand EVO 500gb
@Samsung 960 #NVMe EVO M.2 250gb
@NZXT H440 Mid Tower

Have a great weekend. I know I will!

Bull of the Day: VMware (VMW)

Strong Q4 Results The Zacks #1 (Strong Buy) stock reported impressive fourth quarter fiscal 2018 results last month. Both the top and bottom line topped the Zacks Consensus Estimate. Revenues of $2.31 billion and non-GAAP earnings of $1.68 grew 13.6% and 17.5% year-over-year, respectively. In particular, top line growth was driven by robust performance from the company's #NSX and #vSAN product lines. Breaking sales down by region, U.S. revenues increased 10.9%, while International revenues jumped over 16%; EMEA delivered strong growth in Q4 as well. Taking a look at @VMware's different segments, Services revenues grew 8.4% to $1.24 billion and License revenues surged more than 20% to $1.06 billion. Services represent 53.7% of total revenues, and License make up 46.3% of total revenues. Bookings were looking solid too, with NSX license bookings increasing 24% ad vSAN bookings soaring 100% year-over-year. Guidance is Promising Looking ahead, fiscal 2019 revenues are projected to jump nearly 11%, with License revenues increasing about 11.3% year-over-year. Non-GAAP earnings for the year are expected to be $6.02 per share, and non-GAAP operating margin is anticipated to be 33.3%. Additionally, VMware projects cash flow from operations to grow about 11% to $3.55 billion. Capital expenditures are also expected to be $280 million, with free cash flow of approximately $3.27 billion. Earnings Outlook For VMW, its bottom line is trending upward for the foreseeable future. Earnings are expected to grow over 15% for the current quarter. Eight analysts have revised their estimates upwards in the last 60 days, and none have cut their outlook in the same time period. Fiscal 2018 figures are also looking pretty great, with 13 estimates moving higher in the past two months. The Zacks consensus estimate trend has jumped from $5.56 per share to $5.95 per share. Earnings estimates for fiscal 2020 are on the rise as well, jumping from $6.15 per share to $6.64 per share in the last 60 days. Can VMW Push Higher? Recently, it's been a wild ride for the tech industry, with major sell-offs happening on the Nasdaq and the "FAANG" stocks-Facebook (FB), Amazon (AMZN), Apple (AAPL), Netflix (NFLX), and Alphabet (GOOGL)-experiencing major losses. While it's hard to say exactly what the cause of the tech sell-off is, overall data privacy concerns is likely the culprit, especially in the wake of the Facebook-Cambridge Analytica scandal. Looking at VMware, shares of the cloud infrastructure company have slipped a little over 3% since the start of the year, but have gained roughly 30% in the past one-year period. The Nasdaq has risen 18.2% in comparison

https://m.nasdaq.com/article/bull-of-the-day-vmware-vmw-cm942277

Cisco commits $50 million to end homelessness in Silicon Valley

Homelessness in Santa Clara County has gotten worse, with the overall homeless population increasing 13 percent to 7,394 in 2017 over the course of two years. That puts Santa Clara’s homelessness crisis in the same ballpark as San Francisco’s, which has a homeless population of 7,499, according to a 2017 homeless census and survey. Santa Clara also has the third highest rate of chronic homelessness in the entire country. Today, @Cisco announced a $50 million donation to Destination: Home over the next five years. The idea is to help put an end to homelessness in Santa Clara County — an area of Northern California that is home to the tech industry’s Silicon Valley. This area consists of cities like Cupertino (home to @Apple ’s headquarters), Mountain View (home to @Google/Alphabet), Palo Alto (home to @Facebook), San Jose and Sunnyvale. “We have said for a long time that it is up to all of us to end homelessness in our community,” Destination: Home CEO Jennifer Loving said in a statement. “Cisco has fully embraced that concept, and is stepping up in a big way to provide the type of critical private sector leadership and substantial funding that is necessary to address this crisis head on. We couldn’t be more thrilled or grateful to have Chuck Robbins and the Cisco team at the table.” Cisco has donated an initial $20 million chunk to Destination: Home through its Cisco Fund. The plan is for this money to invigorate Destination: Home’s efforts to achieve its five-year plan to end homelessness, which entails disrupting and transforming homeless response systems, building new housing opportunities and deploying client-centered solutions. Since implementing the plan in 2015, Santa Clara County has been able to permanently house 5,154 people, according to Destination: Home’s March 2018 progress report.

https://www.google.com/amp/s/techcrunch.com/2018/03/26/cisco-commits-50-million-to-address-homelessness-in-silicon-valley/amp/#ampshare=https://techcrunch.com/2018/03/26/cisco-commits-50-million-to-address-homelessness-in-silicon-valley/

Microsoft bullish on Congress' inclusion of CLOUD Act in funding bill

@Microsoft has been arguing for a while that data privacy is something that needs addressing by Congress, not the courts. It looks like Redmond is one step closer to getting its wish. The #CLOUD (Clarifying Lawful Overseas Use of Data) Act was added to an omnibus funding bill on March 21. The CLOUD Act creates a legal framework regulating how law enforcement can access data across borders. It has received backing by members of both parties in Congress, the Department of Justice, and a number of tech companies, including Microsoft, Apple, Google and Facebook. Late last month, the U.S. Supreme Court heard arguments over Microsoft's ongoing battle with the U.S. government over data privacy. According to reports from those attending, the justices didn't seem swayed by Microsoft's claims that data stored overseas should not be accessible to government prosecutors. A ruling in the case isn't expected until the end of June 2018, but may not happen at all if the CLOUD Act renders the issue moot. Microsoft President and Chief Legal Officer Brad Smith was predictably upbeat about Congress taking on the data-privacy issue, as indicated by his blog post about it. But not everyone is onboard. The Electronic Frontier Foundation has opined that the CLOUD Act is a "dangerous expansion of police snooping on cross-border data." The ACLU has said the Cloud ACT "threatens activists abroad, individuals here in the U.S." Microsoft's data-privacy case began in 2013 over emails from a drug trafficking investigation suspect stored in Microsoft servers in Dublin. A federal judge in New York issued a warrant for the emails, and Microsoft decided to challenge the order in court.

http://www.zdnet.com/article/microsoft-bullish-on-congress-inclusion-of-cloud-act-in-funding-bill/

Open Networking Foundation’s Stratum project aims to make generic gear more programmable

The #OpenNetworking Foundation has launched a new #opensource project that aims to create a reference platform for generic networking gear known as “ #whiteboxes” to make them more easily programmable. The project, called @Stratum, will also support a new set of #softwaredefinednetworking interfaces created by the group, which are designed to make the data plane more programmable. Rising adoption of #SDN and also network functions virtualization or #NFV has provided a healthy boost to the success of white-box manufacturers in recent years. By separating the control plane, network operating system and networking tasks from the underlying hardware, organizations have been able to run software on unbranded switches and servers that cost significantly less than branded counterparts from companies such as @Cisco Systems Inc. and @Juniper Networks Inc. The shift has been especially beneficial for the biggest “ #hyperscale” cloud data center operators such as @Google Inc. and @Facebook Inc. Those have been able to push ever-increasing amounts of traffic through their infrastructures while keeping costs under control even as demand increases by using cheaper white boxes from original design manufacturers. Now, the ONF wants to increase the adoption of SDN, NFV and white boxes even more by making unbranded gear easier to use. That’s what the new Stratum project is all about. “The network should be as easy to program as a server,” Nick McKeown, a professor at Stanford University and cofounder of ONF, said in a statement. “By lifting protocols out of hardware and into software, we will unleash a Cambrian explosion of innovation in networking, just as we have seen in compute.” More specifically, the Stratum project aims to create an open-source reference platform that organizations can use to program their white box gear more easily. Stratum is designed to be used with both existing and upcoming hardware, the ONF said.

https://siliconangle.com/blog/2018/03/15/open-networking-foundations-stratum-project-aims-make-white-boxes-programmable/

NVM Express, Inc. Elects Cavium, Cisco, Dell, Facebook, Intel, Microsemi and NetApp to Promoter Board for 2018

@NVM Express, Inc. Elects @Cavium, @Cisco, @Dell, @Facebook, @Intel, @Microsemi and @NetApp to Promoter Board for 2018 Promoters range from OEMs to Enterprise Data Suppliers February 01, 2018 05:21 PM Eastern Standard Time BEAVERTON, Ore.--(BUSINESS WIRE)--@NVMExpress, Inc. elected @Cavium, @Cisco, @Dell, @Facebook, @Intel, @Microsemi and @NetApp to join their Board of Directors, the NVM Express Promoters Group. Existing Promoter members include @Micron, @Microsoft, @Samsung, @Seagate, @Toshiba and @Western Digital. NVM Express elects Cavium, Cisco, Dell, Facebook, Intel, Microsemi and NetApp to 2018 Promoter Board. Tweet this The NVM Express Inc., offering an open collection of NVM Express (NVMe™) specifications and resources on the benefits of non-volatile memory, holds an election for a group of the NVM Express Promoters Group representatives at the end of each year. The elected representatives occupy a two-year term and are eligible for re-election. All NVM Express, Inc. Contributor and Promoter member companies may run for Promoter positions. “Storage architectures are undergoing a sea change with NVM Express playing a pivotal role,” Craig W. Carlson, Senior Technologist, said. “Cavium is a leader in NVMe and has been actively working with the industry on NVMe over Fabric technologies for both Ethernet and Fibre Channel for some time, and with our participation on the Board of Directors we look forward to helping steward the broader NVM Express mission going forward.” The NVM Express Promoter Group is responsible for the operations and financial management of the organization, as well as being the stewards of the specification and standard. To learn more about the NVM Express Promoter Group election process and responsibilities, check out the NVMe Bylaws. About NVM Express, Inc. With more than 100 members, NVM Express, Inc. is a non-profit organization focused on enabling broad ecosystem adoption of high performance and low latency non-volatile memory (NVM) storage through a standards-based approach. The organization offers an open collection of NVM Express (NVMe™) specifications and information to fully expose the benefits of non-volatile memory in all types of computing environments from mobile to data center. NVMe-based specifications are designed from the ground up to deliver high bandwidth and low latency storage access for current and future NVM technologies. For more information, visit http://www.nvmexpress.org. Follow NVM Express on Twitter and LinkedIn Supporting Quotes for NVMe Express, Inc. 2018 Board of Directors Dell EMC “Dell has been working on NVMe since its inception and is one of the leading suppliers of NVMe storage for both enterprise servers and client applications. We believe that NVMe presents a unique opportunity to radically redefine the capabilities of client and server based storage subsystems. As a member of the Board of Directors, we want to work to insure that NVMe has the features and capabilities needed to enable high performance next generation storage subsystems.” — William Lynn, Distinguished Engineer Microsemi Corp., Performance Storage Business Unit “NVMe is the only interface standard that meets the needs of current (Flash) as well as next generation NVM and Microsemi is committed to its success. Microsemi Corp. is a leader in NVMe controllers and custom NVMe board level solution products. Microsemi is also a PCIe switch supplier and is working to enhance PCIe switching to meet the needs of NVMe storage systems.” — Peter Onufryk, Fellow NetApp Inc. “NVMe provides a technology that can be used in a wide variety of environments from laptops to Enterprise data centers. We believe the Promoters Group should contain a balance of representatives from device vendors, host users of storage, and Enterprise data center suppliers. We will bring input that is representative of all those environments to help NVMe become the ubiquitous storage protocol of the future.” — Frederick Knight, Principal Standards Technologist

https://www.businesswire.com/news/home/20180201006644/en/NVM-Express-Elects-Cavium-Cisco-Dell-Facebook

Microsoft Says AI Advances Will Require New Laws, Regulations

The rapidly advancing area of #artificialintelligence will require a new field of law and new regulations governing a growing pool of businesses involved, according to @Microsoft Corp., a 25-year participant in #AI research. Companies making and selling AI software will need to be held responsible for potential harm caused by “unreasonable practices” – if a self-driving car program is set up in an unsafe manner that causes injury or death, for example, Microsoft said. And as AI and automation boost the number of laborers in the gig-economy or on-demand jobs, Microsoft said technology companies need to take responsibility and advocate for protections and benefits for workers, rather than passing the buck by claiming to be ``just the technology platform’’ enabling all this change.  More from Bloomberg.com: Bitcoin Storms Back From Dip Below $10,000 in White-Knuckle Ride Microsoft broaches these ideas in a 149-page book entitled “The Future Computed,” which will also be the subject of a panel at the World Economic Forum in Davos, Switzerland, next week. As Redmond, Washington-based Microsoft seeks to be a leader in AI and automating work tasks, it’s also trying to get out in front of the challenges expected to arise from promising new technologies, such as job losses and everyday citizens who may be hurt or disadvantaged by malfunctioning or biased algorithms. “We are trying to be clear-eyed in talking about the challenges,” said Microsoft President and Chief Legal Officer Brad Smith, who will sit on the Davos panel and co-wrote the introduction to the book. More from Bloomberg.com: Did Bitcoin Just Burst? How It Compares to History's Big Bubbles Microsoft is working on some of these areas through groups such as the Partnership on AI, which includes rivals like Amazon.com Inc., Alphabet Inc.’s Google, Apple Inc. and Facebook Inc. Still, the call for more regulation in an emerging area like AI is unusual for technology companies, said Ryan Calo, a professor at the University of Washington School of Law, who has read the book.

https://finance.yahoo.com/news/microsoft-says-ai-advances-require-030010142.html

Hyperscale data centers reached over 390 worldwide in 2017

Menu #Hyperscale #datacenters reached over 390 worldwide in 2017 Posted yesterday by Ron Miller (@ron_miller) #Hyperscale operators are defined as enormous companies like @Amazon, @Apple, @Facebook and @Google that need to provide computing on a massive scale. You would think that there would be a limited number of this type of highly specialized data center, but recent research from Synergy Research found that 2017 was actually a breakout year for new hyperscale data centers across the world — with no sign of slowing down in 2018. Synergy reported that the year closed with over 390 web-scale data centers worldwide, up from 300 in just one year.  Google was particularly active. Chinese companies Tencent and Baidu also built hyperscale data centers this year. Still, the vast majority are in the US with 44 percent of the total. China is second way back with 8 percent, followed by Japan and the UK with 6 percent each and Australia and Germany with 5 percent each.   Synergy reports that on average, the 24 hyperscale firms have 16 data centers each. The companies with the largest number won’t come as a surprise to anyone with Amazon/AWS, Microsoft, IBM and Google each having at least 45 worldwide. The definition of hyperscale varies, but IDC says it requires at least 5000 servers and 10,000 square feet of available space, but is often much larger. Synergy defines it having “several hundreds of thousands of servers — or sometimes millions.” These operators often build their own equipment to deal with the specific needs of their immense computing requirements. By designing their own hardware and software, these companies can control every aspect of the computing experience to squeeze out the maximum amount of efficiency, which is crucial when you are dealing with the massive scale of these organizations.

https://techcrunch.com/2017/12/30/hyperscale-data-centers-reached-over-390-worldwide-in-2017/

Amazon Web Services backs deep-learning format introduced by Microsoft and Facebook

The #OpenNeuralNetworkExchange ( #ONNX) #deep-earning format, introduced in September by @Microsoft and @Facebook, has a new backer following @Amazon Web Services’ decision to embrace the framework with a new open-source project. #AWS released ONNX-MXNet Thursday afternoon, which sounds like a telecom standard from the 1990s but is actually a method for allowing deep learning models built around the ONNX format to run on the @Apache MXNet framework. In supporting ONNX, envisioned as a standard way to build deep learning models, the cloud computing leader gives a significant stamp of approval to the concept of open data models. Advanced research into deep learning is a relatively new field, even though we’ve been talking about it for several years. There aren’t a lot of go-to methods for building data models, and it can be difficult to switch frameworks upon discovering that your model might benefit more from a different tactic. That was the idea behind ONNX, a bridge between open-source deep-learning frameworks like Caffe2, developed at Facebook, and Cognitive Toolkit, developed by Microsoft. It also supports the open-source PyTorch framework, and now Apache MXNet thanks to the work of AWS. These all seem like positive developments for those at the forefront of artificial intelligence research, when major players in the field agree to forge a common path that makes it easier for people to use deep-learning techniques in their own applications or businesses. All eyes now turn to Google, which has made AI and deep learning a huge part of its cloud strategy but has yet to join forces with the ONNX group
https://www.geekwire.com/2017/amazon-web-services-backs-deep-learning-format-introduced-microsoft-facebook/

Why HPE No Longer Wants to Sell Cheap Servers to Cloud Giants

The recent decision by @Hewlett Packard Enterprise to stop selling commodity servers to companies that operate the world’s biggest cloud platforms and focus its efforts on selling hardware and services to smaller cloud companies, other types of service providers, and enterprises is a smart strategy, industry analysts say. With commodity servers, “there is no money to be made in the Super 7 cloud providers, [but] the tier 2 market is the fastest growing [portion], percentwise, of the cloud market,” said Patrick Moorhead, president and principal analyst of Moor Insights & Strategy. Companies that make up the “Super 7” (an Intel expression) are Alphabet, Microsoft, Facebook, Amazon, Apple, Alibaba, and Tencent. Related: Taiwanese Firms to Sell Latest NVIDIA AI Hardware to Cloud Giants HPE this month announced it is exiting the business of selling commodity Cloudline servers to these companies, also often referred to as tier 1 service providers. But it will continue selling its higher-margin products, including high-end Apollo servers, 3PAR storage systems, and networking equipment to these large providers, a company spokesperson said. The company will continue selling a standardized set of Cloudline servers to smaller cloud companies and service providers, such as Salesforce and eBay. “We are shifting our resources and doubling down on tier 2 and tier 3 providers,” the spokesperson said. Related: HPE Says Storage, Networking Gear Sales Rising Industry analysts say HPE’s decision to abandon the hyper-scale commodity server business is a sound strategy, because the vendor is no longer getting a return on investment from its partnership with original design manufacturer (ODM) Foxconn, which built the custom commodity servers for HPE. HPE and Dell used to sell a lot of servers to the tier 1 providers, Moorhead said. But in recent years, their sales have plummeted, as cloud providers began designing their own servers and buying them directly from Asian ODMs – the likes of Foxconn. HPE partnered with Foxconn in 2014 to get a piece of the action, analysts say. But earlier this year, HPE chief executive Meg Whitman said her company faced significantly lower demand for the commodity servers from one tier 1 customer, which was reportedly Microsoft. The heavy competition from Asian ODMs has forced traditional server vendors, such as HPE and Dell EMC to decide how best to serve these customers or walk away entirely from the low-margin business, said Jeffrey Fidacaro, senior analyst of data center technologies at 451 Research. While HPE is pulling out of the hyper-scale custom commodity server market, Dell has no plans to do so anytime soon, said Ravi Pendekanti, senior VP of server solutions, product management and product marketing, at Dell EMC. “We continue to ship and keep our customers happy,” he said, responding to a question about Dell’s plans in selling hardware to the biggest cloud companies. Aiming at Smaller Providers The smaller providers don’t have the staff to design their own hardware and service it in their data centers like the hyper-scalers do, so they need the hardware and services OEMs like HPE provide, Fidacaro said. The same goes for enterprises. “For second- and third-tier service providers, colocation companies that aren’t very large, and enterprises, the services ecosystem that includes design, procurement, testing, and validation is still important,” Fidacaro said. “It’s something they can’t replicate internally because they don’t have the resources. This is where HPE can offer a lot of value.” Ashish Nadkarni, IDC’s program director of computing platforms, said HPE’s move resolves the awkwardness of having Foxconn build its custom commodity servers while competing with the ODM for the same business. The move is also part of HPE’s continuing effort to focus on higher-margin businesses, he said. Last year, HPE spun off and merged its non-core software with Micro Focus in a deal worth $8.8 billion. “It cleans up their business model,” he said. “HPE is shedding all the low-margin products and becoming more streamlined and nimble and only focusing on areas where it can make money.” For example, the company in 2017 has beefed up its product portfolio and now offers customers hyper-converged infrastructure, technology it gained through its SimpliVity acquisition in January, and flash storage hardware gained through its Nimble Storage acquisition in March. HPE can bolster revenue through its broad portfolio of products for smaller providers and enterprises, Nadkarni said. “There is money to be made with the whole solution.” HPE is a member of the Open Compute Project, the organization through which Facebook, Google, Microsoft, and others open source custom specs and have suppliers compete among each other for large orders of data center hardware. An HPE spokesman didn’t immediately respond when asked whether the company plans to stay in the organization. Moorhead, however, still sees value in HPE staying put. “I still see the value of HPE’s continued participation in OCP, as it can drive specifications for Cloudline, [which is] very much still an entity in tier 2 providers,” Moorhead said.

http://www.datacenterknowledge.com/hardware/why-hpe-no-longer-wants-sell-cheap-servers-cloud-giants

Memory Chips Send Samsung Profit Soaring

@Samsung Electronics Co. Reported another record high in quarterly earnings Tuesday as its semiconductor division posted its highest profit ever and tipped another record-breaking quarter is on the horizon, a breathtaking run for a company that is fighting to get its leader out of jail. The company also announced a monstrous capital investment plan to increase chip production and higher shareholder dividends as the world's largest chipmaker basks in an unprecedented boom in the memory chip industry that helped it overcome a crisis in its leadership. The South Korean company's July-September net income surged 150 percent to 11 trillion won ($9.8 billion), compared with 4.4 trillion won a year earlier. The result beats analysts' consensus of 10.9 trillion won, according to financial data provider FactSet. Sales rose 30 percent to 62 trillion won ($55.1 billion) while operating profit nearly tripled from the previous year to 14.5 trillion won ($12.9 billion), in line with its preview. Samsung, the world's largest maker of smartphones, has continued its record-breaking earnings streak thanks to a global boom in memory chips that resulted from a jump in mobile data traffic and rise of cloud computing, even as its de facto chief was sentenced to five years in prison on a conviction for bribery. Hearings are underway at the appeals court as Lee Jae-yong, Samsung Electronics vice chairman and a grandson of Samsung's founder, contests the conviction for offering bribe to then-President Park Geun-hye and her close friend to ensure a smooth transition of corporate leadership at Samsung. Investors are also waiting for Samsung to name a successor to Kwon Oh-hyun, one of Samsung's three chief executives who are not part of the founding family but climbed the ladder from rank-and-file to overseeing daily management of Asia's largest consumer electronics company. Kwon led day-to-day operations of the world's most successful chip business and was a face of Samsung during Lee's absence. In a surprise announcement earlier this month, he offered to step down when his term ends early next year. While two other chief executives oversee Samsung's phone business and appliance business, leaving little leadership vacuum for Samsung's daily operations, there are worries in and outside Samsung that its long-term business planning is put on hold due to Lee's imprisonment. For now, Samsung is reaping the fruits of its massive investment in the past decade during the days of Lee's father when the cyclical memory chip industry was vulnerable to the global demand for consumer electronics products. As global tech giants such as Amazon, Google, Microsoft, Facebook and Apple expand their data centers and invest in connected cars, artificial intelligence and cloud computing, Samsung's semiconductor division appears invincible. The company's semiconductor division generated 10 trillion won ($8.9 billion) in operating income during the three-month period, accounting for about 70 percent of its entire operating income. The prices of DRAM and NAND memory chips soared as demand outpaced tight supplies. To respond to a surge in demand for memory chips that power data centers and mobile devices, Samsung is spending 46.2 trillion won ($41.1 billion) in capital expenditure this year, including 29.5 trillion won to increase output of chips. That is nearly twice as large as Samsung's last year's capital spending of 25.5 trillion won. With enormous profit growth, Samsung is also earmarking 9.6 trillion won ($8.6 billion) each year from 2018 to 2020 for shareholder returns, nearly double its dividends in 2017. Samsung, the world's largest smartphone maker, generated 3.3 trillion won from its mobile division, lower than the previous quarter but a jump from a year earlier when its fiery Galaxy Note 7 recalls wiped out nearly all of its profit. Looking ahead, Samsung's component business, especially its memory chips and advanced flexible displays called OLED, will drive the improvement of its financial performance during the fourth quarter and next year, Samsung said. The memory chip business will continue to see growth not just from the development of technologies in big data and machine learning, but also from growing sales of mobile devices with dual camera and 3D sensors and artificial intelligence capabilities, it said.

https://www.toptechnews.com/article/index.php?story_id=1330045SZZ4M

HPE To Step Back From 'Tier 1 Service Provider' Server Business While Focusing On Higher Margin Growth Areas

@Hewlett Packard Enterprise is planning to end production of what it calls customizable commodity servers targeting the largest #hyperscale #cloud providers as part of a move to focus on more profitable growth areas. #HPE President @Antonio Neri outlined several initiatives the company has to sharpen its focus on growth during last week's HPE securities analyst meeting. As part of his presentation, which CRN reported on last week, Neri said that HPE will discontinue the sales of custom-designed, commodity servers while continuing to sell higher-margin products. [Related: 10 Server And Storage Technologies Taking Center Stage At HPE Discover 2017] An HPE spokesperson confirmed that the company will end that product line in fiscal year 2018. "In FY18, HPE will cease selling custom-designed, commodity servers to #Tier1 #ServiceProviders, which we define as @Amazon, @Google, @Microsoft, @Facebook, @Apple, @TenCent, @Alibaba and @Baidu.  We will continue selling our higher margin products like storage, networking and higher value servers to these companies," the spokesperson wrote. [Sponsored Suggested Post: Free IoT Virtual Conference Learn how to monetize IoT solutions. Discover new IoT technologies. Build strategic partnerships.] Neri said HPE will stop selling custom-designed commodity servers to the Tier 1 service provider segment, including providers like AWS, Microsoft Azure, and Google, but will continue selling them higher-margin products. "These commodity server deals come at very low margins, and there is no services pull-through opportunity for the company. So, we are modifying our strategy to reinvest these resources towards solutions and services that will drive profitable growth," he said. HPE will focus on gaining share with standardized compute platform for Tier 2 and Tier 3 service providers, including companies such as Dropbox, eBay, and Salesforce.com, who have built their businesses in the cloud, Neri said. "These SKUs will be delivered without a custom-design (including software and services), providing value to our customers while ensuring profitability for our business," he said. "HPE offers channel partners many server options including its ProLiant DL line, its blade servers, and its Synergy composable infrastructure, and so a move to exit the Tier 1 cloud server business would probably have little direct impact on server sales through the channel," said Dan Molina, chief technology officer at Nth Generation Computing, a San Diego-based solution provider and long-time HPE channel partner. However, Molina told CRN, the move will benefit the channel in other ways. "This will free up HPE resources to support its other server lines," he said. "It can be hard for a company to keep up with so many server lines. HPE can now better focus on its roadmap for its DL and Synergy servers, or its SimpliVity hyper-converged infrastructure."

http://www.crn.com/news/data-center/300094536/hpe-to-step-back-from-tier-1-service-provider-server-business-while-focusing-on-higher-margin-growth-areas.htm

If data is the new oil, is storage the new refinery?

Just as oil propelled @StandardOilCo. Inc. to a position of dominant #industrialpower in the late 1800s, #data is doing the same for a number of #technology firms today. Half of consumer online spending in the U.S. is controlled by @Amazon, a company that relies extensively on mining data so it knows what you want before your buy it. Nearly all revenue for digital advertising in America flows to two companies: @Facebook and @Google. The list goes on, but one constant is clear. Just as oil spawned the growth of many industries, data is reshaping the technology stack. From #robotics (think #autonomouscars) to the entire field of #datascience, a new era of innovation is underway. This is forcing any enterprise that handles data to innovate as well, and there may be no better example of this transformation than what is taking place in the storage layer, where the application of machine learning and artificial intelligence is refining data management in new ways. “Pick your domain; there’s going to be innovation across a lot of fronts, across a lot of traditional vertical industries that is all going to be about data and driven by data. And there has to be an infrastructure layer that actually enables it,” said Mike Cordano (pictured), president and chief operating officer at Western Digital Corp. What that infrastructure will look like over the next two to three years was the focus of discussion earlier this month, as Western Digital unveiled new storage technology that is designed to handle the explosion of data across the enterprise stack. Cordano stopped by theCUBE, SiliconANGLE’s mobile livestreaming studio, and spoke with host Jeff Frick (@JeffFrick) at the Western Digital Presents: Innovating to Fuel the Next Decade of Big Data event in San Jose, California. Frick also separately interviewed Janet George, fellow and chief data officer at Western Digital, and Brendan Collins, vice president of product marketing devices, at Western Digital. They discussed key elements of the company’s new storage technology, enterprise trends that are shaping data management tools, the important role of AI and machine learning in bringing compute to data and potential innovations in the future.

https://siliconangle.com/blog/2017/10/18/if-data-is-the-new-oil-is-storage-the-new-refinery-letdatathrive/

3 Reasons Why The 'FANG' Phenomenon Will End Badly

Summary The 1999-2000 tech balloon was not an online-only phenomenon. The "Four Horsemen" that controlled more than half of the market capitalization for the ill-fated @Nasdaq - @Microsoft, @Intel, @Cisco and @Dell - were exceptionally profitable. The notion that @Facebook, @Amazon, @Netflix and @Google/ @Alphabet represent something entirely different in terms of profitability is flawed. The ratio of American households’ net worth to disposable income in 2000 hit 620%, due in large part to tech stock euphoria. Now we have "FANG" stocks emulating the "Four Horsemen" and real estate prices in key parts of the country stretching affordability. Should investors really be dismissive of the ratio in 2017? Those who do not wish to draw any parallels between today's stock market and the 1999-2000 tech stock bubble typically claim that "all of those turn-of-the-century dot-coms weren't making any money. Today's 2017 superstars - Facebook (NASDAQ:FB), Amazon (NASDAQ:AMZN), Netflix (NASDAQ:NFLX) and Google/Alphabet (GOOG, GOOGL) - make money hand over fist!" The problem with this rationalization is threefold. First, the 1999-2000 tech balloon was not an online-only phenomenon. The "Four Horsemen" that controlled more than half of the market capitalization for the ill-fated Nasdaq - Microsoft (NASDAQ:MSFT), Intel (NASDAQ:INTC), Cisco (NASDAQ:CSCO) and Dell (NYSE:DVMT) - were exceptionally profitable. They also were extremely overvalued at a combined price-to-earnings ratio (P/E) of 60. It follows that the notion that Facebook, Amazon, Netflix and Google/Alphabet represent something entirely different in terms of profitability is flawed, especially when one considers an average P/E of 130 for "FANG." Second, the ratio of American households' net worth to disposable income in 2000 hit 620%, due primarily to an 18-year old secular bull market and tech stock euphoria. It was unsustainable, however, largely because the ratio had rarely deviated more than 1 standard deviation from its mean and because asset prices had frequently correlated with after-tax wage growth. Similarly, the ratio hit 650% on residential housing jubilation prior to the financial collapse in 2008. That too was unsustainable. Now we have "FANG" stocks emulating the "Four Horsemen," the median stock bubbling over with froth, and real estate prices in key parts of the country stretching affordability. Should investors really be dismissive of the ratio in 2017?

https://seekingalpha.com/article/4114497-3-reasons-fang-phenomenon-will-end-badly