Showing posts with label Commentary. Show all posts
Showing posts with label Commentary. Show all posts

Friday, June 1, 2018

A look at COBO and the drive for on-board optics

The Consortium for On-Board Optics (COBO) was formed three years ago to develop specifications to permit the use of board-mounted optical modules in the manufacturing of networking equipment, such as switches and even servers. The specifications will cover electrical interfaces, pin-outs, connectors, thermals, etc.  The idea is to drive the development of interchangeable and interoperable optical modules that can be mounted onto motherboards and daughtercards. Microsoft is a founding member of COBO, as are Arista, Broadcom, Cisco, Juniper, and Mellanox. In April, COBO issued its first on-board optics specification targeting 400 Gbps and 800 Gbps data rates and leveraging two electrical interfaces: 8 and 16 lanes of 50-gigabit PAM-4 signals.

I caught up with Brad Booth, president of COBO, at the recent NetEvents 2018 in San Jose, California. In his day job, Brad is also a Network Architect with Microsoft Azure.

Jim Carroll, Optical Networks Daily – Hi Brad, good to see you again. Tell us about your work and the latest developments in the drive toward on-board optics.

Brad Booth, COBO:  I am the president of COBO. That is my volunteer job. My full-time job is a Network Architect for Microsoft Azure. 

A few months ago, COBO, the Consortium for Onboard Optics released our specification 1.0 specification in time for OFC and now we're going around letting people know about it because it is a game changer for how to use embedded optics. Traditionally, embedded optics have been proprietary solutions. We knew that as we progress the technology and move from 10G to 40G to 100G to 400G, and whatever speed comes next, the ability to continue to use faceplate optics is presenting greater complexities. 

First, the thermal environment was getting a little harder to deal with. Signal integrity was becoming more difficult to deal with. The fact that you had people plugging these modules in, which were very sensitive to ASD and sensitive to noise, resulting in damage. And we decided that we needed to start the progression of moving the optics away from the faceplate and closer to the ASIC, if not on to the same footprint of the ASIC eventually. We knew that at 400G it was possible to do faceplate pluggable. We knew in time people would figure out how to do an optics embedded with ASICs, but what was going to be the thing that would allow us that transition in between. What would allow us to have the learnings and the understanding of how this will change our business models. That’s why we came up with COBO.


What is COBO implementation actually going to look like?

Brad Booth, COBO: This is a COBO module. It is a by 16 wide, which means it is a 16-lane interface on one side. This module was designed for the output of four 100G PSM4 signals  This is the medium size version of the module. It comes in a by-8 and a by-16 version. It comes in three lengths: slightly shorter than this, this length, and slightly longer. That allows us to be able to use this module literally from multimode fiber all the way up to long-haul coherent transmission. 
The module at its widest form factor is capable of accepting up to 40 watts of power to drive whatever component tree we need in it. 

It's capable of having a full 1RU heatsink, or greater if we want. You can even do liquid cooling on this if you want. There's the capability of latching either on the high-speed pins or the low-speed pin connector on the bottom. We separated the low-speed power and ground and control away from the high speed to improve the signal integrity. The nice thing is these pins on the bottom are .8mm pitch, which means that we get lots of good contact and power. There is also .6mm pitch, which should be able to handle up to 100 Gbps PAM4 electrical signaling. This means that future iterations of this module will be capable of 1.6 Tbps worth of bandwidth.


What kind of support or feedback are you getting from the industry?

Brad Booth, COBO:  We have over 60 member companies participating in COBO now. Many of them are very well known in the industry. Cisco, Arista, Juniper. Mellanox. We have connector companies involved too., like SEMTECH, TE Connectivity, Luxtera. We have companies that specialize in optical connectors, like SENKO, Huber+Suhner, etc. We have a strong ecosystem. 

We have to figure out how to move those optics from the faceplate further inside the systems to improve the signal integrity, and we have to be able to do it without drawing massive amount of power. And that's one of the advantages of having so many people in the ecosystem that play in this area, we have a lot of good contributions.


Filmed at NetEvents 2018 in San Jose, California

Is COBO targeting 400G and above? Are higher rates going to be supported? 

Brad Booth, COBO: This module is actually a four by 100. So it has 400 gigs worth of bandwidth in it right now. If you were to build it to handle 400G Ethernet, you could put two 400G Ethernets in it, and so that would be 800G worth of bandwidth. When this electrical interface goes to 100G PAM4 in the next version of our specification, this module will be able to do 1.6 terabytes. 

Is that the endpoint? Probably not, because we've actually designed the module to be able to handle the coherent light, or the 400 ZR, which is being worked on by the OIF. That's a 15 watt part and it does 400G per lambda. If the power of that technology continues to drop, you could potentially put more of those in this and actually achieve even greater bandwidth. 

What about the manufacturability of COBO? How is the industry progressing in this regard? 

Brad Booth, COBO: The interesting aspect of building something like this is being able to dissipate the power and the heat. The biggest thing for heat dissipation is actually the maximum height that you can make the heat fin on the heat sink. So, if you get a full 1RU height, you get significantly better thermal performance than just a small one. We've done some general evaluations here. We’ve actually done some thermal modeling. As a next step, we're looking forward to building out true thermal systems and testing it. 

One of the key aspects of this is that you've got a full 1 RU height. Second, you're not blocking the faceplate with optics, and you don't need as much airflow to pull through the system. That means you don't need fans running full out just to be able to pull enough air through to cool it. Another aspect is that COBO will allow you can to change how you implement your system design. You could place these modules in different spots across the motherboard that are potentially cooler or further away from the hotter ASIC. You could implement airflow channels to cool the optics, separate from the actual switch ASIC. 

There is the possibility of using liquid cooling. This module doesn't have a heat sink on it right now, so you could drop a liquid cooling plate on top of it. There's also the expansion capability provided by extra pins on the bottom of the module, which could be used to provide more power and more control signals if necessary for some next-generation technologies. 

What about the challenge of fiber alignment with COBO?

Brad Booth, COBO: Regarding fiber alignments, an interesting aspect is that we expect most people in the first generation to build these modules with pigtail optics. In other words, the optics will already be pre-attached. This module, which is just a mock-up, is shown without them. 

We have work going on to actually do a connectorized interface. We have some people working on how that would actually be implemented. If we can connectorize it here and connectorized it on the faceplate, that means that you wouldn't have to worry about it at manufacturing. You could attach the module in the factory when they build the switch. They could put the optics right in. Then, after the optics and systems are all installed, you could come and hand fiber, trace it out to the pigtails or to the front faceplate and that would actually help eliminate the worry, if you have to replace this, that you have to unstrand the whole pigtail. This provides the option for people to be able to do both. We’ve discussed even short pigtail versions within our optical conductivity study group. 

We invite people to contribute their thoughts and opinions. Come and work with us on this because this is an area that we think is going to become critically important as we start progressing towards actually putting optics embedded with ASICs. 

http://onboardoptics.org/

Tuesday, September 12, 2017

VMware looks to the clouds – part 2

The first part of this article covered the launch of VMware Cloud on AWS, which as a reminder, is the new on-demand service being launched by VMware that runs of bare-metal AWS infrastructure, initially in the AWS US West (Oregon) region.  The service will be expanding worldwide next year, providing what perhaps could eventually become a default migration path for moving virtual machines (VMs) into public clouds. VMware certainly holds a strong position in enterprise data centers and Amazon Data Services continues to run ahead of its competitors in public clouds.

VMware Cloud on AWS is powered by VMware Cloud Foundation, which is the unified software-defined data center (SDDC) platform that integrates vSphere, VMware vSAN and VMware NSX virtualization technologies.

We've been tracking the rise of AWS for some time. As AWS gets bigger, it is almost as if the force of gravity is pulling in more data and applications into its data centers. Although VMware has designated AWS as its primary public cloud infrastructure partner, it is officially pursuing a multi-cloud strategy, where the presumption is that enterprises will continue to operate their own data centers for the foreseeable future while moving selected workloads, applications and data stores to multiple public clouds, one of which would be VMware Cloud on AWS.

Raghu Raghuram, chief operating officer, Products and Cloud Services, VMware, states: "Customers are accelerating digital transformation by deploying applications across clouds, with upwards of two-thirds of enterprises deploying applications on three or more clouds today," said Raghu Raghuram, chief operating officer, Products and Cloud Services, VMware. "VMware Cloud brings a consistent operating model, enterprise control, and investment protection for IT resources and skillsets to a multi-cloud world."

VMware's big picture slide shows vision


Many partners are stepping forward at this week's VMworld show in Las Vegas to join this vision. We profiled some of them in the first part of this article. But could it be that all this activity really benefits just one player in the end? Will this new workload mobility benefit AWS above all?
VMware’s auxiliary offerings

While the AWS provides the bare-metal infrastructure (and integration with its broad cloud services), the VMware Cloud Foundation has many other hooks to keep the customers on platform even once their workloads are in an AWS data center.  These include:

  • VMware AppDefense: a data center endpoint security solution that protects applications by embedding application control, and threat detection and response capabilities into the VMware vSphere-based environments on which applications and data live. By leveraging vSphere, AppDefense gains a deep understanding of the intended state and behavior of the applications running on virtual machines and can detect and respond to unauthorized changes
  • VMware Cost Insight: a cost monitoring and optimization service for public and private clouds that helps IT analyze cloud spend, find savings opportunities, and communicate the cost of services to the business.
  • VMware Discovery: an automated inventory service that improves cloud visibility and tames shadow IT by bringing together inventory information and cloud accounts from multiple clouds, making it easy for IT to search for and identify workloads deployed from their enterprise. Using native cloud tags and properties that have been identified, customers can group cloud resources even if they span across multiple clouds. 
  • VMware Network Insight: a network and security analysis service offering purpose-built for public clouds and software-defined data centers. Network Insight provides comprehensive network visibility and granular understanding of traffic flows to enable cloud security planning and network troubleshooting.
  • VMware NSX Cloud: a service that provides consistent networking and security for applications running in multiple private and public clouds, via a single management console and common API. Micro-segmentation security policy is defined once and applied to application workloads running anywhere -- in cloud virtual networks, regions, availability zones -- and across multiple clouds.  Overlay networking enables more precise control over topologies, traffic flows, IP addressing, and protocols used in public clouds. 
  • Wavefront by VMware: a metrics monitoring and analytics platform that handles the high-scale requirements of modern cloud-native applications. Wavefront by VMware's speed, scale, and flexibility empowers DevOps, and developer teams with instant insight into the performance of highly-distributed cloud-native services. Wavefront by VMware's analytics, query-driven alerts, interactive visualizations, open API, and integrations, all powered by a scalable time-series database, deliver "first pane of glass" visibility to help DevOps teams detect performance anomalies while enabling high availability of key cloud services. 


Rackspace seeks an edge 

Rackspace is jumping into the VMware Cloud on AWS game too, announcing plans to participate in VMware's Managed Service Provider (MSP) program early next year. This is a tricky game for Rackspace. At first blush, we wonder what strategic advantage Rackspace would have in helping customers move workloads into somebody else's data center, especially those of Amazon Web Services. Rackspace already runs one of the largest vSphere footprints in the world.

Some might say that once a workload has been moved out of your house and into the arms of AWS it may never come back. Maybe that's a risk that Rackspace, and all others announcing support for VMware Cloud on AWS, will have to take. However, Rackspace is a Premier Consulting Partner for AWS, so maybe this strategic choice was made some time ago.

Rackspace said it can bring its own spirit of "fanatical support" to mutual customers seeking a multi-cloud choice. Their idea is to let customers run their VMware workloads out of the data center and in the best-fit location, whether in Rackspace datacenters or VMware Cloud on AWS. What can it offer? Rackspace's value-add be in providing architecture, provisioning and management guidance. In the long term, even with fanatical support, will it be enough to retain the customers?

Peter FitzGibbon, vice president and general manager of VMware at Rackspace states "Operating across multiple cloud deployments is relatively new to many organizations, however, and some mutual customers will want support to operate VMware Cloud on AWS effectively. As a leading VMware Cloud Provider Partner and a Premier Consulting Partner in the AWS Partner Network, Rackspace is uniquely positioned to provide multi-cloud expertise and identify customer needs."

CoreSite provides a performance on-ramp

CoreSite is working with Faction, an enterprise-class private cloud and backup infrastructure provider, to deliver expanded multi-cloud service offerings
in both the Northern Virginia and Silicon Valley markets. Specifically, Faction will leverage connectivity to the CoreSite Open Cloud Exchange to deliver high-performance VMware based clouds with Administrator-level access to VMware vCenter that offers unprecedented control, flexibility, and integration capabilities for hybrid & multi-cloud deployments. Faction is launching a multi-cloud NetApp storage solution and Managed VMware on AWS offering. CoseSite's data centers enable the low-latency cloud on-ramps.

In the northern Virginia market, CoreSite operates three highly scalable data centers - one in Washington, D.C. and two on its Reston, VA campus (VA1 and VA2). CoreSite recently announced the expansion of both its Reston and Washington, D.C. campuses, all of which will now total over 1,097,000 square feet of colocation data center space upon full build out.

In Silicon Valley, CoreSite operates seven operational data centers, providing colocation solutions to one of the largest concentrations of Internet and technology companies in the world. These data centers connect more than 185 international and national carriers, social media companies, cloud computing providers, media and entertainment firms, and enterprise customers.

CloudVelox helps speed the migration to VMware Cloud on AWS

CloudVelox, which develops cloud automation and orchestration software for clouds and data centers, introduced its new One Hybrid Cloud software for "workload portability" for customers of VMware Cloud on AWS.

The CloudVelox mission is to provide the flexibility to shift workloads in and out of data centers and clouds without fear of being locked into a single destination environment (cloud or data center). The CloudVelox software can automate the mapping of compute, storage, network and security characteristics of a workload from a source environment to matching infrastructure services in the destination cloud or datacenter. It can be used to migrate workloads from any source to data center environments including between data centers (DC to DC), rack to rack (intra-datacenter) as well as migrate or repatriate workloads from the cloud to the data center (Cloud to DC). This allows users to migrate physical, virtual, or cloud workloads into a VMware virtualized data center, AWS Cloud, AWS GovCloud, or VMware Cloud on AWS.

"VMware Cloud on AWS provides customers a seamlessly integrated hybrid cloud offering that gives customers the SDDC experience from the leader in private cloud, running on the leading public cloud provider, AWS," said Mark Lohmeyer, vice president, products, Cloud Platforms Business Unit, VMware. "Solutions such as One Hybrid Cloud™ enable IT teams to reduce cost, increase efficiency, and create operational consistency across cloud environments. We're excited to work with partners such as CloudVelox to enhance native VMware Cloud on AWS capabilities and empower customers with flexibility and choice in solutions that can drive business value."

Monday, September 11, 2017

VMware looks to the clouds - part 1

VMware is reshaping its entire corporate strategy around the cloud and this year's VMworld conference is the embodiment of its full ambitions.  This statement could have been made at any of the past five years of VMworld shows. Each year there has been a new iteration as the company has evolved from a go-it-alone and own-the-cloud strategy to now being culturally attuned to a multicloud future.

The extent of VMware's cloud ambitions is nicely summarized in the following slide, which was shown by VMware's CEO, Pat Gelsinger, in the opening keynote of this week's VMworld show being hosted in Las Vegas.



The first big announcement from this year's event is that VMware Cloud on AWS is now commercially available in AWS U.S. West (Oregon) region. Other AWS data centers worldwide will follow in 2018. The strategic alliance between VMware and Amazon Web Services was first announced nearly one year ago. Under this partnership, AWS is designated as VMware’s primary public cloud infrastructure partner; and VMware will be AWS’s primary private cloud partner. The joint service will be sold by VMware. It runs on the latest, elastic, bare metal AWS infrastructure.
The big idea with hosting VMware Cloud on AWS is that enterprises will want to move virtual machine workloads and applications across VMware vSphere-based private, public and hybrid cloud environments, all with optimized access to AWS services, which could include S3 storage, cloud-based authentication, compliance and security monitoring, and real-time analytics. In short, VMware customers can use their existing VMware software and tools to leverage AWS’s global footprint and breadth of services, including storage, databases, analytics, etc.

As corporate IT moves applications and data to AWS, VMware has a strategic interest in ensuring that its virtual machine paradigm carrier through to the public cloud. It makes sense for VMware to do anything necessary to ensure that IT teams continue to manage their AWS-based resources with familiar (and profitable) VMware tools. If VMware had failed to make this deal, their very extensive base of Fortune 1000 customers who have begun to erode with every corporate application moved into AWS.

For AWS, VMware is great catch too. At this early stage of the public cloud market, the game is all about moving fast to build data centers, establish sticky services, and bring as many customers onboard before other competitors get to them. VMware is deeply entrenched in enterprise data centers. This could provide a very big on-ramp for moving corporate workloads into the AWS cloud.
Naturally, both companies are making a big deal out of this arrangement.

"VMware and AWS are empowering enterprise IT and operations teams to add value to their businesses through the combination of VMware enterprise capabilities and the breadth and depth of capabilities and scale of the AWS Cloud, providing them a platform for any application," said Pat Gelsinger, chief executive officer, VMware. "VMware Cloud on AWS gives customers a seamlessly integrated hybrid cloud that delivers the same architecture, capabilities, and operational experience across both their vSphere-based on-premises environment and AWS."

"With the availability of VMware Cloud on AWS, for the first time customers can operate a consistent and seamless hybrid IT environment that combines the VMware software they love with the unmatched functionality, security, and operational expertise of the AWS Cloud," said Andy Jassy, chief executive officer, AWS. "The majority of the world's enterprises have virtualized their data centers with VMware, and now these customers can easily move applications between their on-premises environments and AWS without having to purchase any new hardware, rewrite their applications, or modify their operations."

The VMware / AWS alliance is hoping to build an ecosystem of third-party tools and services to fill in the gaps, many of which are significant barriers preventing applications from leaving the corporate data center. The companies said that more than 30 solutions are ready for VMware Cloud on AWS launch, including DevOps tools, application migration tools, data protection software, cloud analytics, security, etc.  Here are a few of the partners.

Chef

Chef, which specializes in automation tools, is introducing a solution for VMware Cloud on AWS.  Chef’s automation tools help data center teams to apply a consistent workflow across their infrastructure. The new capabilities bring together VMware’s enterprise-class Software-Defined Data Center (SDDC) software and elastic, bare-metal infrastructure from AWS. Chef tools can be used to move workloads into production using consistent workflows, which helps in decreasing deployment risk and ensuring procedural compliance. The bottom line is this should be good for VMware and for AWS.

CloudCheckr

Another start-up called CloudCheckr, also introduced tools for helping enterprises migrate VMware workloads to the VMware Cloud on AWS.  The CloudCheckr solution, which is initially being launched in the AWS US West (Oregon) region, gives organizations consistent operating model foe application mobility. IT staff might use the tools to calculate the optimal configurations for running VMware vSphere in the public cloud. The tool converts data center specifications into comparable clusters running the same workload on specialized instances within AWS. This output can be used to understand expected utilization levels for CPU, storage and memory, which shows the anticipated monthly AWS spend of any given configuration.

Datapipe

Datapipe, which specializes in managed services for public cloud platforms, will launch a new service to help manage customers’ AWS and VMware Cloud on AWS footprints. Datapipe sees an opportunity to assist enterprises with planning, building, and running their approach for AWS with its full-service offering. Specifically, Datapipe will provide a unified management framework with a single monitoring platform, network connectivity options, and best practices based security and governance.

Mellanox

Mellanox sees a play to make the movement of VMs running is software-defined data centres more efficient. Their ideas to it combine Mellanox iSER (iSCSI Extensions for Remote Direct Memory Access [RDMA]) networking solutions with VMware vSphere. iSER uses an Ethernet Storage Fabric as a unified connectivity solution for compute and storage. Mellanox says this eliminates the need for Fibre Channel while providing improved performance at a lower cost. Its preliminary benchmark results show that iSER accelerates storage throughput by more than 3x and IOPs by more than 2x when compared to ordinary iSCSI (Internet Small Computer System Interface). Mellanox also claims 2x better efficiency versus Fibre Channel storage connectivity.

Trend Micro

Trend Micro is bringing its server security product to VMware Cloud on AWS, promising seamless visibility and security for virtualized workloads across the SDDC, whether on-premises or in the new VMware Cloud on AWS.

McAfee

McAfee is also jumping aboard VMware Cloud on AWS with its Management for Optimized Virtual Environments (MOVE) AntiVirus, which is a threat protection solution optimized for virtual environments. This eliminates the need to install an agent on every VM. McAfee MOVE AntiVirus offloads all on-access scanning to a dedicated VM that runs McAfee VirusScan Enterprise.

China Telecom is growing but ARPU continues to erode

For the first half of 2017, China Telecom posted total revenue of RMB 184.1 billion, up 4.1% from RMB 176.8 billion a year ago – a faster pace of growth than many other global carriers, but likely slower than China’s overall GDP. While revenue for mobile and wireline services grew a faster pace than the global average, China Telecom’s net profit expanded at an even better pace – up 7.4% yoy to RMB 12.5 billion.

This positive news from China Telecom came about one week after China Unicom finalized a US$11.7 billion investment round with Alibaba, Tencent and nine other leading Chinese companies.  With the Chinese market currently in contention by three rivals (China Mobile, China Telecom and China Unicom), the strategic investment, with the backing of the central government, has the potential to tip the scale in favour or China Unicom.  The cash infusion strengthens the balance sheet and provides the opportunity to leap ahead in network rollout and 5G prepositioning. For comparison, China Telecom’s company-wide CAPEX for the first half of 2017 amounted to RMB 41 billion (US$6.20 billion approximately) and is expected to reach RMB 89 billion for 2017. down 8% compared to 2016. While China Telecom is cutting network expenditures, China Unicom may get a much-needed boost to advance its infrastructure.

China Telecom’s recent past

China Telecom is the original incumbent fixed line operator for China. Until 1995, it was purely a government agency under China’s Ministry of Posts and Telecommunications. In 2002, a historic split took place with some of the provincial and city assets being assigned to the newly-created China Unicom, and some being retained by the newly listed China Telecom Corporation Limited. China Telecom made its entrance into mobile communications by pursuing a personal handy phone network technology, before eventually adopting CDMA 2000 and then in 2015, finally launching 4G LTE for its mobile backbone.

Today’s mission

For its part, China Telecom says its mission is to build “three superior networks, namely 4G network, IoT network and all-fibre network with further reinforcement of network edges.” As the legacy operator for much of the country, China Telecom suffers from the same, steady loss of fixed access line.  In its case, fixed lines have fallen from 137 million two years ago to about 124 million now. However, over the same two-year period, the number of FTTH subscribers has nearly doubled, from 61.92 million in mid-2015 to about 117 million today.  Sometime before the end of this year, China Telecom will have tipped the scales from the past to the future, as the number of FTTH subscribers zooms past legacy copper fixed access.

The big trends for 1H2017

  • Service revenues amounted to RMB165.8 billion, representing an increase of 6.8% over the same period last year
  • EBITDA was RMB52.4 billion, representing an increase of 3.7% over the same period last year while EBITDA margin5 was 31.6%.
  • Net profit was RMB12.5 billion, representing an increase of 7.4% over the same period last year.
  • Capital expenditure was RMB41.1 billion while free cash flow was RMB7.2 billion with remarkable improvement over last year.
  • The net increase of mobile subscribers was 14.85 million, reaching a total of 230 million
  • China Telecom now estimates its national mobile market share at 16.8%.
  • The net increase of 4G users was 30.15 million, reaching a total of 152 million.  This means China Telecom’s penetration rate of 4G users increased has reached 66%.
  • The aggregate handset Internet data traffic increased by 126% compared to the same period last year while the DOU of 4G users reached 1.4GB, representing an increase of 56% over the same period last year.
  • An additional issue is the vast ARPU disparity between China Telecom (along with its home market rivals) with developed markets in the West. The average 4G user in China pays roughly US$11 per month, while users in the United States or Western Europe might pay 4X, 5X or 6X this amount.  The disparity roughly follows per capita GDP differences between the country. However, much of the capital expenditure, at least for network infrastructure, will be closer. 

Usage continues to grow

China Telecom is also reporting one mobile usage trend that may not be common with most other mobile operators – mobile voice usage continues to rise even as handset data traffic screams ahead.  In Q2 2017, China Telecom’s handset data traffic (kTB) reached 633.6 kTB up from 491.6 kTB in Q1.  Meanwhile, mobile voice usage reached 196,735 million minutes in Q2 2017, up from 179,556 million minutes in Q1.  People are talking more even as they become increasing depended on their smartphones.
The net increase of wireline broadband subscribers was 4.98 million, reaching a total of 128 million.
The net increase of FTTH subscribers was 11.24 million, reaching a total of 117 million while the penetration rate reached 92%.

The ARPU quandary

While other carriers have seen a big pop in average revenue per subscriber (ARPU) in the transition from 2G/3G to 4G, the gain at China Telecom is from RMB 56.8 (US$8.55) to RMB 67.2 (US$10.12) As noted above, the migration of the subscriber base to 4G is now 66% complete, so revenue will continue to grow as the remain 34% upgrade their smartphones and move onto a 4G data plan. However, China Telecom’s 4G ARPU is also falling, down 7.7% from RMB 72.8 (US$10.96) in 2016 to RMB 67.2 (US$10.12) today. The rate of decrease in ARPU is greater than the growth in overall revenue. If the recent past is any guide, it’s unlikely that China Telecom will be able to increase 4G prices anytime soon.  With growing competition, ARPU probably will continue to slide.

An additional issue is the vast ARPU disparity between China Telecom (along with its home market rivals) with developed markets in the West. The average 4G user in China pays roughly US$11 per month, while users in the United States or Western Europe might pay 4X, 5X or 6X this amount.  The disparity roughly follows per capita GDP differences between the country. However, much of the capital expenditure, at least for network infrastructure, will be closer.

Usage continues to grow

China Telecom is also reporting one mobile usage trend that may not be common with most other mobile operators – mobile voice usage continues to rise even as handset data traffic screams ahead.  In Q2 2017, China Telecom’s handset data traffic (kTB) reached 633.6 kTB up from 491.6 kTB in Q1.  Meanwhile, mobile voice usage reached 196,735 million minutes in Q2 2017, up from 179,556 million minutes in Q1.  People are talking more even as they become increasing depended on their smartphones.

Tuesday, September 5, 2017

China Unicom's US$11.7 billion strategic investment brings advantages

China United Network Communications Group (China Unicom or 中国联通), the fourth largest mobile operator with over 270 million subscribers, recently closed a deal that will bring in US$11.68 billion in cash from top Chinese tech companies. Under the deal, the Shanghai-based holding company of China Unicom, will sell a 35.2% stake worth RMB 78 billion to 14 strategic investors, including the following.

The deal is backed by the Chinese government, which holds a significant stake in the China Unicom group and its various listed companies, as part of a wider effort to encourage private investment in state-owned enterprises. In this case, the multisided arrangement between China's biggest tech companies also ties the largest players so that their financial interests are closer aligned. After the investments are completed, by way of both new and old share sales, the state will hold approximately 25% of the new China Unicom A shares.

The deal is complex and since it was announced on Wednesday, trading of China Unicom shares has been halted on the Hong Kong market.  The transaction is described as follows: “Strategic investors will subscribe for about 9 billion new shares of Unicom A Share Company and purchase 1.90 billion shares of Unicom A Share Company from Unicom Group, representing in aggregate 35.2% of Unicom A Share Company’s enlarged share capital, at a price RMB 6.83 per share.”

While China Life is expected to invest the large amount, the deal is interesting mainly for the synergy it could generate with the Internet companies. There has already been word from Tencent that a special broadband card option will be available to China Unicom's mobile users giving the unlimited access for Tencent applications.

Current status of China Unicom

In mobile subscriptions in China, Unicom holds second place, far behind China Mobile but ahead of China Telecom. In fixed-line broadband, China Unicom trails behind its two rivals. As seen from China Unicom's 1H2017 financial results (highlighted below) competition is fierce between these top three operators.

In the first half of 2017, China Unicom’s service revenue reached RMB 124.11 billion, up by 3.2% year-on-year. EBITDA amounted to RMB 43.56 billion, up by 5.5% year-on-year, and accounted for 35.1% of the service revenue, up by 0.8 percentage point year-on-year; and profit attributable to the equity shareholders of the company increased by 68.9% year-on-year to RMB 2.42 billion.
Key operating metrics

Total handset data usage for 1H2017 soared 326% over 1H2016, reaching 2,522 MB (monthly average mobile billing handset subscriber DOU.
Handset Internet Access Revenue for 1H2017 reached RMB 42.90 billion, up 22.9% yoy.
Fixed-line service revenue (including legacy voice) was RMB 46.57 billion, basically flat yoy.
Fixed-line broadband access revenue amounted to RMB 21.56 billion, down by 3.0% year-on-year. The number of fixed-line broadband subscribers increased by 4.0% year on-year to 76.92 million.

China Unicom plans to several high-priorities for the remained of 2017 and 2018, including:

Accelerating mobile customer migration onto the 4G LTE network. In the first half of the year, Unicom added 34.26 million 4G subs, bringing it to 138.81 million out of a total customer base of 269.45 million, which means that about 52% of its customers are using the 4G network and 48% remain of 3G or 2G service. The migration is slower than seen by other operators worldwide, but the subscriber base to migrate is much higher, China is a huge country, and subscribers have to agree to a substantially more expensive monthly package. China Unicom’s 4G ARPU is RMB 66.50 (US$9.97) vs the company’s overall mobile ARPU of RMB 48 (US$7.19).
The 4G network rollout is mostly done.  In 2016, China Unicom installed 337,000 4G base stations bringing it to a total of 736,000 4G base stations in service.
CAPEX is down substantially for the second year in a row. The cost reduction trend is accelerating. In fact, in 1H2017 capital expenditure decreased by 49.5% year-on-year to RMB9.14 billion i
Capital Expenditures

  • 2015 – RMB 133.9 billion
  • 2016 – RMB 72.1 billion
  • 2017 estimate – RMB 45.0 billion

To drive ARPU, China Unicom is pushing subscription packages based on a “data + content” model. A recently launched “Unlimited Video Enjoyment”. WO Video, already has about 16 million subscribers. China Unicom will focus on differentiated video partnerships.
China Unicom claims to be the 2nd largest Internet data center operator / cloud provider in China. It has 12 national-standard cloud data centres in operation with approximately 122,000 data cabinets in services. China Unicom also has some 300 “local data centers”. Revenues for IDC/Cloud for 1H2017 reach RMB 5.80 billion, up 22% yoy.
China Unicom now has a pre-commercial NB-IoT network in operation.
In fixed-line home broadband, China Unicom reports that it has encountered extremely intense competition. For 1H2017, fixed-line broadband access revenue fell to RMB 21.557 billion from RMB 22.231 billion a year earlier. To counter this, China Unicom is again looking to pull more subscribers unto fiber access, and to feature IPTV with enriched content package to drive differentiation.

Implications of the newly announced strategic investments

Another interesting dynamic in this deal is that China Mobile and China Telecom is that the government pushed for this investment deal with China Unicom first.  The new investment and the commercial partnership could become a strategic advantage for China Unicom if the other two carriers fail to secure similar deals.  In terms of finishing the 4G rollout or beginning the 5G construction, the new cash will be advantageous. Special bandwidth + bundling arrangements with Tencent and Alibaba could be alluring.

The investment deal involves only private Chinese companies, so in this stage of reform, China is not acting to further open its telecom market to foreign investment. China Unicom has had an ongoing partnership with Telefónica for many years. A strategic alliance between the firms began in 2009 when China Unicom and Telefonica invested US$1 billion in the other party's shares. In 2011, the companies deepened their cooperation by investing the equivalent of US$500 million in the other party through the purchase of each party's shares. China Unicom raised its stake in Telefonica through the acquisition of 21,827,499 Telefónica shares at an agreed value of 17.16 Euros per share. In turn, Telefonica was to buy more shares in China Unicom from third parties. Since then, the companies have collaborated in technical research and in network interconnections. More recently, Telefónica has sold some of its China Unicom shares.  Its current holdings in China Unicom are probably under 2%.

In terms of overseas investments or strategic expansion China Unicom, like China Mobile and China Telecom, has pretty much abstained from large scale venture or acquisitions. Whereas carriers like Vodafone, Orange, Zain, MTN, NTT, Bharti Airtel or Singtel, have ventured widely into all of the countries of Asia and Africa, China Unicom has kept its focus only on its home turf. With the reduced government equity stake in the company, perhaps this will begin to change but it is too soon to know.

Network management policies could evolve in step with these deals

Another area to watch with this transaction involves network management. As it brings on strategic content partners, like Tencent, China Unicom will be incentivized to ensure that subscribers get assured access to bundled programming and applications.

Wednesday, August 30, 2017

Kuwait-based Zain Group positions for digital transformation

Kuwait-based Zain Group, which now has operations in eight markets across the Middle East and Africa, has just raised $846 million in cash by selling a 9.8% equity stake to neighboring Omantel. The all-cash deal adds a measure of liquidity to Zain Group, which is aiming to transform itself into a digital service provider as it prepares for 5G and other advanced infrastructure.

Zain Group, which was established in 1983 as Kuwait’s Mobile Telecommunications company, once pursued a very geographically expansionist strategy. In 2005, it acquired mobile operations in 13 African countries from Celtel International for a reported US$3.4 billion, including networks in Burkina Faso, Chad, Democratic Republic of the Congo, Gabon, Ghana, Kenya, Madagascar, Malawi, Niger, Nigeria, Sierra Leone, Tanzania, Uganda and Zambia.

Five years later, Zain decided to exit these ventures while making a nice profit on the investment. These African businesses were sold in 2010 to India’s Bharti Airtel for US$10.7 billion.

In 2008, the Zain Group raised US$4.49 billion (by issuing new shares) to support strategic expansion into the Kingdom of Saudi Arabia, and Nokia Siemens Networks was awarded a contract valued at US1 billion to rollout the network.

Even now, Nokia continues as a lead vendor to Zain Saudi Arabia, as well as other markets. In May 2017, the companies confirmed the deployment of Nokia’s multi-access edge computing (MEC) platform in Mecca.  A similar installation also uses Nokia centralised RAN technology.to boost network upload speeds at Jeddah's King Abdullah Sports stadium by up to 50%.

Since selling its African operations in 2010, Zain has stayed to closer to home, focusing its external efforts on Bahrain, Iraq, Jordan, Lebanon, Morocco, Sudan, South Sudan, and the very important market of Saudi Arabia. Some of these countries, especially Iraq, Sudan, and South Sudan, are beset by social, political and economic issues – but everyone wants/needs mobile connectivity so demand remains strong.

In a management shake-up earlier this year, Zain’s board of directors appointed Mohannad Mohammed Al-Kharafi as the Chairman of Zain Group, Bader Nasser Al-Kharafi as Vice-Chairman and Chief Executive Officer of Zain Group, and appointed Scott Gegenheimer in a new role as Chief Executive Officer of Operations. Previously, Gegenheimer, a U.S. citizen, served simply as CEO for all of Zain Group since 2012. Before joining Zain, Gegenheimer held leadership positions at several regional operators, as well as with Cisco Systems and Motorola.

At the end of June 2017, Zain Group counted 45.2 million customers. The breakdown by country is roughly as follows:
Iraq 27%
Sudan 27%
KSA 23%
Jordan 9%
Kuwait 6%
Lebanon 5%
Bahrain 2%

Declining revenue and EBITDA for the first half of 2017

Earlier this month, Zain Group consolidated first half 2017 revenues of KD 508 million (US$1.67 billion) down 8% year-on-year (Y-o-Y) in KD terms. The Group’s consolidated EBITDA for the period reached KD 212 million (US$695 million), down 17% Y-o-Y in KD terms, reflecting an EBITDA margin of 41.7%. Consolidated net income remained stable at KD 82 million (US$270 million). Earnings per share for the half-year stood at 21 Fils (US$0.07).   Overall, the company described its financial performance as “in line with expectations” while acknowledging the impact of a significant 61% currency devaluation in Sudan and other factors. (the company says Zain Sudan continues to perform ‘exceptionally well’ in local currency terms),

Some key items and indicators

Data revenues for the group (excluding SMS and VAS) increased 4% Y-o-Y, representing 25% of the consolidated revenues.

Zain launched an over-the-top, streaming video service called “iflix” across several markets. This follows the announcement earlier in the year that Zain and iflix had formed a joint venture entity named ‘iflix Arabia' to be headquartered in Dubai. The JV will trade commercially as “iflix”, adding Zain’s territories of operation to iflix’s global footprint, including Kuwait, Bahrain, Iraq, Jordan, Lebanon, Saudi Arabia and Sudan, with the potential to further extend into additional regional markets. The content catalogue will include highly acclaimed Arabic shows and movies, exclusive Arabic content series, best titles from Hollywood and Bollywood, local programming and children’s shows.

In its home market of Kuwait, Zain’s customer base stands at 2.6 million. Kuwait remains the Group’s most profitable operation with revenues reaching KD 167 million (USD 549 million), EBITDA amounting to KD 66 million (USD 215 million) and net income came in at KD 39 million (USD 128 million). Zain Kuwait’s EBITDA margin stood at 39% at the end of the six-month period, with data revenues (excluding SMS & VAS) accounting for 32% of total revenues.

Zain Kuwait is currently implementing a smart meter project, in one of the sector’s largest ICT projects for the country’s Ministry of Electricity and Water. The Smart Meter project, which runs through 2024, is a key step in the company's strategic plans to deploy smart city solutions in Kuwait and beyond. Ericsson has been selected as the sole technology partner in the Zain led consortium.

In June 2017, Zain launched Cloud Disaster Recovery (Cloud DR) service in Kuwait in collaboration with IBM. The new service provides Zain’s enterprise customers with cloud-based business continuity capabilities and faster disaster recovery of their critical IT systems.

In Saudi Arabia, Zain reports improved financial indicators thanks to a turnaround and cost optimisation program. In H1, 2017, the operator recorded its first-ever half yearly net profit of USD 14 million, compared to net losses USD 154 million in H1 2016. Revenues for the period were up by 9%, reaching USD 1.04 billion. The company recorded a significant 59% increase in EBITDA to reach USD 346 million in H1 2017.

Zain noted that the introduction of a biometric identification requirement for mobile services caused its total customer base to shrink by 15% to stand at 9 million customers at the end of June 2017. Impressively, the operator witnessed a 42% rise in data revenues (excluding SMS and VAS) Y-o-Y, representing 50% of total revenues.  

During Q2, Zain KSA also successfully secured an additional 1800MHz spectrum for expansion of its4.5G LTE network's coverage and capacity.

In January 2017, Zain Group appointed Peter Kaliaropoulos, an Australian national, as CEO of Zain Saudi Arabia. Previously, Kaliaropoulos was the GM of ‘touch’ Lebanon until June 2016, the country’s leading operator that Zain manages on behalf of the Lebanon Telecom Ministry

In Iraq, Zain managed to achieve US$523 million revenues due to the impressive growth in data usage and numerous customer acquisition initiatives in the northern regions of the country. The operation’s efficiency drive saw EBITDA reach USD 179 million, reflecting a 34% EBITDA margin. Net income amounted to USD 11 million for the period. Zain Iraq leads the market serving 12.9 million customers, which represented an impressive 15% Y-o-Y increase.

In Sudan, the 61% currency devaluation this year impacted financial results as measured in USD terms for the first six months of 2017. Nevertheless, in local currency (SDG) terms, revenues grew by 38% Y-o-Y to reach SDG 3.4 billion (USD 213 million, down 44% in USD terms) for the first six months of 2017. EBITDA increased by 22% to reach SDG 1.3 billion (USD 81 million, down 50% in USD terms), and net income increased by 14% to SDG 545 million (USD 34 million, down 54% in USD terms). Data revenues (excluding SMS and VAS) accounted for 15% of total revenues, with an impressive annual growth rate of 69%. The operation saw its customer base expand 3% to reach 12.9 million.

In Jordan, Zain grew its customer base by 3% Y-o-Y, serving 4.2 million customers at the end of June, and maintaining its market leading position despite intense price competition. Y-o-Y revenues increased 2% to reach US$241 million, with EBITDA up 1% to reach USD 116 million, reflecting an impressive 48% EBITDA margin. Net income decreased 5% to USD 48 million for the six-month period. With the continual expansion of 4G services across the country, data revenues (excluding SMS & VAS) represented 37% of total revenues, up by 15% Y-o-Y.

In Bahrain, Zain generated revenues of USD 100 million for the first six months of 2017, up 17% Y-o-Y. EBITDA for the period amounted to USD 30 million, down 8%, reflecting an EBITDA margin of 30%. Net income amounted to USD 4 million, reflecting a 21% decrease. Data revenues (excluding SMS & VAS) increased 36% Y-o-Y, representing 43% of overall revenues.

Monday, August 28, 2017

AT&T preps for its mega-merger with Time Warner – Part 2

Part 1 of this article covered operational metrics from AT&T's Q2 2017 financial report. This week, AT&T confirmed that its $109 billion blockbuster acquisition with Time Warner is on track and likely to close by the end of the year. Although the company did not explicitly say that regulators, including the Department of Justice and the FCC, are likely to approve the merger, one can presume that no major issues have been presented yet that could threaten the deal. As the first mega-merger to face review under the Trump Administration, it has not been exactly clear what level of objections such mergers might face, especially under a president who routinely attacks CNN (a Time Warner property) as 'fake news'. Under the Obama administration, AT&T was blocked from acquiring T-Mobile US because the combination was seen as market consolidation that would be detrimental to consumers.

Key terms of the AT&T – Time Warner combination

As a reminder, AT&T officially announced its intention to acquire Time Warner on October 22, 2016. The deal was to be structured as a stock-and-cash transaction valued at $107.50 per share, representing a transaction value of $108.7 billion. The companies said the reason for the deal is to combine Time Warner's library of content and ability to create new premium content with AT&T's extensive customer relationships, world’s largest pay TV subscriber base and scale in TV, mobile and broadband distribution. Time Warner, which was formed in 1990 through the merger of Time Inc. and Warner Communications, encompasses many premium media properties, including HBO, New Line Cinema, Turner Broadcasting System, The CW Television Network, Warner Bros., CNN, Cartoon Network, Boomerang, Adult Swim, DC Comics, Warner Bros. Animation, Castle Rock Entertainment, Cartoon Network Studios, Esporte Interativo, Hanna-Barbera Productions, Warner Bros. Interactive Entertainment. It also owns 10% of Hulu.

The basic idea driving the merger is for Time Warner to act as the content arm for AT&T, providing mobile and fixed broadband line subscribers with valuable material as part of packaged service bundle. Consumers presumably would purchase an AT&T service bundle based on the perceived quality and value of the package rather than simply the lowest price for mobile connectivity. This will allow ARPU to rise and ensure a 'stickiness' factor that goes beyond the latest mobile handset deals, currently a leading cause for subscriber churn.

As a content provider, Time Warner requires far less ongoing capital expenditures than AT&T, which must invest routinely in its infrastructure. AT&T has stated that it expects $1 billion in annual run rate cost synergies within 3 years of the deal closing due to cuts in corporate and procurement expenditures. As there is little overlap between the companies, its not clear where these savings would come from. Time Warner does maintain its corporate headquarters in high-priced Manhattan, but this is where significant content transactions are negotiated and it is unlikely to be relocated to Dallas.

Rising debt load

One outcome of a combined AT&T and Time Warner is that the investor profile will be very different. An article this week by Bloomberg points out that after the merger is completed, AT&T's net debt will rise to $182 billion. This will transform AT&T into a more leveraged conglomerate, putting much greater pressure on how management can use free cash flow from operations. Over the past few years, AT&T has been using significant amounts of cash to buy back its own shares, thus increasing shareholder value. AT&T has also been one the more generous corporations in terms share dividends, reportedly returning 70% of free cash flow last year, or about $11.8 billion, according to Bloomberg. Going forward, more cash will be needed to service the heavier debt load, so we might expect that less cash will be available for dividends or share buybacks.

Betting on bundling

For the deal to be a success, AT&T and Time Warner will have to generate some real synergy in the minds of consumers. With Net Neutrality principles no longer an area of focus for the FCC under the new chairman, Ajit Pai, AT&T will have more leeway in positioning special service packages for its subscribers. For instance, home entertainment services may be the first big use case for 5G networks starting in 2019. A mobile operator such as AT&T might launch a 4K TV package for residential consumers in certain markets. With connection speeds in the hundreds of megabits, AT&T could deliver a strong 5G TV service featuring exclusive shows from HBO or Castle Rock Entertainment, while also including in the package regular LTE smartphone connectivity for all members of a household. Such an offer would be unmatched by rivals such as Sprint and Comcast.

There is already some movement in this direction. AT&T is ready to launch a 5G video trial with DIRECTV NOW service in Austin, Texas. The trial will evaluate how fixed wireless mmWave technology handles heavy video traffic. Previously, AT&T reported that its 5G lab trials were achieving speeds up to 14 Gbit/s and latency of under 3 milliseconds.

But how is video bundling working so far for AT&T? In July 2015, AT&T completed its acquisition of DIRECTV, making it the largest pay TV provider in the U.S. with more than 26 million customers in the U.S. and millions more in Latin America, including Mexico and the Caribbean. The implied total equity value of this deal was $48.5 billion although the total transaction value was $67.1 billion, including DIRECTV's net debt.

The DIRECTV acquisition also brought exclusive, premier content, particularly live sports programming to AT&T, including exclusive pay TV rights to NFL SUNDAY TICKET, ownership of ROOT SPORTS Networks and minority stakes in the Game Show Network, MLB Network, NHL Network and the Sundance Channel. Soon after completing the merger, AT&T began offering bundles combining cellular service, satellite-TV or U-serve TV over FTTH (in certain markets). These offers have been widely promoted via TV and print advertising, in the carrier's retail stores, and in notices to AT&T's millions of mobile users.

On a regular basis since then, AT&T execs have described the merger as a 'hit', citing higher ARPU from consumers who take both services. By Q2 2016, AT&T added nearly 1 million DIRECTV subscribers. In more recent quarters, this moment appears to have stalled, perhaps because the pay TV segment is no longer growing as more cord-cutting consumers look to over-the-top (OTT) services instead of traditional satellite TV subscriptions. Operating metrics for Q2 showed a loss of DirecTV and U-verse TV subscribers in the quarter. However, AT&T’s over-the-top DirecTV Now service continues to add subscribers every quarter and now stands at 491,000 users. So, if hardware-based bundling is capex intensive and no longer growing, perhaps OTT packages are the way to go.


AT&T appointments executives ahead of merger with Time Warner



AT&T has announced a number of executive appointments in preparation for completing its acquisition of global media and entertainment company Time Warner; the transaction is currently under review by the U.S. Department of Justice and competition authorities in certain foreign countries.Effective August 1st, the following executives will assume new positions and continue to report to AT&T chairman and CEO Randall Stephenson:1.  ...



Thursday, August 24, 2017

Flash Memory Summit – big changes in non-volatile memory part 3

Hyperconverged platforms, such as those offered by Nutanix, have proven to be extremely successful in the market because they integrate compute, networking and storage in a single, scale-out box. They are a new way of looking at the old problem of how best to connect these three resources. In a similar fashion, the Open Compute Project, which was launched by Facebook six years ago, set out to rethink how compute, storage and networking could be optimised at the rack level to build hyperscale data centres.

What we’re seeing now, as evidenced by the 2017 Flash Memory Summit in Silicon Valley, is that non-volatile memory is advancing at a faster pace than other storage technologies, and at faster pace than compute (CPUs and GPUs), or networking. Ethernet has continued to progress in either 10X or 4X steps, but recently, these have taken time. In data centres, 10G backbones are common. Carrier backbones typically run utilise 100G links.

These statements were true a year ago – or even two years ago. We see some 400G pluggable transceiver apparently ready for market this year. But will 400G be rapidly adopted in either data centres or carrier networks? For a variety of network engineering reasons, implementing 400G in a network is not as easily done as deploying new SSDs with 4 times the capacity as last year’s model.

More importantly, Samsung Electronics has a ten-year roadmap showing how its 3D NANDs will evolve from 4th generation to 5th, 6th, 7th, 8th, 9th and 10th over the course of a decade.

The company says the physics of the last two generations in this progression have yet to be solved but so far look possible. At this point, it seems likely that this rapid evolution will deliver 2X or 3X capacity improvements every two years of less. On the networking side, we’ve seen the Ethernet Alliance publish an Ethernet Roadmap that envisions a proliferation of new interface speeds. This roadmap predicts terabit speed interfaces by 2020, scaling up to 10 terabits/second by 2030. Storage innovation may be winning this race.

Here are some other interesting observations on the storage market.

Western Digital pushes 3D NAND to 96 layers

Western Digital, which received the Flash Memory Summit ‘Best of Show’ award for its BiCS4 technology, has now pushed its 3D NAND technology to 96 layers of vertical storage capability. This marks several years of continuous improvement. In 2016, WD announced 64-layer 3D NAND after achieving 48-layer 3D NAND in 2015. Last month, the company also announced the development of its first four-bits-per-cell (X4) 3D NAND technology. More layers translate into more capacity.
Toshiba faces uncertainty but moves to 64-layer, triple-level Flash

Toshiba’s semiconductor division has been a state-of-turmoil due to restructuring and likely sale. Various suitors have been suggested and apparently rejected either by the company or the Japanese authorities.  Most recently, Toshiba’s management appears to be nearing a deal to sell the business to a consortium led by Bain Capital, although this too may be at an impasse.  The joint venture with SanDisk (a division of WD) focused on flash memory has become mired in legal disputes. Apparently, Toshiba will not ship its latest generation of 96-layer BiCS modules to SanDisk.
Nevertheless, at Flash Memory Summit, Toshiba America Electronic Components (TAEC) introduced its first enterprise SSDs utilizing the 64-layer,3-bit-per-cell TLC (triple-level cell) technology flash memory: the PM5 12Gbit/s SAS series and the CM5 NVM Express (NVMe) series. Toshiba’s PM5 series will be available in a 2.5-inch form factor in capacities from 400GB to 30.72TB], with endurance options of 1, 3, 5 and 10DWPD (drive writes per day). Toshiba also introduced its own consumer SSDs for PCs and laptops based on the same 64-layer technology. Capacity options include 256GB, 512GB, and 1024GB.

Toshiba is also introducing the first MultiLink SAS architecture, enabling up to 3350MB/s of sequential read and 2720MB/s of sequential write in MultiLink mode and 400,000 random read IOPS in narrow or MultiLink mode.

Mellanox looks to NVMe over Fabrics

Mellanox Technologies is pushing ahead with its BlueField System-on-Chip (SoC) for NVMe over a network fabric. BlueField integrates all the technologies needed to connect NVMe over Fabrics flash arrays. It provides 200 Gb/s of throughput and more than 10 million IOPS in a single SoC device. In addition, an on-board multicore ARM processor subsystem enables flexible programmability that allows vendors to differentiate their software-defined storage appliances with advanced capabilities. The BlueField chip can be used to control and connect All Flash Arrays and Just-a-Bunch-Of-Flash (JBOF) systems to InfiniBand and Ethernet Storage fabrics. The Mellanox SoC combines a programmable multicore CPU, networking, storage, security, and virtualization acceleration engines into a single, highly integrated device. Refence storage platforms are now ready.

“By tightly integrating high-speed networking, programmable ARM cores, PCIe switching, cache, memory management, and smart offload technology all in one chip; the result is improved performance, power consumption, and affordability for flash storage arrays. BlueField is a key part of our Ethernet Storage Fabric solution, which is the most efficient way to network and share high-performance storage,” stated Michael Kagan, CTO of Mellanox.

Seagate revs its Nytro Flash storage

Seagate Technology introduced enhanced versions of two flash technologies to boost performance and capacity for mixed data center workloads. The updated solid-state drives — including the 2 TB Nytro 5000 M.2 non-volatile memory express (NVMe) SSD and the Nytro 3000 Serial Attached SCSI (SAS) SSD — address different segments of the cloud and data center markets. The latest Nytro 3000 SAS SSD offers a dual-port SAS interface to maintain data integrity in the event of an unexpected communication channel loss. Capacity is 15TB, more than four times the capacity of the previous version.

Seagate also previewed plans to offer a 64-terabyte (TB) NVMe add-in card (AIC). This forthcoming product boasts a read speed of 13 gigabytes per second (GB/s) — the fastest and highest-capacity SSD ever demonstrated.

“Large-capacity SSDs are in high demand in hyperscale computing, a market that is growing faster than any other sector,” said Jim Handy, general director of research firm Objective Analysis. “Seagate’s new SSDs, with their high-performance interfaces and high capacities, should find ready acceptance in this market and other data center applications.”

WekaIO, a start-up based in San Jose, California with R&D in Israel, introduced a cloud-native scalable file system that scales to exabytes of data in a single namespace while delivering a big performance boost to applications, processing four times the workload compared to IBM Spectrum Scale measured on Standard Performance Evaluation Corp. (SPEC) SFS 2014. A key innovation is that WekaIO eliminates bottlenecks and storage silos by aggregating local SSDs inside the servers into one logical pool, which is then presented as a single namespace to the host applications. A transparent tiering layer offloads cold data to any S3 or Swift cloud object store for unlimited capacity scaling, under the same single namespace.

In partnership with Intel, WekaIO is now demonstrating a native NVMe-oF system using the new “ruler” form factor for Intel SSDs. The companies said this enables a storage capacity of beyond 1PB in 1U while delivering more than 3 million IOPS.

Tuesday, August 22, 2017

Flash Memory Summit – big changes in non-volatile memory - part 2

Over the past year, we’ve seen that HDD capacity increases have plateaued. Spinning disks have been surpassed in storage capacity by SSDs. Performance comparisons between the two is not even a topic of debate. For CIOs, the deployment of flash storage arrays is easy and offers an immediate boost in IOPs for critical applications. More importantly, all the innovation in new drive development has shifted to flash. We are now seeing many approaches being tried in the market to boost SSD performance even further, to scale up to new drive capacities and new array architectures, to adopt new form factors for better rack-scale integration, and increase manufacturing volume to finally meet market demand.

In the first part of this article, we covered Samsung’s rapid progression with 3-D NAND technology. With the arrival of its 5th generation 3D NAND next year we will see 2.5” SSDs soar into the 128TB range. The company says its on-track for 5 more generations of 3-D NAND in the coming decade. In this second part of the article, we’ll look at innovations from another giant, Intel, which has also set its sights on bringing non-volatile memory technologies to the forefront of server, system and data centre design, as well as developments from Nimbus Data and the Gen-Z consortium.

Intel’s non-volatile memory advancements

Intel began shipping its first SSDs as early as 2008 and has been on a continuous improvement path ever since. In 2010, Intel and Micron Technology entered into a partnership focused on NAND flash memory. In 2015, Intel and Micron announced 3D XPoint technology, which was described as the first new memory category since the introduction of NAND flash in 1989, with promises to be up to 1,000 times faster and up to 1,000 times greater endurance than NAND, while being cheaper than DRAM and non-volatile. Intel then adopted the "Optane" brand for products based on thistechnology, while Micron adopted the QuantX brand. Optane is fundamentally different from NAND and uses a combination of unique Intel memory + storage controllers, Intel interconnect IP, and Intel software.

Introducing the memory ruler

Intel's big news at last week's Flash Memory Summit was its new "ruler" form factor for SSD. Instead of the traditional, 2.5" or 3.5" rectangular box for disk drives, Intel's ruler is a long, thin box designed to slide in to a 1" server chassis, plugging in via a PCIe interface at the end of the ruler. It is a slick design. Apart from looking better, the long, thin shape dissipates heat easier. Intel showed a 1” RU server chassis accommodating 32 of these SSD rulers, creating up to 1 petabyte of storage. Intel could offer Optane SSDs and/or 3D NAND SSDs in this form factor.

It’s been a while since a new storage drive format gained widespread acceptance. Intel will need to bring its new form factor to standardization, perhaps via the Open Compute Project, although this was not confirmed. The ruler design should prove to be particularly useful in hyperscale data centres, where plug-n-play convenience is especially useful when 100s of thousands of servers need to be maintained. Intel also noted that its ruler form factor could be used for plug-in accelerators, perhaps FPGA boards optimized for specific functions. No timeline was given for when the ruler might enter the market.

Intel and Attala Systems also announced an FPGA-based accelerated RDMA over Converged Ethernet (RoCE) networking solution designed to serve as high-performance, composable storage infrastructure with features such as self-learning orchestration and provisioning capabilities. The idea is to create an adaptable storage infrastructure that is essentially an elastic block storage (EBS) solution, accelerated. Attala Systems is a start-up based in San Jose, California that was founded by Sujith Arramreddy, who previously co-founded ServerEngines (acquired by Emulex for $250 million in 2010) and ServerWorks (acquired by Broadcom for $1.4 billion in 2001). Attala's CEO is Taufik Ma, who previously was co-GM of Intel's Server System business unit before leaving for a storage/networking start-up. Nimbus Data sees 500 TB SSDs by 2020

Nimbus Data is a privately-held develop of all-flash arrays based in Irvine, California. The company observes that 40 million nearline/high-capacity HDDs are shipped per year, and all of them use the 3.5” form factor. At Flash Memory Summit, Nimbus Data introduced a software and multiprocessor solution for OEMs developing next-generation solid state drives for data centres. Whereas conventional SSDs are based on a single flash controller, Nimbus ExaDrive is based on a distributed multiprocessor architecture. Inside an ExaDrive-powered SSD, multiple ultra-low power ASICs exclusively handle error correction, while an intelligent flash processor provides wear-leveling and capacity management in software. Nimbus sees an opportunity for its ExaDrive being used in super capacity SSDs that let data centers rip-andreplace HDDs with flash. ExaDrive supports the standard SAS interface and is optimized to fully utilize the volume of the 3.5” form factor.

Nimbus said its ExaDrive is used by Viking Technology and SMART Modular Technologies in 50 TB and 25 TB SSDs for cloud infrastructure, technical computing, and digital content storage. The company predicts that its ExaDrive software-defined architecture will enable SSDs as large as 500 TB by the year 2020, achieving up to 600 petabytes in a single rack. This represents a 50x increase over what is possible with HDDs today. “ExaDrive’s software-defined multiprocessor architecture for SSDs delivers a game-changing leap forward in capacity, density, and energy efficiency that HDDs will never be able to recoup,” stated Thomas Isakovich, CEO and Founder of Nimbus Data. “ExaDrive broadens the appeal of flash memory to tier 2 and nearline use cases, enabling flash to become the dominant data center storage media.”

Gen-Z consortium targets data centres

The Gen-Z Consortium is a vendor-led group that is developing an open systems interconnect with memory semantic access to data and devices via direct-attached, switched or fabric topologies. Its major members include AMD, ARM, Broadcom, Cray, Dell EMC, Hewlett Packard Enterprise, Huawei, IDT, Micron, Samsung, SK hynix, and Xilinx. At this year’s Flash Memory Summit, the group had planned it’s the Gen-Z multi-vendor technology demonstration, connecting compute, memory, and I/O devices. Despite the unfortunate fire at a vendor booth on the opening day of the event, the demo was still able to occur in a nearby meeting room.

The demo showed FPGA-based Gen-Z adapters connecting compute nodes to memory pools through a Gen-Z switch, creating a fabric connecting multiple server vendors and a variety of memory vendors. Such a highperformance and scalable fabric/interconnect could be implemented in future data centres. The demo also featured a scalable prototype connector defined by the Gen-Z Consortium, running at 112 giga-transfers/sec. “We are excited to showcase the first technology demonstration of Gen-Z that includes solutions from multiple member companies, including a variety of servers, memory and I/O devices, all connected with a Gen-Z fabric,” said Kurtis Bowman, President of the Gen-Z Consortium. “The consortium continues to meet the planned development schedule and we expect to see initial Gen-Z products in the 2019-2020 timeframe.”

http://genzconsortium.org/

Thursday, August 17, 2017

Flash Memory Summit – big changes in non-volatile memory - part 1

Can you imagine a 128 TB SAS SSD? It is coming soon from Samsung in the familiar 2.5” disk drive package and destined for the next generation of cloud data centres. Leading companies and start-ups from across the storage industry met at this week's Flash Memory Summit in Santa Clara, California. A key takeaway from the event is that solid state storage continues to improve at a rate much faster than networking technologies. Solid state drives surpassed spinning disks in total capacity some time ago - Samsung announced a 16 TB SDD in August 2015 and currently offers a 32 TB SSD, but prices remain high.



The market is driven by unrelenting demand for flash drives in laptops, desktops and servers, especially in cloud data centres where there has been an uptick in spending over the last few quarters. NAND prices on a $/GB are significantly higher than they were 12 months. According to data from Objective Analysis, contract prices for NAND averaged $0.30 per gigabyte on July 2nd, compared to $0.20 per gigabyte a year ago. Looking at Amazon.com, the street price of a 500 GB SSD is about the same in mid-2017 as last summer. Meanwhile, with higher prices and relentless demand in the current market, the leading manufacturers of 3D NAND are doing quite well. For Samsung Electronics, this translated into very strong revenue and earnings for its June financial report, which predicted that a tight market for DRAM and 3D NAND will continue for the rest of the calendar year.

In a presentation at Flash Memory Summit, Jim Handy of Objective Analysis predicted that NAND prices will remain stable at these rates through mid-2018, but will then suddenly collapse due to a saturation of new supply entering the market. His argument goes that all vendors have begun to ship 3D NAND but only in limited volume due to the complexity of mastering 3D NAND manufacturing. Over time, these complexities are being ironed out, manufacturers will move to add additional layers of stacking and the cost per GB will become cheaper for 3D NAND than for 2D planar NAND. Objective Analysis expects a steep oversupply of 3D NAND by late 2018, even before significant new manufacturing facilities in China come online.

Disruption at Flash Memory Summit

This year’s Flash Memory Summit was disrupted on opening day by a fire in the exhibition area, apparently an electrical issue at one of the vendor stands. Thankfully no one was hurt, but the exhibits were cancelled for the remainder of the event. Conferences and keynotes were the forum for technological disruptions, of which there are plenty in this rapidly evolving segment.

Firstly, Samsung made several important announcements and previewed that massive 128 TB SSD. At a fundamental level, Samsung said its 3D NAND roadmap is progressing on schedule. Last year, Samsung introduced its 4th generation, 64-layer triple-level-cell V-NAND flash memory. This has now gone into production and is being used for products such as the 32TB SSD. Drive capacity and performance are expected to scale up with the upcoming v5 generation of 3D NAND. Samsung has already started work on v6 and v7, with an assumed 18-month interval between each generation. Samsung executives seemed confident they will be able to squeeze at least ten generations out of 3D NAND technology, which provide another decade of continuous improvement if Flash SSD. Beyond that, other non-volatile memory technologies will need to be developed.
Samsung's 1 TB V-NAND chip

Samsung also announced a 1 TB V-NAND chip, slated for commercial production next year, that will enable 2 TB of memory in a single V-NAND package. This is achieved by stacking 16 x 1 TB dies – an advancement the company considers 'one of the most important memory advances of the past decade'.

Samsung is introducing a 16 TB NGSFF (next generation small form factor) SSD that is designed for use in 1U rack servers. Measuring 30.5 x 110 x 4.38 mm, the Samsung NGSFF SSD aims for improved space utilisation and scaling. The company showcased a 1U sample design, codenamed Mission Peak, that pack 36 of the units for a total capacity of 576 TB in the 1 RU appliance. Samsung is looking for partners on this new drive form factor.

In addition, for extreme SSD read/write performance, Samsung introduced its first Z-SSD product, boasting 15 microseconds of read latency time, which is approximately a seventh of the read latency of an NVMe SSD. At the application level, the company estimates its Z-SSDs can reduce system response time by up to 12 fold compared to using NVMe SSDs.

Samsung is also introducing a technology it calls Key Value SSD. Whereas today's SSDs convert object data of widely ranging sizes into data fragments of a specific size called 'blocks', the new Key Value SSD technology allows SSDs to process data without converting it into blocks. Samsung said its Key Value instead assigns a ‘key’, or specific location, to each value, or piece of object data, regardless of its size. The key enables direct addressing of a data location, which in turn enables the storage to be scaled.

Tuesday, August 8, 2017

Intel sees rapid shift from enterprise to cloud, increased NFV spending

Intel beat financial expectations when it released its Q2 2017 financial results in late July.  The company cited strong growth in its client computing (up 12 percent) and data-centric businesses (up 16%). The good earning report builds on the marketing momentum it established in the quarter with the launch of its Intel Core X-Series family of processors, which are designed for advanced gaming, AR and VR client applications, as well as its Intel Xeon Scalable processors for data centres, artificial intelligence (AI) and other data-intensive workloads. The recent Xeon launch was covered here previously.

Because Intel holds such a dominant and strategic position in the IT ecosystem, its quarterly report is often an excellent measure of the industry’s overall health and an early indicator of significant trends that will impact global network traffic.

The Q2 review

For Q2 2017, Intel reported revenue of $14.8 billion, up 9% year-over-year. After adjusting for the Intel Security Group (ISecG) transaction, which was spun out as an independent company on April 3rd and now known by its original name of McAffee, Intel’s Q2 revenue growth was even better – up 14% from a year ago. Operating income was $3.8 billion, up 190% year-over-year, and non-GAAP operating income was $4.2 billion, up 30%. EPS was $0.58, up 115% year-over-year and non-GAAP EPS was $0.72, up 22%. For Q2, Intel generated approximately $4.7 billion in cash from operations, paid dividends of $1.3 billion, and used $1.3 billion to repurchase 36 million shares of stock.

To top off the good news, Intel raised its full-year revenue outlook by $1.3 billion to $61.3 billion and raised its EPS outlook to $2.66 (GAAP) and $3.00 (non-GAAP), a 15 cent increase over the previous guidance.

Key Business Unit Revenue and Trends
Quarterly Year-Over-Year
Q2 2017
vs. Q2 2016
Client Computing Group
$8.2 billion
up
12%
Data Center Group
$4.4 billion
up
9%
Internet of Things Group
$720 million
up
26%
Non-Volatile Memory Solutions Group
$874 million
up
58%
Programmable Solutions Group
$440 million
down
5%
*Data-centric businesses include DCG, IOTG, NSG, PSG, and all other

Clearly a lot of hot areas and promising technologies at Intel

For the Data Center Group, Intel said its current 9% annual growth rate in Q2 probably can be sustained for the whole year – a fantastic result considering that service provider spending overall, including for mobile infrastructure in developed markets, appears to have stalled. Capex budgets may not be restored to normal levels as a percentage of carrier revenue until the 5G upgrade cycle gets under way.

On its Q2 investor conference call, company execs commented that by 2021 the silicon opportunity for data centres could be worth $65 billion per year and that Intel is currently less than 40% of the total available segment today. Beyond its Xeon processors for cloud servers, Intel is chasing adjacent product categories, including Ethernet, Silicon Photonics and its 3D XPoint memory. Its goal is to rule the full data centre rack, and not just the server motherboard.

Cloud and communications service providers

For Intel's DCG, sales for public cloud zoomed up 35% year over year. On the other hand, enterprise data centre spending declined 11%. The two figures are clearly related, with a rapid shift of workloads to the public cloud underway. One can presume many of these to be new workloads. At the time of deployment, companies are signing up for public cloud capacity instead of buying new servers for their enterprise data centre. Or simply, when servers are ready to be retired, enterprises are moving their workloads to the public cloud rather than buying new servers.

Intel cited communication service providers as another growth vector for DCG. Revenues here rose 17% year over year. For Intel, this is good news as it would seem to indicate that network functions virtualisation (NFV) is finally taking hold. Previously, there have been statements from AT&T and Orange revealing an accelerated schedule to migrate large percentages of their network function workloads onto virtualised infrastructure, i.e. x86 platforms.

Leading deployments with the top-tier communications service providers have been underway for the past year. Intel’s 17% growth rate for CSPs seems to indicate a broader adoption base for NFV. If this growth can be sustained, one should expect other companies in the NFV ecosystem to start showing results as well, such as companies offering virtual network functions (VNFs) such as firewalls and load balancers. The NFV movement has had a very long incubation cycle, and now the real spending by CSPs for Intel gear will be a boost for many players.

Together, the cloud and CSP segments make up nearly 60% of Intel’s total DCG revenue. The other segments include the IoT Group, where Q2 revenues were up 26% to $720 million; the non-volatile memory solutions group (NSG), where sales were up 58% to a record $874 million; and the programmable solutions group (PSG), formerly Altera, where revenue declined 5% to $440 million. Intel completed its $16.7 billion acquisition of Altera in January 2016, so it now has a track record of over one full fiscal year in managing the group’s business. The group specialises in field-programmable gate array (FPGA) technology.

A further note regarding Intel and CSPs

On its Q2 investor conference call, Intel CEO Brian Krzanich said the company is making inroads with its 5G strategy as there are now five ongoing trials underway with global service providers and 15 more in the pipeline. We know from earlier announcement that Intel’s office in Austin, Texas became the first customer site for AT&T’s pilot 5G network in December 2016. The 5G fixed wireless pilot in Austin is delivering and ultra-fast Internet connection and DIRECTTV NOW using Ericsson's 5G RAN and the Intel 5G Mobile Trial Platform.

Mobileye acquisition approaches completion

As Intel makes its transition from a PC-oriented company into a data-centric company it is seeking adjacent opportunities either through internal development or acquisitions. One topic on everyone's mind in Silicon Valley is autonomous driving. In this area, it looked like NVIDIA was moving faster to capture the huge opportunity in next gen transport systems. For instance, the newly unveiled Tesla Model 3 is powered by the NVIDIA DRIVE PX 2 AI computing platform. To counter this, in March Intel announced plans to acquire Mobileye, a developer of machine vision systems for automated driving, in a deal valued at $14.7 billion.  Mobileye, based in Israel, claims to be the leading market position in computer vision for Advanced Driver Assistance Systems (ADAS). Its portfolio includes surround vision, sensor fusion, mapping, and driving policy products. Mobileye's EyeQ chips are already installed in 16 million vehicles as of 2016. Mobileye currently has OEM relationships with GM, VW, Honda, BMW, PSA, Audi, Kia, Nissan, Volvo, Ford, Renault, Chrysler, SAIC and Hyundai. Mobileye reported 2016 revenue of $358 million and gross margin of 76%. For2017, it should bring in more than $1.6 billion in revenue. Intel said its Mobileye acquisition will be completed by the end of the year.

See also