Tuesday, 19 August 2014

Telcos doomed by dumb pipes? Don’t be so sure.

The death of telcos, starved of revenues and relegated to being providers of dumb pipes while over the top providers cream off all the revenues, has been predicted for far longer than the likes of Google have been around.

In the wake of both Telstra and SingTel Optus releasing annual results last week the issue has once again come to the fore, in a lengthy article in the Australian Financial Review last Saturday. It painted a graphic picture of "a future in which humans are making more phone calls, sending more messages and downloading more content than ever before. And yet the big phone companies, such as Telstra, that for more than 100 years have made it happen are reduced to utilities providing little more than a network of 'dumb pipes'."

The message was reinforced a couple of days later with the release by IDC of its Australia Mobile Services 2014–2018 Forecast and Analysis report, accompanied by a press release quoting senior market analyst, Amy Cheah, saying that connectivity is no longer enough to provide revenue growth. "While operators must continue to invest in network capabilities to protect their core revenue they must adapt their strategy to become more like OTTs [overt the top service providers]; to create new streams of revenue growth by creating new business and deliver new customer experiences."

That's another piece of advice that has been repeated ad nauseam for several years. It was accompanied by some more, from IDC research manager, Siow-Meng Soh, who said: "Mobile operators need to form the right partnerships and train their sales force to be able to sell mobile solutions to different verticals instead of just selling connectivity — which is increasingly being commoditised."
He added: "Some of these applications are machine-to-machine (M2M) applications targeted at specific verticals (eg, utilities, manufacturing, and transportation)."

Both these articles, and I have seen many over the years, start from the premise that the pipes are dumb, that value can be added only through what they carry and the issue for the owners and operators of those pipes is that they get some of that value instead of letting the OTT providers have it all.

But what happens when that network ceases to be dumb and becomes smart, very smart indeed? Last week when Telstra and Optus were announcing their results, over in the US, in Las Vegas TMC was staging a conference on software defined networking (SDN) and network functions virtualisation (NFV). On its web site there's a Q&A with conference chairman and TMC CEO, Rich Tehrani. It's well worth a read, here are a few snippets.

He's asked how the specifications for NFV will impact the market. "These specifications describe how carriers can grow their revenue by providing virtualised services to their enterprise customers by placing solutions in the customer's cloud. By providing virtualised customer equipment, carriers will be able to cost-effectively compete with OTT cloud-communications vendors more effectively."

He's asked what the opportunity will be for telcos. "Massive – huge – incredible – I haven't seen an opportunity in the telco ecosystem space this big since 1998 when I launched Internet Telephony Magazine and subsequently watched circuit switched products become legacy thanks to packet-switched. This exact sort of transformation is going to happen again. ... The OTT providers and other competitors know this and are certainly not slowing down their assault."

It's less that two years since a bunch of telcos, including Telstra, presented the first white paper on NFV. Since then progress has been astonishingly rapid. Telstra is already trialing virtualised CPE functionality with Ericsson and yesterday Ericsson announced that would continue to supply optical transmission equipment to Telstra, along with gear from its partner, Ciena. All their press release could talk about - albeit very vaguely - was how the technology would enable Telstra to implement SDN and NFV.

In other words those dumb pipes are about to become very smart indeed, and the telcos will be the brains. Reports of their death are, in all likelihood, greatly exaggerated.


Monday, 18 August 2014

Vodafone Australia going after M2M market

On 10 September Vodafone Australia is holding a press briefing "Machine-to-Machine – Vodafone insights on adoption, growth and future markets" at which it promises to look at "the global [M2M] landscape based on insights from Vodafone’s M2M Adoption Barometer 2014" and share "new Ovum research, commissioned by Vodafone, [that] will more closely examine expectations for the Australian market." This could be the one area where Australia's beleaguered number three mobile player can gain an edge on Telstra and Optus.

Vodafone Group was an early entrant into the global M2M market. In January 2009 it launched a global machine to M2M service platform to help companies deploy and manage large, wireless M2M projects, supported by a team charged with growing the company’s M2M business worldwide The move came from Vodafone Global Enterprise, a business unit set up in 2007 to serve the mobile communications needs of multinationals across multiple countries.

That initiative appears to have paid off. In June Machina Research, a UK based company that claims to be "the world's leading provider of strategic advice on the newly emerging M2M, IoT and big data markets" published its annual review of the M2M operations of the major global telcos, saying that Vodafone had taken the top spot.

The report's author, Matt Hatton said: “Vodafone’s scale, growth and customer wins are testament to its ongoing lead in the sector. If we just look at the numbers, this year it overtook AT&T and Verizon to become the biggest global M2M provider, in terms of SIMs. It was also the fastest growing of the fifteen we studied."

He added: "We’re not just counting numbers of connections to determine current, or future, success. We expect initiatives such as its SOBE product, its plans for licensing the GDSP platform, and the added bonus of funds from Project Spring to provide further impetus in the next few years.”

SOBE (Simple Out-of-Box Experience) is designed to make it easier for users of smart devices to sign a prepaid data contract on the fly. For example, the Amazon Kindle Fire come equipped with Vodafone's SOBE technology. GDSP (Global Data Services Platform) is an online service designed to help Vodafone clients manage all their M2M connections and activate, block and disable devices. Project Spring is a global network investment program.

In a further move into the M2M market, Vodafone Group announced in July that it would expand its long-standing partnership with Japanese telco, NTT Docomo, to include the delivery of M2M services to global enterprises. However the two gave no details of their new collaboration.

Also in July Vodafone announced plans to acquire Italian company Cobra Automotive Technologies, a provider of security and telematics products to the automotive and insurance industries. Vodafone said the move was in line with its strategy to expand its M2M capability beyond connectivity. “Cobra’s telematics products and expertise will enable Vodafone to provide a more comprehensive range of end-to-end services to automotive customers.”

So back to the 10 September event. I guess we'll have to wait until then for Ovum's research on the Australian M2M market but Vodafone's barometer was published in July. The headline is that the percentage of companies saying that they are using M2M has grown by over 80 percent since Vodafone's first study a year ago. Vodafone says: "Over a fifth of companies that we spoke to already have at least one M2M-based solution in place. Fifty five percent of organisations that we talked to said that they expect to have an M2M solution in place within two years."


Vodafone's success in the M2M market will be good news for NetComm Wireless (ASX: NTC). In October 2012 it signed an agreement with Vodafone Global Enterprise to supply M2M modems, saying: "This is an ongoing supply agreement and the size of revenues will directly relate to the success that Vodafone has in securing M2M contracts."

Friday, 15 August 2014

Digital security? No such thing, says Gartner

Good headline eh? A beat up? Don't be so sure. Gartner has just released its 2014 Hype Cycle for Emerging Technologies, claiming that it "maps the journey to digital business." It puts digital security at the very start of that journey and on a slow road to the destination.

This journey is important. All the analysts are saying that businesses must become digital to survive. I spent most of yesterday at Forrester's CIO Summit in Sydney at which that message was driven home time and again.

It's neatly summed up in the introduction to a new Forrester report The State Of Digital Business In Asia Pacific In 2014, due for release later this month. "Regional CIOs must incorporate digital as a core technology imperative. CIOs who ignore the impact of digital disruption do so at their own peril. The only way to weather dynamic industry changes is to incorporate systems that help your organisation win, serve, and retain customers."

I'm sure you're familiar with the Hype Cycle. A new technology starts with an innovation trigger, rises relatively rapidly to the 'peak of inflated expectations' before descending equally rapidly to the 'trough of disillusionment' and then, assuming it survives rises relatively slowly up the 'slope of enlightenment' to the 'plateau of productivity'.

There are close to 50 emerging technologies spread across this hype cycle, each individually coded with Gartner's estimate of how long that technology will take to reach the plateau of productivity.

The one that caught my eye is right at the start of the cycle with a 5 to 10 year time frame to productivity. It's labelled 'digital security'. Surely not? The issues with security are well publicised but it is an established and generally successful technology, if you take it to be a blanket term covering the whole gamut of techniques and technologies used to secure data in the digital world. That at least is how Wikipedia defines it (although that article is flagged as having multiple issues).

Without access to Gartner's full report on the hype cycle it's hard to know exactly what Gartner means by 'digital security' but references in Gartner blogs etc suggest it to be a fairy general umbrella term.

And indeed it seems that Gartner is forecasting the end of security as we know it. In Gartner's top 10 predictions for 2014 there is one slide (slide 13) which predicts that: "By 2020 enterprises and governments will fail to protect 75 percent of sensitive data [and] will declassify and grant broad/public access to it."

It goes on to say: "Enterprises and government should accept that sharing many seemingly sensitive data is neither dangerous or unprofitable, politically and economically." It argues that the growth of data will exceed protective mechanisms and that the best form of protection is having nothing to protect.

Trouble is the same technologies than can be used exfiltrate data can be used to infiltrate and disrupt systems that do much more than store and process data: systems that control things, things like electricity supply, lifts, life support systems etc, etc.

Another of Gartner's 2014 predictions is that "by 2024 at least 10 percent of activities potentially injurious to human life will require mandatory use of a non-overidable smart system."

Non-overidable of course does not mean secure and non-hackable. But hey, Gartner reckons digital security will reach have reached the plateau of productivity by then so we can all relax.

Friday, 8 August 2014

Does this space drive break the law of physics?

There's been quite a bit of coverage on news services these last few days of a new technology that appears to violate the laws of physics and that could revolutionise space travel. I've read a few of these and found the explanations a bit confusing. Discussing it with people over dinner last night also suggested that just there's a bit of confusion as to just why it is so significant. So I thought I would try and explain it in simple terms.

The technology is called the EmDrive. It's been in development by British company SPR Ltd for several years amid much scepticism. It's hit the headlines this month because NASA engineers, apparently, have studied it, confirmed it works as advertised and published their findings in a conference paper. However they have not, reports say, explained exactly how it works, simply that it does work.

This has produced headlines like this one "‘Impossible’ space drive tested by NASA foretells future of deep-space travel"" The story went on to say: "NASA has conducted long-awaited experiments to prove that the fabled space drive, capable of generating its own thrust and breaking a fundamental law of physics, works. If the find survives fresh scrutiny, space ship construction will be revolutionised."

So, why is this so exciting and why does it appear to violate the laws of physics? Here's the problem. You're sat in your spaceship, stationary out in space. The laws of physics say that if you want to move forward you have to chuck something out the back end. The more you can throw and the faster you can throw it the faster you will go forward.

You can have an infinite supply of energy with which to throw stuff, but if you ain't got anything to throw you're going nowhere. This creates another problem. The mass you want to throw out the back later has to move forward with you initially. So the further and faster you want to go the more mass you need to carry and the more you need to throw out the back (or the faster you need to throw it) to even get started. It's a case of diminishing returns.

This is why you need a massive space rocket to launch a small satellite. (In this case the source of energy, the rocket fuel, also provides the mass that gets shoved out the back to move the rocket forward.)

This state of affairs is formally expressed as the law of conservation of momentum. It states that, in a closed system such as a spaceship in space, the product of mass x velocity remains constant. Put simply if you want to get a mass of 1kg moving one way at 10km/minute you need 1kg moving in the opposite direction at 10km/minute, or 100gms at 100k/minute. And that 100gms must come from the spaceship.

What the EmDrive claims to have done, according to its web site, is to "convert electrical energy directly into thrust" that is move the object forward without throwing anything out the back end. "Thrust is produced by the amplification of the radiation pressure of an electromagnetic wave propagated through a resonant waveguide assembly."
At first blush this would seem to be in conflict with the law of conservation of momentum. SPR claims this is not the case. According to the FAQs on its web site "The EmDrive does not violate any known law of physics. ... The EmDrive cannot violate the conservation of momentum. The electromagnetic wave momentum is built up in the resonating cavity, and is transferred to the end walls upon reflection. The momentum gained by the EmDrive plus the momentum lost by the electromagnetic wave equals zero. The direction and acceleration that is measured, when the EmDrive is tested on a dynamic test rig, comply with Newton's laws and confirm that the law of conservation of momentum is satisfied.
I have no idea what that means. But I guess eventually this will turn out to be either the breakthrough of the century, or the con of the century.


Thursday, 7 August 2014

The soon-to-be-global smartphone player you've never heard of...

...And almost certainly can't pronounce. Its name is Xiaomi. It is only four years old and has just, according to market analyst Canalys, soared to the top of China's smartphone market, which in Q2 of 214 accounted for 37 percent of global smartphones shipments, some 108.5 million units according to Canalys.

"In little over a year, Xiaomi has risen from being a niche player to become the leading smart phone vendor in the world’s largest market, overtaking Samsung in volume terms in Q2," Canalys said.

"Xiaomi took a 14 percent share in China, on the back of 240 percent year-on-year growth. With Lenovo, Yulong, Huawei, BBK, ZTE, OPPO and K-Touch, the eight Chinese vendors in the top 10 together accounted for a total of 70.7 million units and a 65 percent market share."

Ninety seven percent of Xiaomi's sales were in Mainland China, but that's unlikely to be the case for very long. According to Canalys Xiaomi is now looking to expand into other markets, with Indonesia, Mexico, Russia, Thailand and Turkey in its sights for the second half of the year.

Canalys analyst Jessica Kwee says Xiaomi faces considerable challenges in trying to crack the global market, and she has some advice for the company: "Xiaomi needs to build its international brand, and will need to localise its services offering with MIUI [its version of Android] for the different markets into which it expands, else its differentiation, value proposition and service-oriented revenue streams will be eroded. And it must tailor its marketing and largely online sales channels accordingly."

She concludes: "Xiaomi does have the potential to be a disruptive force beyond China and international vendors should take note." I'd say that is a very significant understatement.

Xiaomi's MIUI android-based OS is already gathering a global following with, 26 global 'fan sites' including one for Australia. You can download and install the MIUI OS on numerous Android phones. There was also at one time a Xiaomi store in Australia importing selling the company's products, but not connected with Xiaomi. However it has closed down.

Rival forecaster IDC seems to see the threat to other smartphone vendors as being far greater than Canalys. It has just released its Worldwide Quarterly Mobile Phone Tracker for Q2 of 2014. The top five vendors from one to five were Samsung, Apple, Huawei, Lenovo and LG. But program director Ryan Reith doesn't expect that ranking to remain. "Right now we have more than a dozen vendors that are capable of landing in the top five next quarter," he says. "A handful of these companies are currently operating in a single country, but no one should mistake that for complacency – they all recognise the opportunity that lies outside their home turf."

And to ram home that message, senior research manager, Melissa Chau, adds: "As the death of the feature phone approaches more rapidly than before, it is the Chinese vendors that are ready to usher emerging market consumers into smartphones. The offer of smartphones at a much better value than the top global players but with a stronger build quality and larger scale than local competitors gives these vendors a precarious competitive advantage."

More to the point, Canalys and IDC are at odds on the global smartphone vendor rankings. According to Canalys Xiaomi had 14 percent of a total of 108.5 million smartphone shipments in China in Q2, which works out to be 15.75 million units. According to IDC. Number three global vendor Huawei shipped 20.3 million units in Q2; number four vendor Lenovo 15.8 million and number five vendor LG 14.5 million. So, depending on whose figures are correct Xiaomi could already be the number four global smartphone vendor.


I'd say all those vendors should do more than take note of Xiaomi. They should be very afraid.

Wednesday, 6 August 2014

The data centre of the future could be built from smartphones

There's an article just published in Wired magazine, titled: "The Data Centers of Tomorrow Will Use the Same Tech Our Phones Do." The data centre of the future, it says, "is about eliminating all vestiges of the proprietary hardware used in networking and storage in favour of commodity components available through the mobile supply chain. It’s about this commodity hardware performing the function of proprietary systems today."

According to Wired, the data centre of the future will be "a bunch of dirt cheap, cell-phone-like machines—all connected together with sophisticated software—instead of those power-sucking, refrigerator-sized boxes."

This indeed is the next logical step beyond the architecture espoused by Cloudera, which has its origins in Google: masses of commodity 'pizza boxes' comprising CPUs, memory and storage, where redundancy and data replication, managed by software, replace reliability, and which I wrote about last week. Cloudera CTO and founder Amr Awadallah claims that costs of storage in its technology can be as little as 100th those of conventional systems.

The vast majority of CPU chips in data centre servers are supplied by Intel, but as equipment densities in data centres increase, power and cooling become bigger issues. Much of the demand for capacity in data centres is coming from those mobile devices: many of the billions of apps on those billions of devices rely heavily on resources in the cloud.

Nowhere are power issues more critical than in smartphones: where the ever increasing demands of applications and higher speed communications are pushing the limits of battery and semiconductor technologies.

And who holds the lion's share of the smartphone processor market? Not Intel, but ARM. In 2013 it claimed to have a 90 percent share in smartphones; a 95 percent share in feature and voice phones; and, a 50 percent share in mobile computing devices including tablets, net books and laptops.

Some analysts think that Intel will start to make inroads into the mobile device market. Morningstar senior analyst Andy Ng wrote in June "While Intel has had limited success in penetrating the smartphone and tablet processor market so far, it has only begun to use its manufacturing technology advantage in its Atom product line. Intel's new Silvermont Atom chips are the first Atoms manufactured using cutting-edge process technology, as the firm used older-generation technologies for prior Atom products. As a result, we think Intel will have opportunities to achieve some success in mobile device processors as it fully harnesses its moat in that market."


Intel will be working very hard to achieve that success before its dominance of the data centre market is threatened by the evolution of the data centre into "a bunch of dirt cheap, cell-phone-like machines—all connected together with sophisticated software—instead of those power-sucking, refrigerator-sized boxes," or even instead of low cost Intel-powered pizza boxes.

Tuesday, 5 August 2014

Speed bumps on the road to digital transformation

Cloudera founder and CTO Amr Awadallah, account of the challenges it and its customers face as they strive to exploit the full potential of the Cloudera Enterprise Data Hub technology is yet another example of the problems enterprises face as they embark on journey of 'digital transformation'.


When I interviewed Awadallah last week he showed me a slide depicting how Cloudera sees customers progressively embracing and extending their use of the Cloudera Enterprise Data Hub technology. "The journey starts from using the system for better operational efficiency, just for cheap storage," he explained. "Then they start to use it for what we call 'extract, transform and load': transforming the data from its unstructured form into a structured form that they can use inside their database. That's a big bottleneck today for a lot of organisations."

This is followed by data warehouse optimisation. Awadallah claims that the cost of storage in conventional data warehouse technology means that much data is relegated to the archives and for all practical purposes lost. "The archive is the graveyard of data, he says.

Based on the claim that the Cloudera approach offers storage at 1/30th to 1/100th cost of a data warehouse, he quips: "We offer economy class storage. Data Warehouse is first class. With a data warehouse it's fly or die."

Once organisations have achieved this level of usage, he says, they can start to exploit the real potential of the technology, analysing both structured and unstructured data, applying sophisticated data science techniques and finally applying converged analytics. This, he says "is when you have achieved enlightenment as an organisation. Where you have a single place with all your data and your workloads all come to the data as opposed to the data going to the workloads."

He claims that this is typically a four year journey for an organisation. "It can be a ten-year journey and for some organisations, it can be a one-year journey." Cloudera is only six years old, so none of its customers has yet completed the journey, but many are at the 'data science' stage, Awadallah says.

Not the least of the challenges for an organisation is how it makes the transition from the technology being in the domain of IT to the domain of lines of business - a transition shown in Awadallah's slide as a dotted line between enterprise 'data warehouse optimisation' and 'agile exploration'.

I asked him how Cloudera gets its customers across that line. How they make the transition from Cloudera technology being the domain of IT to the domain of lines of business.

The answer, it seems is not IT pushing the business along the journey, telling them how they can make better use of the technology. Nor is it the lines of business pressuring IT to provide the additional functionality. "IT does not know how to do the sale. In fact IT is afraid of this as there are new skills they have to learn," he says. "At a minimum it forces them to do something new. IT sometimes fights this change."

This places Cloudera in an awkward position: from having an established relationship with IT, it moves to selling the benefits of its technology to the business, potentially undermining established relationships. "We try to make friends with the IT guys" Awadallah say. "We give them training so they are less resistant."