Sep 232016
 
weta_digital

Dealing with the massive wave of data flowing into businesses will be one of the defining management issues of the next decade. One company that is already dealing with this is New Zealand’s Weta Digital.

Wellington based Weta that’s best known for its work on Lord of the Rings and is part owned by director Peter Jackson employs 1400 staff for its movie special effects work and has won five visual effects Academy Awards over its 23 years of operations.

Kathy Gruzas, WETA Digital’s CIO, spoke to Decoding the new Economy at the Oracle OpenWorld forum in San Francisco this week about some of the challenges in dealing with the massive amount of data generated by the movie effects industry.

“We have some very heavy loads.” Kathy states. “We push our systems to the limit.”

Applying powerful systems

One challenge is the sheer computing power required, ‘the render frame processes one frame per server until you have four seconds of footage. Sometimes that takes over night or even longer and for that we use a lot of storage,” Kathy says. “The render farm being six thousand servers will write 60 to 100 terabytes of data a day and read a quarter to half a petabyte each day.”

“We need systems that will be very large to handle the volume of data we generate but also be very quick to handle those read and writes.”

“One render could use a thousand computers, sometimes more, and all of those will be reading and writing against the same block of storage so we have our own software layer that directs those loads but we try to minimise the load on our storage but we have the worst work load you can imagine with lots of servers, lots of small reads and writes and many of them random and concurrent with pockets of hot files.”

Despite the automation, the business is still extremely capital intensive. “In visual effects you probably need at least three hundred artists to work on one film, it’s a very labour intensive process to do the artistry and much like a production line.”

Going mobile

The nature of modern movie production means the effects teams are now part of the shoot which adds another level of complexity for Weta. “Although we are visual effects which is largely post-production we do go out with crews when they’re shooting the movie so we can do reference photography,” says Kathy.

“We do 3D scans so if we need to do something digitally and we do motion and facial capture as well,” she says. “There are 240 muscles that we tweak individually to get the expression. That’s a huge amount of data to capture.”

To do this, Weta created their own ‘road case’ that contains everything they need to grab the shots and store the data they need, “you can’t ask the director retake the shot because we missed something.”

Into the forest

“We have to take the case into the forest and into the rain and everywhere. It’s good having that roadcase that has storage, networking and servers in it.” The case, which was self assembled by Weta’s team is “probably the most travelled Oracle system on the planet,” laughs Kathy with “lots of data capture and sub-rendering.”

Weta’s story illustrates just how managing data is becoming a critical issue for companies. While movie special effects is very much a specialised field that’s far ahead of the curve in its technology use than most businesses, they do show the importance of managing and securing their data.

For other businesses, lessons from Weta is understanding your company’s – including staff and customers’ – needs then investing in the right tools to deliver is essential.

One important difference between technology intensive businesses like Weta and most other organisations is the New Zealand company is doing most of its processing and storage in house. Those without the same needs will almost certainly be shifting these tasks onto the cloud.

Sep 212016
 
mark_hurd_oracle_ceo

Oracle CEO Mark Hurd’s keynote at the company’s Open World conference in San Francisco yesterday illustrated a problem facing businesses around the world and its effects on enterprise software vendors like the one he heads.

“Standard and Poor’s top five hundred companies’ revenue growth is at one percent, their earnings growth is five percent.” “It means what? Expenses are going down.”

“This is the problem that the CEO has,” he says. “Why is it hard to grow revenue. All your investors want you to grow earnings and deliver growth. They have little patience for any long story about why it’s so hard.”

“They don’t care about any issues you may have. Grow earnings, grow cash flow, grow stock price. That’s it.”

Growing in a slow market

As a result of that the easiest way to grow earnings is to grow revenues but when global GDP and markets are flat, the only way to grow is to gain market share, Hurd says. “We have to know the customer better, we have to do a better job of marketing and we have to do a better job of aligning our goods and services to what our customers want. We have to improve our products and processes.”

That imperative for companies to cut their operating costs has had a brutal effect on enterprise IT budgets, “over the past five years, the growth in enterprise IT has been flat.” Hurd says, “the growth in spending has been basically zero.”

Customers drive the market

Like many things in the tech industry, the sector’s growth focus has shifted to consumers, “consumer spending on IT has almost quadrupled in the past decade. So while companies are sort of flat, consumers have been spending like crazy.” Hurd observes, “consumers are more sophisticated, more capable, more knowledgeable and expect better services than ever before.”

“Your customer experience is not being defined by your competitors but by technology fuelled consumers. For instance, AirBnB may be defining customer experience for the hospitality industry.”

“People are using a lot of social technologies in their personal lives,” “we expect ease of use, simplicity, clean interfaces are now things we expect in the enterprise side.”

Crimping innovation

In the enterprise IT sector, Hurd believes the flat market means many companies catering to the corporate market are skimping on Research and Development which in turn is crimping innovation, a factor compounded by cloud providers taking an increasingly larger share of the market.

This is underscored by cloud leader Amazon Web Services spending over ten billion dollars a year on R&D. Hurd’s boast that Oracle is spending half of that shows how the legacy players are struggling.

What stands out in Hurd’s keynote is how legacy providers see cloud computing as their salvation. However Amazon’s dominance in that space is a major obstacle for them.

For consumers, big and small, the shift to the cloud has been a good thing in shaking up the existing industry and making new technologies more accessible to smaller customers. For existing businesses like Oracle, there’s a challenge in adapting to a lower margin, commoditized and quickly changing market.

A bigger question though facing all large corporations, not just software companies, is this new normal of low economic growth. Succeeding in that environment is going require a completely different management and investor mind set to that of the last seventy years.

Sep 182016
 
commonwealth-bank-in-lockart-nsw

What would happen if the world’s richest people invested in startup businesses? Bloomberg Business ran an interesting, if flawed, thought experiment looking at how many nascent companies each country’s richest individuals could invest in.

It’s surprising how low those numbers are and, if anything, the result underscore how the 1980s and 90s banking sector ‘reforms’ caused the world’s financial system to pivot from its historical purpose of funding commercial enterprises into speculation, rent seeking and manipulating markets.

Apart from a smattering of venture capital not much has replaced the banks in funding the SME and entrepreneurial sectors, if anything it has been those ultra high net wealth individuals who have been financing the investment funds providing capital to entrepreneurs.

How the finance industry evolves in the face of the fintech boom and a world that’s slowly becoming less indulgent of the industry’s greed will be one of the defining things of next decade’s business environment. For the small business and startup sectors getting the funding right will also be a key factor.

The biggest question though is job creation, being able to fund new and innovative investments will be one a critical concern for societies dealing with the effects of an increasingly automated economy.

Sep 132016
 
google-data-center-mayescounty-employee2

One of the little discussed reasons for the US tech industry’s success is the role of military and intelligence spending by the government. Not only are various agencies funding research and enthusiastically buy technology, they are also being strategic investors in many companies.

In Sydney last week Dawn Meyerriecks, the CIA’s Deputy Director for Science and Technology, gave an interesting insight into the agency’s investment philosophies at the SINET61 conference.

The conference was aimed at drumming up interest in the technology security industry along with showcasing the connections between Australia’s Data61 venture and the US based Security Innovation Network (SINET).

SINET itself is closely linked to the United States’ security agencies with chairman and founder Robert Rodriguez being a former US Secret Service agent prior to his move into security consulting, venture capitalism and network-building.

Compounding the organisation’s spook credentials are its support from the US Department of Homeland Security along with the UK’s Government Communications Headquarters (GCHQ), so it was barely surprising the Australian conference was able to attract a senior Central Intelligence Agency officer.

Investing in flat times

“Flat is the new up,” says  Meyerriecks in describing the current investment climate of thin returns. In that environment, fund managers are looking for good investments and the imprimatur of the CIA’s investment arm, In-Q-Tel, is proving to be a good indicator that a business is likely to realise good returns.

“If you can predict a market – and we are good predictors of markets – then the return on investment is huge,” she says.

“In-Q-Tel really leverages capital funding for good ideas. We get a twelve for one return, for every dollar we put in it’s matched by twelve dollars in venture capital in emerging technologies.”

Attracting investors

For the companies In-Q-Tel invests in along with those that supply technology to the organization, the CIA encourages them to seek private sector investors.

“What we’re telling our supply chains is you go ahead and tap into the capital markets,” Meyerriecks says. “If you can turn that into a commercially viable product then will will ride the way with the rest of the industry because it’s good for us, it’s good for the country and it’s good for the planet.”

Adding to the CIA’s attractions as a startup investor are the opportunities for lucrative acquisition exits for the founders, she believes. “Not only are we using that venture capital approach for emerging technologies but our big suppliers are sitting on a ton of cash.”

Diversity as an asset

Another lesson that Meyerriecks believes will help the planet, and the tech industry, is diversity. “Globalisation has show isolationism doesn’t work,” she says.

“Back in the day when I was a young engineer the best way to make sure your system was resilient was to harden its perimeters. the best ways to be ‘cyber resilient in the old days was by drawing the barriers to keep the bad guys out.”

“The best way to be cyber-resilient in the old days was to draw big boundaries around yourself to keep the bad guys out. The latest studies look at other things because you want to be resilient, you want high availability.”

Now, system diversity is seen as an asset.“Biologically the three factors that contribute to resilience are the ability to adapt, the ability to recovery and diversity,”  Meyerriecks says. “We look to deliver high availability among components that may not themselves have high reliability.”

The future of investment

“I think we’ll see commercialisation still driving investment for applied R&D in particular,”Meyerriecks said in a later panel on where the agency is looking at putting its money.

“The big game changers will be around the edge, taking SDN (Software Defined Networking) to its logical extreme giving everyone their own personal networks, not just in data centres but at the edge of the network.”

“I think there’s lots of things that the commercial industrialisation of the technology and physical system are going to force us to grapple with on many levels.”

Risks in managing identity

An interesting aspect of Meyerriecks’ talks at SINET61 was her take on some of the technology issues facing consumers and citizens, particularly in the idea for individuals having their own personalised network.

“This opens up a whole range of things, ” she suggests. “Do I eventually not just be an IMSI or EIMI (the mobile telephone identifiers) but do I become an advertising node, does that become my unique ID? Do I a become a gaming avatar?”

“Then we get into the whole Big Data area. Computational anonymity is a phrase we use. At some point people start saying ‘this is crossing the line’ – it crosses the ‘ooooh’ factor.”

Changing Cybersecurity

“I think the definition of cybersecurity will be expanded to much more beyond wheat we’ve classically thought about in the past.”

Meyerriecks’ presentation and later panel appearance was a fascinating glimpse into the commercial imperatives of the United States’ intelligence community along with flagging some of the areas which concern its members as citizens and technology users.

The US security community’s role in the development of the nation’s tech sector shouldn’t be understated and Meyerriecks’ observation that private sector investors tend to follow the CIA’s investment path underscores their continued critical role.

Sep 092016
 
Meg Whitman (center), President and Chief Executive Officer of Hewlett Packard Enterprise, rings the opening bell of the New York Stock Exchange on November 2, 2015. CREDIT: Eric Draper

Over this week I’ve been posting a series of interviews with the candidates for this week’s Sydney Lord Mayoral election. All of the teams have interesting schemes and ideas on how they can improve the city’s profile as a global tech centre.

While each team’s plans are worthy, it’s worth asking exactly what governments can do to make their communities more attractive to businesses and whether short term subsidies and incentives can help.

There is some evidence they can, prior to San Francisco changing its tax rules the city took second place to Silicon Valley in the southern Bay Area. In the last ten years, the city has become the focal point for the tech industry.

However there is a counter argument that San Francisco benefited on a generational shift of lifestyle preferences away from the leafy suburban lifestyles of Palo Alto and San Jose to the grungy but walkable communities of the Mission and SOMA.

The Bay Area though is a special case, Silicon Valley’s success as a tech hub is based upon massive Cold War tech spending that drove the region’s industry and its that high level support that probably tells us more about government support.

In the case of London and Singapore, the successes have been due to the national governments putting in broader economic reforms and incentives. Also their proximity to Europe and East Asia respectively has made both cities attractive.

On balance it’s those broader economic factors that make regions attractive as industries clusters – local incentives count little compared to access to factors like markets, capital and skilled labour. Taxation is, at best, a secondary issue.

The biggest challenge for Sydney, and most Australian cities, is the the crippling cost of property. In 2013, staff.com released a survey showing Sydney to be second only to Zurich in the cost of establishing a startup.

In many respects, the cost of property doesn’t really matter to prosperous industry hubs – San Francisco, London, Singapore and New York are all eye wateringly expensive and yet they still thrive – however all of those cities have better access to capital and markets, if not labour, than Sydney.

Addressing Sydney’s chronic shortage of affordable accomodation is firmly in the state and Federal governments’ remit and beyond giving property developers a green light to build high rise apartments neither level of government has shown any interest in addressing it.

Similarly, the tax structures which penalise Australian employees of high growth businesses and dissuade investment in early stage ventures are totally the responsibility of the Federal government and it’s hard to see that changing in the term of the current dysfunctional administration.

The relative powerlessness of local governments leaves initiatives by the City of Sydney limited in scope and schemes to promote the city or offer incubator space are peripheral to the factors that encourage the development of a global industrial centre.

Ultimately though, the question has to be how much any government can do to create a Silicon Valley, factors such as labour availability and access to capital come down as much to the community’s attitudes and business’ risk tolerances.

So perhaps we focus on what governments can do for business. Maybe just providing a level playing field can be the best we can hope for.

Aug 302016
 
we need to treat chinese markets carefully

At the recend Autodesk University event in Sydney I had the opportunity to talk with Pat Williams, the company’s senior vice president for Asia Pacific.

Williams’ beat covers all of Asia and he’s based out of Shanghai where he’s been based for the last eight years and prior to that he spent a decade in Japan.

Having spent so much time in North East Asia, and heading to the PRC the following week myself, it was interesting to hear Williams’ views on how industry is changing in China and ther country’s attitude to American software companies.

“There’s a lot of noise that gets made in China about their local IP and the local vendors and what I say is ‘the Chinese companies are competing in a global market and they are under the same competitive pressures as everybody else in the world so when they find a better tool they use it. Despite all the noise, business is quite good there.”

For the Chinese economy, the aging and increasingly expensive workforce presents a problem, something addressed by the China Manufacturing 2025 plan which sees the country increasingly competing in high tech sectors such as aerospace, telecommunications and biotech fields.

“China’s kind of an anomaly,” says Williams of the country’s immense growth rates. “From a government perspective there’s a lot of horsepower behind the things that they do – China 2025, their manufacturing initiative, you’ve got what they’ve been doing with Building Information Modelling (BIM) and our architectural tools.”

They’ve really kind of spearheaded what we’ve been talking about on things like 3D printing of houses. China on its own is just this mushroom that’s happening.”

While the industrial shift in China and the rest of Asia is promising opportunities to companies like Autodesk, that change is affecting their workforce as well with the company announcing plans to lay off ten percent of their workforce earlier this year.

Those cutbacks are part of the adjustment to a new market reality says Williams, “it was part of right sizing the business.” He observed “we realised our margins were going to be compressed as we move to a subscription model.”

Autodesk’s shifts illustrate how the opportunities in the new economy don’t come without costs even for the companies that seem to be winners in a shifting marketplace.

In China, American companies are finding they have to a unique proposition – companies like Apple and Autodesk are good examples – and as the country moves its economy further up the value chain all foreign businesses are going to have to show how they add value.

Succeeding in a changing economy isn’t without uncertainty. And it certainly isn’t without risks.

Aug 292016
 
Australia National Broadband Network CEO Bill Morrow

One of the most stunning examples of Australia’s uncompetitive, post-mining boom economy is its National Broadband Network.

Announced in 2009 to provide high speed data access to the nation to address the effects of thirty years of poor decisions and poorly thought out policies by successive governments, the project was intended to upgrade the telecommunications network and break the near monopoly of the incumbent telco, Telstra.

Sadly the project quickly foundered as the managers of the company set up to build the network made a series of poor decisions that stemmed from their underestimating of the project’s scope and their arrogant hubris in rejecting the advice of those who did.

To compound the problem, the project was politicised by the intellectually lazy and opportunistic Liberal opposition who promised they could build it for less by utilising existing telephone and Pay-TV infrastructure. On becoming government, the then communications minister and now Prime Minister changed the scope to do that and promised a quicker and cheaper rollout.

Last Friday, the folly of the Liberal Party’s plans were shown when the National Broadband Network company, nbn™, issued their updated business plan that detailed a further retreat from both the original project scope and the government’s promises.

The Melbourne Age’s Lucy Battersby illustrated how completely Malcolm Turnbull and the Liberal Party bungled their costings, showing just how mediocre and dishonest the government and Prime Minister have been in estimating the cost of the project.

However, NBN Co underestimated the cost of using existing hybrid-fibre coaxial [HFC] cables laid by Telstra and Optus in the 1990s. Last year it calculated an average cost of $1800 per house. But detailed field work discovered the cost was actually $2300.

In 2013 the Coalition estimated FTTN connections would cost about $900 per premise and this was raised to $1997 in a 2014 strategic review, and raised again in 2015 to about $2300.

In the real world, being out by nearly 300% would cost an estimator or executive their job and for a small business could well see them being put out of business, but in the carnival of mediocrity that marks modern Australian politics, those responsible for such mistakes only thrive, as do the managers of nbn™ who recently awarded themselves fat bonuses.

Adding insult to injury for the long suffering Australian taxpayers and broadband users is that the nbn™’s management have revised the scope again to overcome increased costs and now only 21% of consumers will get a fibre connection as opposed to the 40% claimed when the new government changed the scope.

Those scope changes beg the question why anyone bothered in the first place. Had the network been left with Telstra there’s a reasonable chance 20% of customers would have ended up on fibre by early next decade as the economics of maintaining and installing the technology overtook the older copper system.

Probably the biggest insult though to Australian customers though are the desperate attempts to make the new network profitable with plans to gouge the nation’s telco users as Fairfax’s Elizabeth Knight reported.

Data use per user is anticipated to grow at a compound rate of 30 per cent per cent to 2020.

At first blush these increases in usage might look exaggerated – but wait. Only last year NBN was working off the expectation that this year its existing customers would consume 90 gigabytes per month. But the current rate of consumption is actually 131 gigabytes per month – and rising.

Thus as the years progress towards 2020, NBN not only gets an increase in customers, it get an increase in revenue per customer .Monthly average revenue per user is forecast to increase from $43 this year to $52 in 2020..

 

So Australians will be expected pay more for their substandard connections to help an organisation that has consistently failed to meet its promises and targets. It should also be noted that rising Average Revenue Per User (ARPU) is the opposite of what’s been happening in the real world over the last twenty years as revenues, and profits have fallen.

To be fair, it’s not just Australia that has struggled with rolling out fibre networks. In the US, Google Fiber is going through blood letting and scope changes as the company struggles to meet targets and keep costs under control. That same experience has been repeated around the world.

However when it comes to missed targets, broken promises and the sheer scale of money wasted, Australia’s National Broadband Network dwarfs them all.

Australian taxpayers, voters and telecommunications users should be asking hard questions of their political leaders