UK: Technology Predictions 2010 (Part 1)

Foreword

Welcome to the 2010 edition of Technology Predictions. This is the ninth year in which the Deloitte Touche Tohmatsu (DTT) Global Technology, Media & Telecommunications (TMT) Industry Group has published predictions on the TMT sectors.

This year's report has three principal themes: the economy, digitization, and CleanTech.

First, the global economy: if there was one advantage to making predictions for 2009, it was the consensus that most major economies were expected to fall into recession. (They did, with a few notable exceptions, such as Brazil, India, and China.) In 2010, the picture is far more mixed. While it is generally agreed that most economies should recover, there do not appear to be enough shapes or letters available to describe the possible permutations that recovery may take — will it be a U, a V, a W, or a square-root recovery? And the possibility remains of a double-dip recession, once the stimulus ends. At the time of writing, governments appeared bullish, corporations more bearish, and economists divergent.

The uncertain economic outlook is likely to keep enterprises focused on cost. This emphasis on cost is a key driver of three of our Predictions: the return of thin client computing to enterprise computing, the adoption of employee-driven technology procurement (with cost savings as a key factor), and the keen interest in, but moderate adoption of, cloud computing.

Digitization's momentum precedes the current recession by decades, but it continues to reinvent the global technology sector. This fundamentally simple transition — the conversion of analog data into digital form and its distribution via digital networks — not only changes the balance of power within the industry but can also reset the scope of other sectors. Anticipating digitization's impact is commendable, but timing is everything and preparation is paramount. Being caught off guard by the pace of digitization has been — and will continue to be — catastrophic.

Companies that have launched netTabs — devices that fit between smartphones and netbooks — over the last two decades are all too aware of this. Our view is that 2010 should, however, be a key year for the device. Moore's Law, one of the key yardsticks of digitization's progress, is expected to remain in effect in 2010, but with a focus on size, price, and cost, rather than power.

In 2010, we predict a recovery for CleanTech, following a battering by a storm of recession-induced pragmatism in 2009. But the return to growth is unlikely to be uniform. One nascent sector that may see major progress this year is carbon-negative cement. Given cement's emissions (5 percent of the global total), any advances in green cements may be significant.

I am often asked about Predictions' track record. We are never likely to be 100 percent right. However, a focus on pragmatism and an aversion to hype has meant that we are more often right than wrong. We never include a prediction only because it will come true in the next year. Rather, our focus is on identifying potential "black swans" whose impact could have major, strategic ramifications for companies in the sector.

As a result, each prediction is designed to start or stoke a further conversation — not to stop it. And we trust that the Predictions' launch, expected to take place in over 50 cities around the world in 2010, reaching over 5,000 industry executives, serves precisely this purpose.

I wish you every success for 2010 and beyond.

Smaller than a netbook and bigger than a smartphone — net tablets arrive

DTT TMT predicts that in 2010–2011, tens of millions of connected portable devices will likely be purchased by consumers with sore eyes and sore arms. Net tablets, or netTabs, will be based on a new form factor and feature significant processing capacity.¹ They will aim to offer an appealing balance of form and function. Priced between $400 and $800, they are likely to weigh less than 1.1 pounds (500 grams) and measure about 8 in by 5 in (20 cm by 12 cm by 2.5 cm). They are expected to include cellular and WiFi access, full-color touch screens, and well-populated app stores.

NetTabs are expected to meet specific consumer needs compared to smartphones on the one hand (which are still a bit small for watching videos or even Web browsing), and notebooks, netbooks, and ultra-thin PCs, on the other (which are too big, heavy, or expensive).

NetTabs, the Goldilocks of devices (not too big, not too small), may be "just right" for many users.

This is counter to consensus expectations. The segment lying between the smallest available laptop and largest phone has historically been a technological dead zone. Recent efforts from successful PC manufacturers have met with limited consumer acceptance at best.² As a result, many observers have concluded that there is no current consumer demand for tablet computers — nor will there ever be.³

We believe that previous attempts at the tablet form factor failed (or were less than resounding successes) for many reasons: the graphics, software, and user interface were underwhelming, not well connected to cellular or WiFi networks, and, most important, they have largely served as work-oriented data-entry devices. By contrast, an easy-to-use, consumer-focused device used primarily for media and Web browsing is much more likely to be widely accepted by the market.

The next 18 months are likely to see a proliferation of these devices from two sources: tablets that were designed to be tablets from day one and stand-alone single-purpose devices that will be repurposed as multifunction devices by adding enhanced processing and software.

Although none have been launched yet, leaked information suggests that custom-designed tablets are likely to be released by start-ups, some existing successful phone and PC makers, netbook leaders, and various smaller manufacturers using open-source phone operating systems.⁴

While it is difficult to forecast sales of devices whose specifications are unknown, some analysts estimate 12-month sales from launch of over $1 billion. This is larger than global sales of personal navigation devices (PND).⁵ Some industry surveys have shown that consumers are willing to consider the new devices — even at relatively high price points. As many as 40 percent of respondents in one survey said they would be willing to pay more than $800 for one model of netTab that has been the subject of much speculation.⁶

That said, there are tens of millions of devices that have generated billions of dollars in sales that already have something close to the form factor of potential netTab units.

These are popular consumer and enterprise-friendly devices with strong brand names.⁷ They include global positioning system (GPS) units, eReaders, portable entertainment devices, gaming devices, and various handheld computers in different vertical sectors. In many cases, the multifunctional smartphone and netbook, with their wide array of software applications, are finding their targets among customers who currently own one or more of the single-function devices.⁸

For example, the current PND market is roughly 40 million units and a billion dollars in annual revenues.⁹ However, competition from GPS services on mobile handsets is expected to disrupt this market extensively, and analysts forecast standalone PNDs will go from 90 percent market share today to less than 30 percent by 2013.¹⁰

It would be difficult for PND manufacturers to respond to the netTab threat by offering more precise location functionality in the consumer space. Centimeter-level or inch level positioning accuracy is useful for the military and agriculture industries, but not for the average user. One PND manufacturer has already introduced its own smartphone,¹¹ and it is likely that larger multifunction devices (with larger screens for maps), such as netTabs, are sure to follow.

Other manufacturers of various tablet-sized but single-function "dumb" devices are likely to add more intelligence to their products through ARM-based or x86 processors like the Atom. These multifunction devices would offer Web browsing, support email reading, and almost certainly support widgets, applets, and the various stores for purchasing them.

Moore's Law is alive and well in 2010

DTT TMT predicts that in 2010 Moore's Law¹² will probably continue to work just fine. The traditional ability of the global semiconductor industry to double the number of transistors in a square centimeter of silicon every 18 to 24 months will not likely come to a screeching halt. In fact, it may not even slow down. But that increased density is unlikely to be used to produce larger or more computationally powerful chips. Instead, it could permit the production of "good enough" chips that use less electricity, cost less money, or are smaller.

The current consensus on Moore's Law forecasts a gloomier scenario, but historically, it has proven unwise to foretell the end of the Law. Roughly every decade since the Law's declaration in 1965, various pundits have proclaimed that "the immutable laws of physics" would force the rate of chip-manufacturing progress to slow or even stop, only to see the industry continue its advance. But the past year or two has seen an uptick in the dire predictions.¹³

Those anticipating the end of Moore's Law have cited various pieces of evidence. Some of the largest chip manufacturers have publicly stated that "scaling doesn't scale" the way it once did.¹⁴ That is, the costs of doubling transistor density have certainly risen faster than in the past.¹⁵

Most damningly, many have noticed that the "performance" (as measured by clock speed in GHz and number of cores) of the various central processing units (CPU) and other chips has almost leveled off.¹⁶

The integrated circuit has permitted computing devices to become much smaller, more powerful, cheaper, and more efficient than the machines that came before them. And for years, the semiconductor industry marketed its generational advances primarily in terms of processing power, specifically clock speed. Up until 2005, the flagship CPUs found inside top-end PCs and servers ran about 10 times faster, and cost roughly the same ($500), but used more than 10 times the electricity than similar devices from a decade earlier.¹⁷

The increased power usage has not, until recently, slowed sales for servers or desktops.¹⁸ But now those markets are relatively saturated, and growth rates are low. Meanwhile, the new growth markets for chips are being driven by mobile computing, smartphones, and other consumer devices. And in those markets, the primary requirements for each generation of chips are cost and power reduction. Speed, on the other hand, is generally now considered sufficient for most current user requirements.¹⁹ In fact, the most successful recent CPU chips have been those targeted at the netbook and smartphone markets, and both are focused on price and power rather than speed.

Moore's Law has been misunderstood for about as long as it has been around. The popular misinterpretation is that processors become twice as powerful every two years. But the original formulation of the Law, and subsequent clarifications by Gordon Moore, never once mention speed or power. The Law simply refers to technology and engineering advances enabling greater transistor density. Greater density — if one is relatively indifferent to considerations of cost or electrical consumption — can translate into more computing speed. But it can also translate into lower prices or lower power usage. From the perspective of a designer, Moore's Law allows a chip to be optimized for any one of performance, price, or power — but usually not all three simultaneously.

With current growth in mobile computers coming in the form of lower-cost laptops and ultra low-cost netbooks, we predict that the next few generations of PC chips are likely going to be optimized for price, with some consideration given to power consumption and almost no focus on performance. Other strong-selling devices, such as smartphones and possibly tablets, are likely to be optimized primarily for power consumption, with some attention paid to price. Again, performance may be almost irrelevant. Although some chips will continue to be performance-driven, this segment will not likely see as much growth.²⁰

Moore's Law is alive and well in 2010 and should remain so for at least the next few generations of semiconductor manufacturing technologies. But progress may well be measured by very different metrics to those used in the last two decades.²¹

Cloud computing: more than hype but less than hyper

DTT TMT predicts that in 2010 cloud computing²⁵ is likely to grow much faster than most other technology verticals²⁶ but will still fail to reach the heights its more enthusiastic supporters have suggested. We also expect to see it grow the fastest in the consumer and small and medium enterprises (SME) market, rather than in the large enterprise and government markets. Although estimates vary (usually due to definitions), in 2009 cloud services revenue was around $55 billion,²⁷ and we predict that it will likely grow by more than 20 percent in 2010 to roughly $70 billion, taking share away from alternative software models.

Before moving on, we may need to give some examples of cloud computing. There are personal applications (often advertising-supported) such as Gmail, Hotmail, Facebook and Twitter; consumer applications such as Google, eBay and Amazon; and business applications such as Saleforce.com and NetSuite.²⁸

The consensus on cloud is polarized. One group positions cloud computing at the very apex of the hype curve, but still two to five years away from mainstream adoption.²⁹ The other suggests that in 2010 "cloud computing is about to explode."³⁰

Our prediction is not just an average of the two extremes. We believe that there are significant advantages offered by cloud computing for some, but not all, applications and situations. There will likely be growth in those segments (non-critical file storage or customer-facing applications that see large spikes in demand), where offering dynamically scalable and virtualized resources over the Internet and as a service makes good business sense.

However, many surveys show that buyers of IT services remain hesitant about moving to the cloud. Among other things they cite concerns about security, reliability, and data portability and are sometimes skeptical about the longer-term cost benefits.31 These are serious concerns, and whether or not a shift to cloud computing actually increases these risks compared to whatever solution is currently being used is almost beside the point. A leading software chief technology officer said in late 2009, "Perception is absolute reality, and the perception is that security is still an obstacle, and we cannot deny it."³²

For reliability reasons, adoption of cloud is most likely to occur in markets with well-developed, secure, and reliable IT infrastructure: North America and Western Europe are currently the biggest users of cloud, and this is likely to persist for at least the short term. In the medium term, the developing world will likely catch up as cloud solutions are ideal for markets with few installed traditional hardware and software bases.³³

Although the media tend to focus on enterprise and government adoption³⁴ (or non-adoption) of cloud computing (specifically, systems infrastructure delivered as a service), the reality is that 60 percent of cloud today is cloud-based advertising, while the infrastructure component is only about 6 percent.³⁵

The prevalence of advertising, and other more consumer-centric applications, is consistent with the perceived benefits and risks offered by cloud computing.

When looking at the possible risks of cloud computing, enterprise and governments are legitimately obsessed with data security and maintaining near – 100 percent uptime, whereas consumers are likely to be more accepting of occasional outages, and the potential value of lost or stolen information from a single consumer is usually a great deal less than from a bank or government.

Equally, the benefits offered by cloud should benefit consumers disproportionately. Over the last decade, most enterprise and government IT departments have adopted the latest technologies, such as virtualization, to utilize their IT assets relatively efficiently. They usually have sourcing solutions that allow capacity to scale on demand, and they almost always have highly redundant storage of information. Consumers, on the other hand, tend to be inefficient, inflexible, and inadequately backed up.

Additional impediments to enterprise and governmental adoption are the issues of control and responsibility. In a traditional IT environment where the chief information officer (CIO) has followed best practices for data security and an employee loses a laptop or physical security is breached with consequent data loss, then the consequences fall on the shoulders of the employee responsible for building security. But in a cloud implementation, and even though the CIO cannot directly control security, the blame for any data loss will certainly fall on the IT department that made the decision to choose a cloud-computing approach.

Thinking thin is in again: virtual desktop infrastructures challenge the PC

DTT TMT predicts that in 2010, Virtual Desktop Infrastructure, a computing model based on "thin" or stateless clients, centralized applications, and processing power, will be taken far more seriously than in previous years, even if it does not outsell its thick client counterpart.

In 2010, over one million seats may go thin client,⁴² with the largest deployments involving tens of thousands of seats.⁴³ A pure thin client approach centralizes processing power and data storage and replaces a user's computer with a dumb terminal whose role is limited to sending keyboard and mouse inputs and receiving screen inputs. For the users, there should be little change in experience — the same business applications that they used in their former thick client environment should all look and feel largely the same.

In 2010, the majority of medium to large corporations are likely to consider the case for shifting to Virtual Desktop Infrastructure. By 2015, thin client may reach 10 percent of all enterprise client devices.⁴⁴ Cynics may note that it has been the "Year of Thin Client" for at least the last 13 years; but in each of those years, the case for thick client computing has remained more persuasive. Our view is that in 2010 an unprecedented confluence of enterprise priorities such as cost reduction, energy efficiency, security, and personal productivity is likely to make the case collectively for thin client strong enough for significant adoption.

Cost scrutiny is likely to include a continued focus on the IT budget.⁴⁵ Thin client could help deliver this via direct and indirect savings.

The principal direct cost savings is due to the ability to shrink support and maintenance and move-related costs. New recruits would not need a technician to build and install a PC for them, and human resources could simply hand out a standard thin client as part of the employee orientation. Employees could even pick up a unit from a stationery cupboard. Technical staff would not be required to undertake moves and changes. Project teams wanting to work in the same area would just need to be allocated a number of adjacent seats. Major software refreshes, such as an operating system (OS) upgrade, could be undertaken overnight, with no need to bring everyone's PC back to base. There would be less need for technical support to fix mechanical failures, as thin client units have no moving parts. Capital costs for end-user equipment may also be lower — cost per unit for devices is about a few hundred dollars.⁴⁶

The cost per software and operating system license could also be lower in a thin client environment. Economies could be achieved through having a central pool of virtual desktops that are shared between users but only activated when in use. If 10 percent of users are not using their desktops on any given day, a commensurate reduction in software costs is possible. With a thick client environment, licenses are typically paid for on a per device basis.

Indirect cost savings could include lower real estate costs as well. Thin clients are typically smaller and lighter than PCs and can be suspended under tables, which means workers could be allocated smaller desks yet still feel they have more desk space. Moves to new office premises or major personnel relocations that accompany the adoption of thin client computing could also facilitate "hot-desking", which can lower real estate requirements.

A further key source of indirect savings could be lower power consumption. Thin client units' power consumption can be as low as 4 watts.⁴⁷ By contrast the current generation of PCs consumes about 100 watts.⁴⁸ Lower power output also means less need for air conditioning. And an increasingly important benefit of lower power is a reduced carbon footprint at a time when emissions are expected to become an increasingly prominent key performance indicator for enterprises.⁴⁹ Lower CO2 emissions also reduce companies' spending on carbon credits or fines due to above-average emissions.⁵⁰

Thin client would lead to some power consumption simply due to being moved from PCs to data centers. But the use of shared processing facilities should create a more energy-efficient environment overall.⁵¹

Security is also expected to remain a major challenge this year, particularly for enterprises with flexible workers, or with project teams now spread across continents. Thin client means that no data are stored on the device. It also enables anti-malware software to be deployed centrally, rather than in the device, which can make it nearly impossible to copy a file. Thin client should address many of enterprises' key security concerns in 2010, including the following: the impact on reputation from the loss of data; the increasingly professional approach to the creation and dissemination of malware, phishing, and other forms of digital attack; and the ubiquity and capacity of portable data storage, facilitating illicit copying of files.

The fourth key factor that may make the case for thin client is productivity. Thin client could enable workers to waste less time in booting up, shutting down, or receiving software updates. Over the course of a week, total time savings could reach a couple of hours. Working at home using a thin client could present exactly the same working experience as in an office — available in an instant.

IT procurement stands on its head

"If you win the consumer, you will win the enterprise." DTT TMT predicts that 2010 will likely see many enterprise technology and telecommunications purchasing decisions based on the preferences of individual employees, rather than traditional IT department criteria.

In the past, technology and telecommunications hardware and software manufacturers targeted their latest products toward the enterprise market, and specifically the gate-keeping IT department. Enterprises tended to be early adopters, sophisticated users, and those willing to pay premium prices for premium products. A few years later, the solutions could be sold to the consumer market where typically less demanding users paid lower prices. The manufacturer made lower margins, but these were offset by larger volumes.

This traditional business model is now being reversed. Large chip companies are using their most advanced manufacturing techniques for devices aimed at the consumer, not the enterprise market.⁵³ What's more, new enterprise software applications for internal and external social networking were first implemented and validated by the consumer market.⁵⁴

The first segments to experience this reversal of historic procurement processes have been computing and telecommunications devices, although traditional enterprise software applications are beginning to see some signs of employee push-back.⁵⁵

Historically companies would support one, or perhaps two, standard desktop or laptop computers across the entire enterprise. They would almost always run a single version of a single operating system and browser across tens of thousands of workstations. The emerging pattern appears different. Some companies that used to limit their support to one type of PC or operating system are now responding to employee demands by supporting others.⁵⁶ Other IT departments are even more thoroughly disintermediating themselves from the PC-buying decision: they give employees the cash and allow them to choose the PC they want.⁵⁷

Of course, enterprise IT departments are unlikely to do this simply because employees ask them to. Advances in various operating systems, application programming interfaces (APIs), administration tools, hardware, and virtualization mean that administering a heterogeneous corporate client environment in 2010 should not be the nightmare it might have been just a few years back.⁵⁸

Even more of a bottom-up procurement trend is taking place in mobile telecom — especially smartphones. Total telecom costs are usually in the top three expense items in an enterprise, and the bulk of those costs are now mobile, as land line and long distance costs have fallen worldwide.⁵⁹ It has been standard practice for IT managers to control costs by standardizing on a single carrier and offering a very limited selection of devices that the firm will both support and pay for. Often these phones were not what the users would have chosen for themselves.

One response to this was the rise of the "prosumer" market — enterprise employees who bought the phone themselves, but used it for both work and play. This is a large market, estimated at 14 million users in the United States alone, and growing.⁶⁰ Some leading smartphone manufacturers now get more than half of their total new subscribers from prosumer sales.⁶¹ However, in the past, some enterprises did not allow prosumer phones to connect to the corporate IT network.

But many employees would prefer to have their smartphones better supported (and paid for) by their employers. Once again, evolutions in technology mean that devices that were formerly neither secure nor interoperable enough are now being allowed — if not always welcomed — into the enterprise IT world. Email, security, virtual private networks, and configuration management are all now likely to be supported to varying degrees.⁶²

As a result, more and more enterprises are likely to allow employees to choose their own corporate phones, or at least allow prosumer-selected phones to integrate better with the enterprise networks.⁶³

Of course, it isn't just PCs and smartphones. Employees are picking their own browsers,⁶⁴ insisting on the ability to use social media sites even while at work,⁶⁵ and even choosing their own voice communications solutions, including VoIP services.⁶⁶

Bottom line

The most obvious probable effect of a more consumer-centric procurement process is to make the traditional sales approach of the enterprise hardware, software, and telecom vendors much more difficult — or different, at least. The classic enterprise sale was not without challenges, but they were known challenges, and sales techniques were designed to sell to monolithic buyers whose concerns were enterprise in scale. If consumers become the new chief procurement officers, enterprise-focused vendors will need to learn new skill sets.

It is also likely to shift the existing balance of power between incumbents and new entrants. A traditional tool has been the bundle: a vendor will agree to provide a key product at a good price, but only if the enterprise customer buys a specified number of the vendor's other products. Devices from competitors are prohibited, even if the enterprise's employees want them. This sort of tied selling may be less successful in the future, and therefore vendors will have to become more competitive and flexible.

IT departments may have to become more flexible as well. Although new technologies have made administering and provisioning heterogeneous IT environments easier, it is still more complex than a homogeneous workplace. Further, traditional IT best practices are likely to continue to be necessary. Even when the employee has picked the PC or smartphone, some of the information that resides on those devices still belongs to the enterprise and needs to be deleted if employment changes.⁶⁷

Although being responsive to user preferences is a good thing and should make for happier employees, enterprise IT executives need to keep in mind the faddish nature of consumer sentiment. Workers may clamor for hot new devices, but there is a need for processes that reduce device and software proliferation and churn. Supporting too many choices becomes onerous to administer. Excessive churn can reduce productivity and escalate costs.

The future of many enterprise computing and telecom tools will likely be one where the line between work and personal lives is blurred. Some of that is at the request of the enterprise ("please be available 24/7") and some at the request of the employees ("OK — but please let me choose my smartphone").

Footnotes

1 High end ARM or low end x86 (Atom) processor equivalents have the power to support all the tasks required of netTabs.

2 Various examples including the Apple Newton and Microsoft Tablet are cited in Paul Atkinson Paul (2008) A Bitter Pill to Swallow: The Rise and Fall of the Tablet Computer. Design Issues, Autumn 2008, Vol. 24, No. 4, Pages 3-25, MIT Press.

3 A bitter pill to swallow: The rise and fall of the tablet computer, MIT Press Journals, Autumn 2008: http://www.mitpressjournals.org/doi/abs/10.1162/desi.2008.24.4.3

4 Archos 5 Internet tablet Android-based PMP, Register Hardware, 11 November 2009: http://www.reghardware.co.uk/2009/11/11/review_media_player_archos_5 /; and Vega tablet beats Apple and Crunchpad, Wired, 13 November 2009: http://www.wired.com/gadgetlab/2009/11/vega-tablet-beats-apple-and-crunchpad /

5 Analyst predicts tablet sales provides mockup, World of Apple, 8 August 2009: http://news.worldofapple.com/archives/2009/08/08/analyst-predicts-tablet-salesprovides-mockup /

6 Step aside Apple fans, Apple needs to sell tablets to the PC crowd, Retrevo Blog, 23 October 2009: http://www.retrevo.com/content/blog/apple-tablets-pc-users

7 Sales slowdown prompts re-evaluation among PND players, iSuppli, 14 October 2009: http://www.isuppli.com/News/Pages/Sales-Slowdown-Prompts-Reevaluation-Among-PND-Players.aspx

8 Google keeps disrupting, Forbes, 13 November 2009: http://www.forbes.com/2009/11/13/chrome-operating-systems-technology-internet-google.html

9 PND revenues to peak in 2010: iSuppli, GPS Obsessed, 24 June 2008: http://gpsobsessed.com/pnd-revenues-to-peak-in-2010-isuppli /

10 High growth reported for the World GPS market forecast to 2013, Reuters, 23 April 2009: http://www.reuters.com/article/pressRelease/idUS110111+23-Apr-2009+BW20090423

11 Garmin Nuvifone G60 adds cellphone to GPS unit, USA Today, 29 October 2009: http://www.usatoday.com/tech/wireless/phones/2009-10-28-garmin-nuvifoneg60-cellphone_N.htm

12 Moore's Law 40th Anniversary, Intel press kit: http://bit.ly/Intel_press_kit http://www.intel.com/pressroom/kits/events/moores_law_40th/index.htm?iid=tech_mooreslaw+body_presskit

13 Moore's Law to die at 18 nm, analysts predict, TG Daily, 16 June 2009: http://bit.ly/TG_Daily http://www.tgdaily.com/hardware-features/42874-moores-law-todie-at-18-nm-analysts-predict

14 Hot Chips panel looks for inflection point into silicon scaling, EDN, 24 August 2009: http://www.edn.com/blog/1690000169/post/150048215.html

15 Is Moore's Law becoming academic?, iSuppli, 16 June 2009: http://www.isuppli.com/News/Pages/Is-Moore-s-Law-Becoming-Academic.asp

16 Intel to play down clock speeds in new naming plan, eWEEK.com, 19 March 2004: http://www.eweek.com/c/a/Desktops-and-Notebooks/Intel-To-Play-Down-Clock-Speeds-in-New-Naming-Plan /

17 The mother of all CPU charts 2005/2006, Tom's Hardware, 21 November 2005: http://www.tomshardware.com/reviews/mother-cpu-charts-2005,1175.html

18 Server demand fuels record CPU sales, IT news, 23 January 2008: http://www.itnews.com.au/News/101651,server-demand-fuels-record-cpu-sales.aspx ; although lately power and cost are becoming issues even in high end CPUs; source: Internet Services: Researchers Save Electricity With Low-Power Processors And Flash Memory, ScienceDaily, 15 October 2009: http://www.sciencedaily.com/releases/2009/10/091014122056.htm

19 Intel's Atom architecture: the journey begins, Anand Tech, 2 April 2008: http://www.anandtech.com/showdoc.aspx?i=3276&p=1

20 Unless, of course, there is some novel application (like artificial intelligence) that requires much more powerful processors. http://www.bettscomputers.com/requirementsofai.htm

21 The focus on clock speed is more a 1985-2005 phenomenon. Prior to that it was more focused on increasing transistor count and the bus width.

22 This assumes a greenfield 18 nm plant of 1 million sq feet of fab, 300-500,000 of clean room, and roughly 35,000 per month of wafer starts, starting in 2013. Any reuse of real estate or equipment would lower costs.

23 A company that packages semiconductors: at one time most of the "rocket science" in semiconductors was in the silicon. Now the packages that contain the silicon are of equivalent sophistication and importance. As one example, current generation chips produce a great deal of heat in a very small area, which needs to be dissipated. Sophisticated packaging techniques need to be utilized.

24 There was an old software axiom: when designing next year's application, use this year's leading edge machines. By the time the software came to market, the "average" machine would be capable of running it. If hardware becomes less powerful, this technique will no longer work.

25 Many discussions of 'cloud computing' can degenerate into definitional debates. For simplicity, we have referred to the NIST definition found at http://csrc.nist.gov/groups/SNS/cloud-computing/index.html

26 IT spending is expected to be up 3 percent in 2010. Source: Gartner says IT spending to rebound in 2010 with 3.3 percent growth after worst year ever in 2009, Gartner Newsroom, 19 October 2009: http://www.gartner.com/it/page.jsp?id=1209913

27 Gartner says worldwide cloud services revenue will grow 21.3 percent in 2009, Gartner Newsroom, 26 March 2009: http://www.gartner.com/it/page.jsp?id=920712

28 All examples from Deloitte Report on Cloud, page 5. http://www.deloitte.com/assets/Dcom-Ireland/Local%20Assets/Documents/ie_Consulting_CloudComputing_09.pdf

29 Gartner's 2009 hype cycle special report evaluates maturity of 1,650 technologies, Gartner Newsroom, 11 August 2009: http://www.gartner.com/it/page.jsp?id=1124212

30 Cloud computing special: Ready to blow, vnunet.com, 8 October 2009: http://mobile.vnunet.com/accountancyage/features/2250858/cloud-computing-special-ready

31 Above the Clouds: A Berkeley View of Cloud Computing, UC Berkeley Reliable Adaptive Distributed Systems Laboratory, 10 February 2009: http://d1smfj0g31qzek.cloudfront.net/abovetheclouds.pdf

32 Large crowd gains insight into future of cloud computing, eWEEK, 3 November 2009: http://www.eweek.com/c/a/Cloud-Computing/Large-Crowd-Gains-Insightinto-Future-of-Cloud-Computing-894912/1 /

33 The oh-so-slow tech revolution, Business Today Egypt, November 2009: http://www.businesstodayegypt.com/article.aspx?ArticleID=8691

34 L.A. cloud contract goes to Google over Microsoft, Top Tech News, 28 October 2009: http://www.toptechnews.com/story.xhtml?story_id=020000OTW410

35 Gartner says worldwide cloud services revenue will grow 21.3 percent in 2009, Gartner Newsroom, 26 March 2009: http://www.gartner.com/it/page.jsp?id=920712

36 Does cloud computing need malpractice safeguards?, CNET News, 1 November 2009: http://news.cnet.com/8301-19413_3-10387879-240.html?part=rss&tag=feed&subj=TheWisdomofClouds

37 http://blog.gogrid.com/2009/11/10/gogrid-announces-new-pricing-more-volume-discount-plans /

38 Intel's next target, Forbes, 28 September 2009: http://www.forbes.com/2009/09/26/enterprise-processors-ibm-technology-cio-network-intel.html?feed=rss_technology

39 Cloud computing and the consumer, xchange magazine, 28 January 2009: http://www.xchangemag.com/articles/498/cloud-computing-and-the-consumer.html

40 Verizon's cloud-based computing as a service moves telecom solutions forward, SearchTelecom.com, 3 June 2009: http://searchtelecom.techtarget.com/generic/0,295582,sid103_gci1357914,00.html#

41 See: Page 28 ; http://www.deloitte.com/assets/Dcom-Ireland/Local%20Assets/Documents/ie_Consulting_CloudComputing_09.pdf

42 In 2009, APAC deployments of thin client may approach 1 million, if H2 09 installations match those of H1 09. Also see: Businesses Take Another Look at Virtual Desktops, The Wall Street Journal, 30 September 2009: http://online.wsj.com/article/SB125417207134047337.html

43 For example, see: http://www.japancorp.net/Article.Asp?Art_ID=22065

44 Facing big budget cuts, county turns to thin clients, SaaS, Computerworld, 8 October 2009: http://www.computerworld.com/s/article/9139142/Facing_big_budget_cuts_county_turns_to_thin_clients_SaaS

45 The principal motivation, common to 77 percent of decision makers at North American and European enterprises and small and medium businesses for considering an alternative to the traditional PC environment is "lowering costs." Sample: 609 PC decision makers. Source: Forrester, Client Virtualisation Adoption Trends, October 2009.

46 One company that moved from thick-client to thin-client estimated that its average cost per desktop fell from £3800 to £800. Source: IT shakeup is just the ticket at Transport for London, Computing.co.uk, 18 June 2009: http://www.computing.co.uk/computing/analysis/2244354/shakeup-ticket-tfl-4724924

47 For examples of power consumption, see: http://www.sun.com/sunray/sunray2/index.xml ; http://www.linuxfordevices.com/c/a/News/Fujitsu-Futro-S100 /#

48 The next generation of PCs could have half the power consumption. For information on current PCs' power consumption, see: http://www.jisc.ac.uk/publications/documents/greenictqueenmargaret.aspx #

49 For example, a league table of the UK's 5,000 largest organizations' carbon footprint will be published: http://www.decc.gov.uk/en/content/cms/what_we_do/lc_uk/crc/crc.aspx

50 The UK's carbon reduction credit program will require large organizations, from fiscal 2010–2011, to purchase allowances equal to their annual emissions: http://www.decc.gov.uk/en/content/cms/what_we_do/lc_uk/crc/timings/timings.aspx

51 Cost savings including pro rata usage of data center are up to 50%. Source: Thin Clients Trump PCs On Energy Consumption, Environmental Leader, 26 March 2009: http://www.environmentalleader.com/2008/03/26/thin-clients-trump-pcs-on-energy-consumption /

52 Review: 3 Windows 7 touch-screen laptops, Computerworld, 9 November 2009: http://www.computerworld.com/s/article/9140385/Review_3_Windows_7_touch_screen_laptops?taxonomyId=125 ; and Toshiba announces touchscreen notebooks, netbooks for Windows 7 launch, Engadget, 14 October 2009: http://www.engadget.com/2009/10/14/toshiba-announces-touchscreen-notebooksnetbooks-for-windows-7 /. For a video view, see: http://www.eweek.com/c/a/Windows/Windows-7-Touch-Screen-Functionality/?kc=rss

53 Intel to rev up Atom development, executive says, CNET News, 28 September 2009: http://news.cnet.com/8301-13924_3-10362652-64.html

54 The 'ROI' case for Web 2.0: return on information, Internet Evolution, 18 September 2009: http://www.internetevolution.com/author.asp?section_id=796&doc_id=181952&f_src=internetevolution_gnews

55 Why younger workers shun business software and what to do about it, itbusniess.ca, 4 November 2009: http://www.itbusiness.ca/it/client/en/Home/News.asp?id=55111

56 The Mac in the gray flannel suit, BusinessWeek, 1 May 2008: http://www.businessweek.com/magazine/content/08_19/b4083036428429.htm

57 40% of Citrix employees choose Mac for home and work, MacNN, 29 July 2009: http://www.macnn.com/articles/09/07/29/citrix.byoc.program /

58 http://www.forrester.com/Research/Document/Excerpt/0,7211,46634,00.html [link doesn't work]

59 Economic downfall forces enterprises to cut telecom cost, ArticlesBase, 29 June 2009: http://www.articlesbase.com/management-articles/economic-downfallforces-enterprises-to-cut-telecom-cost-1000622.html

60 Prosumers: A new growth opportunity, Cisco IBSG, March 2008: http://www.cisco.com/web/about/ac79/docs/wp/Prosumer_VS2_POV_0404_FINAL.pdf

61 Smartphones like BlackBerry give rise to the "prosumer," Mobile Marketer, 21 December 2007: http://www.mobilemarketer.com/cms/news/manufacturers/244.html

62 Top 3 iPhone features for the enterprise, iPhoneCTO, 3 September 2009: http://iphonecto.com/2009/03/09/top-3-iphone-features-for-the-enterprise /

63 Employees are driving iPhone enterprise share gains — at the expense of BlackBerry, TBI Research, 6 November 2009: http://www.tbiresearch.com/theemployee-invasion-employees-driving-iphone-enterprise-share-gains-2009-11

64 IBM backs Firefox in-house, CNET News, 12 May 2005: http://news.cnet.com/IBM-backs-Firefox-in-house/2100-7344_3-5704750.html?tag=nefd.top

65 Montgomery employees to have access to Facebook, Washington Examiner, 9 August 2009: http://www.washingtonexaminer.com/local/Montgomeryemployees-to-have-access-to-Facebook-52841652.html

66 Skype use in the enterprise, Processor, 9 October 2009: http://www.processor.com/editorial/article.asp?article=articles%2Fp3125%2F32p25%2F32p25%2F32p25.asp

67 Their phone, your headache, IDG News Service, 16 October 2009: http://news.idg.no/cw/art.cfm?id=5DE1761A-1A64-67EA-E43676FE78710F42 http://cleantech.com/about/cleantechdefinition.cfm

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.