Why Is the Intel of the "Dum, Dum-Dum-Dum" Chime Drifting Away From Us?

How did it miss its chance?

According to our backend statistics, there is a 90 percent chance you are not reading this article on a device stamped with Intel Inside.

That is no surprise. The last time you saw a phone with an Intel processor was probably eight or more years ago (if you ever saw one at all). According to the earnings report Intel released this morning, its PC and mobile processor business fell 8% over the past three months from the same period last year. You can't tell how much money Intel actually lost on mobile, because last quarter Intel merged the figures for its mobile and PC businesses. The mobile losses are now hidden.

Earlier last month, Intel cut its first-quarter revenue forecast, and at one point more than US$7 billion was wiped off its market value.

Over the past three months, Intel's only bright spot was its data center chip business, which grew 19% and has become the core of its revenue. For the second quarter, Intel said growth would again have to come from this business.

But this is not good news. The server processors used in data centers, or so-called cloud computing platforms, mostly go to big customers such as Google, Amazon and Facebook, which leave Intel no fat margins. In the old PC market, by contrast, hundreds of brands around the world revolved around Intel, and the hard-earned money went mainly into the pockets of Intel and Microsoft.

In the mobile era, the chip giant that discovered Moore's Law, built Silicon Valley, and stamped the "dum, dum-dum-dum" chime and the blue Intel Inside logo into your memory is drifting further and further away from us.

The world went mobile, and Intel missed the chance

In 1999, Intel's market value peaked at more than US$500 billion. Today it is under US$150 billion. Yet the world's shift to mobile was already foreshadowed in Intel's golden age.

In 1992, Apple Computer produced its first personal digital assistant (PDA), the Newton. It ran on an ARM 610 RISC chip. The device never found its market, though, and was a commercial failure. Four years later, PDAs got their first hit, the Nokia 9000 Communicator. It did not use an Intel processor either; it chose Intel's rival AMD.

Not until 2002 did Intel team up with Microsoft to supply chips for its PDAs and smartphones. For this, Intel prepared the XScale series of chips, based on ARM architecture. Meanwhile, Apple's iPod music player, launched in 2001, had set the consumer electronics market alight. It used an ARM processor.

Over that decade, smartphones gradually replaced PDAs, and devices for getting online or for entertainment became ever more mobile. Intel did have chances.

In 2005, Microsoft smartphones running Windows Mobile 5.0 used the XScale processor. The following year, Windows Mobile's market share rose by 20%. Yet that same year Intel sold the XScale series to Marvell Technology Group and went back to focusing on its own x86 processors.

Microsoft's smartphones later sank below 3% market share, which lets Intel say it has no regrets. But what about the iPhone? According to Intel's former CEO Paul Otellini, besides selling off XScale, Intel also turned down the chance to make the processor for the first iPhone. It could not see a future in smartphones.

After Apple released the first iPhone in 2007, the world's turn toward mobile sped up. By then, Intel had already missed its chance twice.

For a long time afterward, Intel still did not take the mobile market seriously. Compare Intel's PC processors with its mobile processors of the same tier, and it is easy to see that PC processors, including those for netbooks and Ultrabooks, always got new technology first. In November 2011, for example, Intel's 32-nanometer processors went to PCs first. Phones got processors of the same tier only two or three months later.

Intel tried to reverse the PC's decline, and failed

Andy Grove, Intel's founder, once wrote a memoir titled Only the Paranoid Survive. Perhaps it was this corporate culture that made Intel cling to the PC business and x86 technology to the end.

With mobile devices popular, Intel set out to make PCs mobile too. In February 2008, it introduced the netbook. The device was small, portable and fairly cheap. But anyone who has used one knows how frustrating its low specs were: web pages loaded slowly and movies looked poor. Consumers had no reason to stay fond of such a product for long.

In October 2010, Apple released a redesigned MacBook Air that dropped the mechanical hard drive for a solid-state drive. Fast startup and shutdown changed how people experienced their computers. A thinner, lighter body and longer battery life were also appealing. These qualities warmed up the high-end laptop market. Right after, at Computex in Taipei in 2011, Intel introduced the Ultrabook. You guessed it: the "new" selling points of the so-called Ultrabook matched those of the MacBook Air exactly, right down to the far-from-cheap price.

While Apple was reinventing how people used PCs, it also launched the iPad tablet in 2010. During this period, Intel's Ultrabooks, which offered nothing distinctive, drew no response from the market. The iPad, meanwhile, sold better and better, came to define the tablet market and made the most of its head start. Later, Intel had no choice but to push the hybrid concept of an "Ultrabook-tablet two-in-one." Such muddled positioning will not win over both market segments at once, unless the price is low. So from here on you will see more low-priced PCs switching to Intel chips. But judging by the trend, isn't the market with a future the high-end laptop?

Longtime partner Microsoft was a beat behind on operating systems

As you can see, Intel's efforts to reverse the PC business's decline always came just a step too late. And Intel's longtime partner Microsoft has also always been a beat behind.

Start with the PC. Surely some people are still running pirated copies of Windows XP. The system was such a classic that it lasted 13 years; updates stopped only in April 2014. Why? You have surely tried its successors, Windows Vista, Windows 7 and Windows 8. Most people's reaction after using them was: what the heck? It's not even as good as XP!

Windows Phone users, meanwhile, have long been stuck waiting, and waiting in vain. In five years, Microsoft released only three versions of its mobile operating system: Windows 7, Windows 8 and Windows 10. Many basic features were slow to arrive, such as VPN settings, quick viewing of background apps and mobile data usage statistics.

On top of that, apart from Microsoft's own office apps, other useful and fun apps always reached the app store a step late. Take Alipay Wallet. Its app store listing reads: "Sorry to keep you waiting, dear users, we're here!" Yet its latest version is stuck at 8.2.1.806. That is why Windows Phone users stayed unusually calm around Chinese New Year: they had no red envelope feature at all.

Basically, the apps people use most in daily life, such as WeChat and Didi, were slow to update. And apps that suddenly take off generally don't offer a Windows Phone version at all, or show up belatedly, after the craze has passed.

Besides slow updates, the operating system's versions don't work with one another. Existing Windows Phone 7 phones cannot be upgraded to Windows Phone 8 because the two use different kernels. Phones running Windows Phone 7.5 can only be upgraded to Windows Phone 7.8. Low-end models are left out too: the newly released Windows 10 for Phone supports no low-end phone with 512MB of RAM.

Intel tried to build its own operating system and lost too much time

On PCs and on mobile alike, slow-moving Windows plainly held Intel back. So Intel set out to build its own operating system.

In 2007, Intel began work on Moblin. Two years later, at its fall Intel Developer Forum (IDF) in San Francisco, Intel formally unveiled Moblin 2.1, aimed at portable mobile devices. That year, Dell's Mini 10v netbook offered Moblin as an option. Rumor had it that many PC makers and contract manufacturers, including Acer, Asus and Foxconn, were willing to adopt Moblin. In the end, nothing came of it.

In 2010, Intel teamed up with Nokia, which was also in trouble. Squeezed between the iPhone and Android phones, Nokia was losing global smartphone share fast, and its Symbian system was running out of steam. The two quickly agreed on a new system, MeeGo, which merged Nokia's Maemo and Intel's Moblin platforms. MeeGo ran on both x86 and ARM architectures and raised high hopes. Yet the only product ever to reach the market was the Nokia N9. Later, even Nokia chose to partner with Microsoft on the Lumia series and abandoned MeeGo.

Intel did not give up. In 2011, MeeGo was combined with Samsung's LiMo system to create Tizen. You may have heard of it only recently, because in 2014 Samsung planned to use it to revive its flagging phone business, and also wanted to push Tizen onto smart TVs and smartwatches. But what is the point of developing a system with no products to run it on? Better to save the effort. That same year, for the sake of its tablet business, Intel began putting R&D into porting Android: making the Android platform support the x86 processor architecture and providing developer tools so that developers could build apps for its tablets.

Intel's attempt to build its own system was far too winding. After a long detour, it still ended up in Android's bowl.

Tablets, its only mobile gain, are faltering too

Intel put a great deal of effort into tablets, from the underlying technology and operating system to marketing. In September 2013, Intel released Bay Trail, a 64-bit low-power system-on-chip. The ability to switch between Android and Windows was the highlight of Intel tablets. Tablets were also split into categories for different users, such as office, travel, calling, video and gaming. The results were decent. According to Intel's 2014 annual report, its tablet business grew 18% over the previous year.

In fact, as with smartphone makers, most tablet brands did not want Intel chips. So Intel paid subsidies to win partners, finding Chinese makers such as Ramos, Teclast and Aigo, as well as traditional PC makers like Dell, Asus and Acer. Intel also teamed up on R&D with Chinese chip design companies such as Spreadtrum and Rockchip to tackle costly components and long development cycles.

But as in the PC business, global tablet sales growth was slowing. Apple's results for the first quarter of fiscal 2015 showed global iPad sales falling 9% from the previous quarter. It was the second drop in iPad sales. With big-screen phones and laptops with powerful chips selling well, tablets, whose purpose had grown blurry, had less and less room to survive.

Next, Intel set its sights on wearables

PCs and tablets were nearly finished, and phones offered no opening. Intel needed new products to start over, such as wearable devices. In 2013, Brian Krzanich, the new CEO, first admitted Intel's strategic mistakes in mobile devices as soon as he took office, and then went straight on to wearables.

The topic is hot. Take smartwatches alone: there are Google's Android Wear, the Motorola Moto 360, Apple's Apple Watch, and the crowdfunded Pebble. None of them uses Intel chips. In other words, the most promising wearable devices today have nothing to do with Intel.

Didn't Tag Heuer, the watch brand owned by luxury group LVMH, just announce a smartwatch with Google and Intel? That looks more like a publicity stunt. A more practical approach would be to use the Edison processor, designed specifically for wearables, to develop products with real market appeal as quickly as possible, rather than interesting exhibits. And since no one knows whether treating wearables as another gateway to internet services will work, Intel may also have to explore new business models.

For now, Intel's new world is far from clear.