Microprocessors, he said in an article published more than a decade ago, "are a dream business, with wonderful margins and a wonderful market position. How could anything else compete here for resources and profitability?"
Barrett might have been talking about Intel of 2000 when he made those comments, but the statement still rings pretty true for the Intel of today.
The latest casualty? Intel Media, the company's fledgling
Earlier this year, Intel Media head Erik Huggers appeared on stage at a tech conference and said the company would. A few months later, he was scrambling to find a buyer, according to people familiar with the matter. Verizon has emerged as the most likely bidder, they said, and a deal should be announced by the end of the year.
What changed so quickly? Signing deals with content companies turned out to be much harder -- and pricier -- than expected. And there was no sign Intel Media would make money anytime soon. Add to that uncertainty who didn't want to shoulder all of that risk on his own.
In some areas, Intel is willing to gamble, like spending billions of dollars on new chip factories without really being able to accurately predict demand, or creating chips for devices that hadn't yet been designed.
But Intel has been much more hesitant to take risks outside of its core processor business over the years. In 1998, the company launched an ambitious program to create new businesses that would bring in revenue beyond its PC chips. The company dipped its toes into tablets and smartphones years before
In each instance, however, Intel pulled back before any of those trends became popular, closing most within a couple of years. In some cases, such as mobile, the decision to exit proved to be the wrong choice.
As Intel faces a future with slower PC sales -- still its core market -- the company may need to take more risks. What those will look like is unclear. But what we do know is some risks it has decided not to take over the years.
Here's a look at some opportunities Intel has missed:
Intel's history in the TV industry has been rocky. Intel Media, while the latest push, wasn't the first. In 2005, Intel unveiled its Viiv technology to allow videos and music to seamlessly move between a PC and a TV or portable device. If a PC carried the Viiv brand, it was supposed to serve as the hub for a digital media universe, sending content to televisions around the home.
But the first PCs to arrive with the Viiv brand were only incremental updates to existing Windows Media Center PCs. And follow-up products didn't live up to the technology's promise, either. In 2007, the company decided to scale back the Viiv brand, and it disappeared soon after.
Intel then became an early partner for Google TVs and other smart TVs, with its processors powering a Sony TV and a Logitech Google TV set-top box. However, such products -- and Google TV in general -- flopped. Logitech even called its Google TV box "."
Intel gave up on trying to get its processors into TVs in late 2011 after failing to gain traction. Huggers, the head of Intel Media, later revived the company's TV efforts but took Intel down a different path to providing a complete software, hardware, and content package.
An increasingly glaring Intel misstep is in smartphones. Virtually every mobile device runs on chips based on a design by rival ARM Holdings, created by companies such as Qualcomm and Samsung Electronics, with Intel left to compete for scraps.
Intel had a chance to get into the chip business earlier, and actually gained the rights to use ARM's designs in 1998 and acquired chipmaker DSP Communications. Intel renamed the business XScale and courted customers such as BlackBerry, Palm, and Motorola.
Over the next four years, Intel spent more than $11 billion acquiring three dozen cell phone and communications chip companies, but it couldn't make inroads against rivals such as Qualcomm. With concerns mounting about competition in its core PC chip business, Intel decided to dump the cellphone chip operations to Marvell Semiconductor in 2006 for $600 million.
That decision might have appeared smart at the time -- Intel was believed to be losing money in the cellphone market -- but in hindsight, it has proved to be one of the company's biggest mistakes.
Intel now pushes its own X86 processors for smartphones and tablets, but it has yet to secure any design wins that ship in high volumes. Intel has said 2014 will be a better year for its smartphone business, but whether that's the case remains to be seen.
Believe it or not, Apple's tablet wasn't the first
The devices were boxed and ready to ship to retail partners such as Circuit City later that year when Intel decided to cancel the project. Its PC partners were upset with the device, as well as the company's other consumer electronics products, and didn't want Intel to release something that would compete with them.
Intel ended up pulling the shipments and selling the devices in a fire sale to its own employees. Then in 2010, Apple released its first iPad and revolutionized the computing market. Much like the smartphone market, Intel went from a potential early player to a company forced to play catch-up. It has
"I couldn't say then that Intel was making the wrong decision," Ed Arrington, venture lead in Intel's new business initiatives group and one of the three initial employees working on the Intel Web Tablet, told CNET. "Now I look back on it and say if we'd done a tablet back then, look at where we'd be today."
In the late '90s, Intel started looking for ways to expand its reach in the consumer market, particularly with children. The company in 1998 decided to partner with toy maker Mattel -- of Barbie and Hot Wheels fame -- to create a "Smart Toy Lab" in Portland, Ore. The new lab's task was to create high-tech toys that combined Mattel's toy and marketing know-how with Intel's technology expertise.
The two companies in 1999 introduced an Intel Play brand of technology toys, starting with the $99 QX3 Computer Microscope and the $69 Me2Cam Virtual Game System. The QX3 was rated the top-selling multimedia toy of the 1999 holiday season despite being priced higher than most other products in the category.
However, in May of 2000, Intel and Mattel decided to go their separate ways. Mattel wanted to focus on its core market of traditional low-tech toys, and its vision about the future of smart toys and the Smart Toy Lab started to differ greatly from Intel's. Intel continued its investment in the toy lab, releasing two products in the following months.
Intel discontinued the Intel Play line in early 2002 after selling it to Prime Entertainment, which continued the product line under the Digital Blue brand.
Other consumer electronics
Intel didn't just sell toys in the late 1990s and early 2000s. It had an entire line of consumer electronics products, including digital cameras, wireless keyboard and mouse sets, and portable music players. Ultimately, all of its consumer electronics devices were designed to work with PCs, boosting the company's core market.
"Intel offers consumer products that are innovative, fun, and affordable," Don Whiteside, then vice president of Intel's connected products division, said in an October 2001 press release.
In early 2001, the company released the Pocket Concert Audio Player, a device to play songs recorded in MP3 or Windows Audio Player formats, as well as FM radio. The Intel Personal Audio Player 3000, which came with half as much flash memory as the earlier product and cost half as much, debuted a few months later.
Intel also offered digital cameras. One, also introduced in October of 2011, was called the Intel Pocket Digital PC Camera. The images could be posted to a personal Web site or sent in e-mail messages. Users also could print photos using their PC connection, as well as use the device as a video phone for face-to-face calls or as a webcam.
Around that same time, Apple unveiled its first
Intel eventually axed its consumer electronics unit. The company in October 2001 said it would phase out its business for digital cameras, audio players, and toys, selling the existing inventory and halting manufacturing. It said at that time that the business didn't meet its requirements for long-term growth potential.
In the late 1990s/early 2000s, the PC market faced some uncertainty. Big software companies were arguing that the PC was dead -- sound familiar? -- and computing of the future would come through sort of dummy terminals. Rather than a full-blown PC, most of the computing would be done online.
To get in front of that possibility, Intel in mid-2000 launched its Dot.Station Internet appliance. The countertop terminal was powered by Red Hat Linux and let people surf the Web, exchange emails, and make phone calls.
However, the market already was crowded with rivals, and Intel only offered the product through service packages with telecom companies, Internet service providers, and similar businesses. Intel had some success selling it overseas, but it didn't catch on broadly.
Intel in late 2001 said it would phase out the product along with the rest of its consumer electronics business.
So far, the prediction that dummy terminals would kill full-blown PCs hasn't proved true. However, devices such as
Intel provided Web hosting, called Intel Online Services, while Amazon was still a new company. The company first considered running data centers to house other companies' Web sites in 1996, but executives decided against it. Two years later, Renee James, then the chief technical aide to Chairman and CEO Andy Grove, looked at the concept again. Intel approved her plan in late 1998 and opened its first data center the following year.
Web hosting may be very different from the chip industry, but Intel believed it was well positioned because of its ability to manage factories all over the world. Its facilities were to become "bit factories" that would process data and conduct Internet transactions on behalf of large companies and telecom providers. Intel hoped to gain insight into e-commerce trends and protect itself against the possibility that dummy terminals would replace PCs.
Two years later, Intel redeployed its workers and shut down the business. It had struggled to attract customers, and it wasn't making as much revenue per customer as it had anticipated. The closure came at about the same time Intel shut down several other businesses to focus on its core operations. Earlier in 2001, it also terminated iCat, an e-commerce hosting service for small- and medium-size businesses.
Now, Web hosting and cloud computing are two of the biggest trends in tech. Amazon has become a powerhouse in the sector, providing the backbone to run popular sites such as Netflix and Pinterest. The company doesn't provide many details about Amazon Web Services, but analysts estimate the business generates about $4 billion in annual revenue.
In May 2000, Intel unveiled Internet Media Services, which aimed to provide a network for streaming media. The business rented airtime and other services to companies that wanted to broadcast stockholder meetings, training courses, or other events over the Web.
Initially, Intel planned to invest $200 million into the project to set up broadcasting centers in the US and Europe and install servers in key locations around the globe. Early customers included Nasdaq.com and Golf Magazine.
However, in February 2001, Intel shuttered the business. Like many of its other service businesses, Intel landed only a few customers and never established itself as a major player in the market. In addition, the streaming business hadn't grown as fast as projected in 2000. Many companies had scaled back plans because of a downturn in the US economy.
Intel, at the time it announced it was closing Internet Media Services, said it "didn't see a timeline to get to profitability that would be adequate for Intel."
As much as Intel has tried, it just can't seem to shake that pesky creosote bush. In the future, it may wish it had tried a little harder.