Intel is arguably the most efficient, most advanced manufacturer in the world and one of the most sophisticated companies when it comes to bringing cutting-edge technology to the mass market.
It just isn’t very good at branching out.
The Santa Clara-based chipmakers inability to move into new markets has come up again amid rumors that it is working on services to provide TV programs over the Internet.
“The chip maker envisions the service as a bundle of television channels akin to the packages provided by cable and satellite television providers, according to a person with knowledge of the discussions, who did not want to be identified talking about unannounced company plans. Instead of using the service via a set-top box, consumers would tune in with devices that run on Intel chips,” wrote the New York Times.
It sounds dreamy. Intel takes its know-how in digital technology, manufacturing and marketing to team up with content providers to effectively create a parallel universe for content, a shadow alternative to cable networks. In fact, it eliminates the need for cable companies. With this, content providers could deliver their own programs to consumers directly, splitting the cost of the transmission and backbone with the big I and a few other select partners. Plus, the company’s bank account and relentless drive will help the effort steer through tough times.
On paper, it’s great. Add stock analyst jibberish here.
What could go wrong? Plenty, says the historical record. Despite all of its tangible strengths, Intel has mostly fumbled, or at least not made much of a dent, when venturing beyond the PC market. The bulk of the company’s revenue, in fact, still comes from processors and chipsets for servers, desktops and notebooks. A few samples from the Hall of Shame:
--In 1999, it unveiled Intel Online Services, which promised to provide backbone communication services. The idea was that Intel knows how to efficiently run factories on a global basis, so it should know how to monitor web services. It terminated IOS in 2002 with a $100 million charge.
--Between 1999 and 2003, Intel bought 37 companies for over $11 billion, mostly to compete against Texas Instruments, Broadcom and others in communications chips. The idea was to produce silicon for things like cell phones and those big boxes in Telco datacenters. It subsequently sold most of the acquisitions.
--At CES in January 2004, CEO Paul Otellini announced that Intel would make LCOS chips for sharp, digital TVs. It killed the effort in October 2004.
--ViiV. Class, discuss.
--Intel at various times has marketed videoconferencing systems, MP3 players and digital cameras under its brand. If you own one, you're in a select group, involuntarily.
--The company along with Micron invested millions into Numonyx, a startup that promised to produce an energy-efficient, high-performance memory chip that could replace DRAM, flash memory and even hard drives. The original concept dates back to the 70s. After several delays, Numonyx came to market, and sold some chips to makers of Pachinko machines.
--SpectraWatt, a solar company spun out of Intel in 2008 and backed with over $91 million in VC funds, declared bankruptcy in 2011.
So what gives? Intel is a great component maker: it can offer pricing, performance and product availability better than competitors like AMD. Chalk it up to relentless manufacturing methodologies. It can also comprehensively lay the foundation for standards like USB and WiFi that help expand the usefulness and demand for PCs. Digital technology simply would not have expanded as fast as it did without these top-down standards efforts funded largely by Intel.
Intel also does a great job of expanding into adjacent markets. In the early 90s, Intel had only 3 percent of the market for server chips. Ten years later, Sun Microsystems, IBM and Digital had been pushed to the side: Intel controlled all but 3 percent of that market.
But here’s the problem. Intel’s size and track record scares the heck out of incumbents and prompts them to put up whatever effort they can to block Intel from even getting a toehold in their neighborhoods. They’ve all read Andy Grove’s “Only the Paranoid Survive.” Potential customers also tend to get leery about becoming increasingly dependent on a supplier.
As a result, Intel tends to succeed when it can fundamentally tie its existing product lines to the new one. Case in point: graphics chips. Back in the late 90s, Intel bought a Chips & Technologies and a few other outfits to make graphics chips for PCs. Nvidia and ATI Technologies circled the wagons and it flopped.
Intel bowed out of standalone graphics chips, but said it would continue to make chipsets with integrated graphics. Chipsets, which allow the microprocessors to communicate to hard drives and other components, are one of the least glamorous, but most important, pieces of silicon on the market. Intel has been a leader for years.
Integrated graphics chipsets don’t provide the same clarity and performance of standalone graphics chips, but to PC makers and consumers they provide another advantage. They’re a lot cheaper. Within a handful of quarters, Intel was the largest provider of graphics silicon on the market, all from the strength of its chipsets.
Additionally, some of these new businesses can get lost in the bureaucracy. Sources have said over the years that Intel’s push into healthcare was stymied a bit by a lack of interest in the topic by CEO Paul Otellini. Intel can put ten microprocessors into every wind turbine, but the total number of turbines produced is miniscule compared to the number of PCs.
The company might be able to wrangle all of the right partners to remake the world’s home entertainment grid. But don’t cancel your cable subscription just yet.
