Revenue from Nvidia's mobile processors over the next four years will grow tremendously, Jen-Hsun Huang predicts, as they get embedded in new mobile devices.
Nvidia, best known for its high-end graphics chips, will generate a vast majority of its revenue from its now burgeoning mobile-processor business, according to outspoken Chief Executive Jen-Hsun Huang.
Huang, speaking to a roundtable of reporters today, said he expects the mobile-chip business to grow tenfold, to $20 billion by 2015. In comparison, the graphics processor business is expected to grow 75 percent, to $7 billion in the same time frame. He added he expects to maintain a strong share in both businesses.
"We'll be quite a force to contend with," Huang said.
Over the past few years, Nvidia has exited the computer chipset business and moved into the faster-growing mobile-processor business with its Tegra chips, found in devices such as Samsung Electronics' Galaxy S II phones. Huang said that in the smartphone and tablet businesses, Nvidia and Qualcomm are the only real serious providers of chips to consumer electronics equipment manufacturers.
"We're the only person actively on the dance floor with Qualcomm," he said.
There is an addressable market of 100 million devices that need mobile processors this year, a figure that could balloon to 1 billion, he said. The growth is expected to be driven by more mass-market smartphones, computers running on more efficient ARM processors, and the proliferation of energy-efficient ultrathin notebooks.
Huang estimates that Tegra chips are found on half of all high-end Android smartphones and 70 percent of Android tablets, and he added that he expects Nvidia to maintain a high market share in both businesses. He also sees an opportunity to provide chips for mass-market smartphones due to its acquisition of Icera, which enables the company to combine a cellular radio with a processor, much like Qualcomm does with its line of Snapdragon chips.
Despite Android tablets' slow start, Huang said they will make up half of the market in four years, or roughly the same amount of time it took for Android smartphones to win widespread acceptance in the market.
Huang's confidence underscores the tremendous growth in the mobile-devices business, which has been a boon to some companies, even as other traditional players have flailed. He added that companies need to have a mobile strategy to succeed in the computing business.
"If you don't have a mobile strategy, you're in deep turd," Huang said. "If you're not in mobile processors now, you're seven years too late."
Hewlett-Packard's decision to spin off its PC business and shutter its WebOS mobile division is indicative of not having a clear strategy in mobile, he said. Likewise, he doesn't believe that Intel poses a major threat because its low-power Atom chips don't run on the same architecture as the ARM processors that power most mobile devices now.
"They're speaking the wrong language," he said. "We're not worried about them at all."
ARM processors hold a strong advantage because a majority of mobile applications are written using that standard, and not the one Intel uses. Likewise, he said he believes that apps written for Windows Phone 7 will run on Windows 8, which is an operating system designed for both PCs and tablets.
While the company's graphics chip business isn't seeing as much growth, Huang said he sees continued demand as the need rises for better graphics performance. He noted that cameras in smartphones didn't kill off the digital-SLR business, since it only drove the need for higher picture capture capabilities.
He added that the company took some of the benefits from the Tegra processor--including the ability to completely shut down when not used--and is looking to move them to its graphics processors.
While Intel is pushing Ultrabooks as a competitor to the MacBook Air in the ultra-thin category, Huang said he expects more "clamshell devices" with keyboards and more powerful mobile processors to eventually win out. He said those devices would most likely be priced at or below current tablets. The Ultrabooks are expected to retail around $1,000, while tablets can be bought for less than half that price.
With $2.5 billion in cash on the books, Huang said Nvidia isn't done buying other companies. He noted that while the valuations for Internet start-ups are high, chip start-ups can be had for decent value. Nvidia announced its acquisition of Icera for $367 million in May.
"We're always looking for a good opportunity," he said.
Corrected at 9:17 a.m. PT: The $20 billion figure refers to the total mobile business, and not just Nvidia's share of the market.
Updated at 3:02 p.m. PT: to clarify that Nvidia was exiting the computer chipset business.