The endorsement couldn’t be any sweeter. Intel Corp. is at least five years and “probably” more ahead of the competition in process technology, says John Daane, chairman, president & CEO of FPGA vendor Altera Corp. That’s why Altera will be sourcing semiconductor wafers from Intel starting in 2014, shifting wafer sourcing for high-end products from long-term foundry partner Taiwan Semiconductor Manufacturing Co. Ltd., to secure a technology edge against key rival Xilinx Inc.
Intel’s technology leadership in semiconductor manufacturing is widely acknowledged in the industry and this could bring other chipmakers to its side as it accelerates a widely expected push into the contract wafer production business. The deal with Altera is being carefully watched by other chipmakers and analysts who want to know how deeply Intel could insert itself into the foundry market and the implications for dedicated foundries, including TSMC, GlobalFoundries and United Microelectronics Corp.
“We are moving to Intel in 2014 [because] nobody is getting close to where Intel is in process technology,” said Daane during a presentation last month at the Citi Global Technology Conference. “If you look at the roadmap of the semiconductor industry from a foundry perspective, there is nothing that catches up with Intel’s 14 nanometer for the next four to five years.”
Industry observers and analysts agree Intel is indeed ahead of the competition in the volume shipment of next-generation products. Take 22 nanometer FinFET process, for instance. Intel began producing chips using this technology in 2011 and shipping to customers started in the first half of 2012. “No competitor has even demonstrated a working FinFET product,” said Ashraf Eassa, a technology analyst in a report. “Do I believe that the foundries will eventually get there? Of course. What I don't see is Intel losing its process edge. In fact, I expect it to widen over time.”
Intel’s foundry business has a minuscule market share today but the company is already sending tremors through the semiconductor world as it slowly expands operations in a direction most never imagined. The world’s biggest microprocessor supplier is eyeing opportunities in markets adjacent to its regular PC business -- consumer electronics, for example – but it is also flexing muscles in the contract manufacturing of semiconductor wafers and using its considerable resources to take on the biggest players in the sector.
Still, Intel won’t be rolling over entrenched players like TSMC. Altera, for example, has had a long relationship with TSMC and isn’t about to scrap this even as it engages with Intel, according to Daane. The company plans to produce certain products with TSMC and a different set with Intel, giving itself opportunities for second sourcing while guaranteeing productivity advantages for areas where the American semiconductor giant is ahead of its Taiwanese foundry rival.
“We are going to have two foundries,” Daane said. “We’ll have some products with TSMC and some products with Intel. Right now the high end is certainly being centered with Intel to take advantage of their leadership and process.”
A New Foundry Giant?
The technology edge Intel enjoys could result in the company, within the next decade, becoming a major player in contract wafer manufacturing although it may not achieve the same level of dominance it has in the PC and enterprise computing space. The Altera deal can help Intel gain credibility in the market thereby attracting even bigger chip manufacturers, according to Altera’s Daane.
“Ultimately I think they have grander ambitions within the foundry industry,” Daane said. “They’d like to obviously take on some other big customers as well. Naturally, because of that I think they are going to try to make us very successful to hold up to others as an example what they can do.
In fact, the company has a greater chance of dislodging the leading foundries in this market than it probably does in the mobile (tablet) computing and wireless communications markets, areas where it has so far failed to gain the kind of traction it wants despite spending billions of dollars in investments over the last years.
Ironically, the strengths – process efficiency and leading-edge/next-generation technology – that Intel deployed so efficiently to sideline competitors in the microprocessor market are the same ones it can rely upon in the foundry sector. In order to maintain a competitive edge, Intel has invested tens of billions of dollars in process technology over the last decades. The company is notorious for raising investments even during downturns on the presumption of increased payoffs once the industry emerged from the recession. Most of the time, it was correct and that strategy helped it sideline Advanced Micro Devices Inc., as a major rival in the microprocessor market.
For a $53 billion revenue company Intel can sometime demonstrate extreme paranoia about its future, a fact noticeable in its investment habits. It’s not that Intel executives believe the company isn’t competitive (they know it is) but that they act as if the sky might fall on their heads if they aren’t investing in processes and products that would keep it far ahead of competitors 10 or more years into the future.
So, Intel is constantly acquiring businesses and product lines and making other strategic investments that the management think would make a difference to its future through Intel Capital. Its R&D and capital equipment expenditure exceeds most rivals and it is constantly pushing into new or adjacent markets – wireless and communications, for example – all in a bid to keep the profit margins healthy.
Intel’s R&D budget ($10.1 billion in 2012 vs. $8.4 billion in 2011) and capex are huge not simply because it believes this would help it win in the market. That was only one part of its calculation. Behind the huge investments is the fear prior Intel leaders have passed onto the next generation of management. Investments in process technologies, they believed, were not only necessary but unavoidable if Intel wants to maintain its leadership position; Failure to invest would have catastrophic implications for Intel.
“We aim to have the best process technology,” Intel said in its 2012 filing with the U.S. Securities and Exchange Commission. “This in-house manufacturing capability allows us to optimize performance, shorten our time to market, and scale new products more rapidly. Most of our competitors rely on third-party foundries and subcontractors such as Taiwan Semiconductor Manufacturing Company, Ltd. or GlobalFoundries Inc. for their manufacturing and assembly and test needs, creating, among other risks, the potential for supply constraints and limited process technology differentiation between competitors using the same foundry.”
The huge investments haven’t always panned out. Intel has spent billions on mobile handset ICs but it is still grasping for a toehold in the sector although it is seeing signs of improvements. It trails products based on IP cores from ARM Plc. in mobile handset and tablets and has made excruciatingly slow advances against the British company, as CFO Stacy Smith acknowledged during a Sept. 4 presentation to investors.
“Today, our share in [wireless] phones is close to zero. Our share in tablets is probably unfortunately 4 percent or 5 percent based on some of the recent designs we've won, like the Samsung Galaxy,” said Smith. “I would expect that in time we’ll grow into significant share and we’ll do it because we can provide more performance and integration.”
Intel’s experience in the foundry market could play out quite differently depending on whether it truly wants to become a major force in the sector. It’s not clear this is Intel’s objective, though, and executives have dismissed suggestions the company is offering foundry services to fill its fabs.
“We don't need foundry to fill fabs,” Smith said. “It's not like we're sitting there with excess capacity, and said, ‘let's go pull in a foundry partner’. So it's a myth that it's a fab filler. We do it, simply put, because we make money at it. It's a good business and it's very consistent with our strategy.”
It may sound simplistic but that strategy can be summed up in two words: paranoia and profit. The first drives the second; the fear of being overtaken in technology, process and manufacturing, resulting in the loss of healthy margins, have kept Intel in fighting shape over the years. Getting into the foundry business simply makes sense on this basis, according to Smith.
“The combination of our IP and manufacturing will pull in as much of the world's production volume as possible into our factories over the next five years. Foundries are a nice adjunct to that,” said Smith. “The recent deal with Altera is a great example. They believe that landing their products on our process technology gives them a significant competitive advantage. They're going to have great products [and] they're going to pay us a nice margin for the manufacturing. It's good for them and it's good for us.”