By Joanne Itow
Today, foundries supply more than 20% of the silicon used to produce all the semiconductor products sold. The foundry impact has grown from only 10% in 1997 to 24% today. The significance of foundries is even more evident when focused on logic wafers alone.

Figure 1. Foundry Wafers as a Percent of Total, IC’s, and IC’s Minus Memory

Source: Semico Research Wafer Demand Model

When discretes and memory products are subtracted from the total, the impact of the foundry business model increases to more than 40%. For the top foundries, the big revenue generators come from the high-volume advanced technologies. Foundries are making a much bigger dent in the advanced technology arena. In 2012, more than 70% of the total advanced technology logic wafers were run at foundries.

Figure 2. Foundry Wafers as a Percent of Total Advanced Logic Wafers

Source: Semico Research Wafer Demand Model

TSMC, GlobalFoundries, Samsung and UMC are the main providers of manufacturing capacity for products that require advanced process technologies. The benefits of going fabless or fab-lite became very clear for semiconductor manufacturers prior to the transition to the 45/40nm process technology. Texas Instruments made the switch for its designs at the 45/40nm node. AMD sold off its fabs before ramping up its 32nm process technology.

Supplying leading-edge technology can reap big benefits, but it is adds high risk. It is costly to develop advanced technology and build the fabs. TSMC still reigns as the leading foundry supplier and outranks its closest competitor by almost 4X revenues. The big foundries all are targeting customers in need of large volumes and cutting-edge technologies. So how do the other foundries compete when market share is weighted so heavily in favor of one supplier and the risks continue to rise?

GlobalFoundries is rolling out technology as quickly as it can. Admittedly, the company stumbled a bit with its 32nm HKMG, but it’s once again on a very aggressive roadmap to be the first to deliver a low-power FinFET option. Its 14nm-XM process is on the roadmap for Q1 2014 risk production. The Common Platform continues to deliver. Samsung has successfully taken the jointly developed Common Platform technology to produce two generations of foundry products for its highly recognized mobile customer. However, there are risks to being so highly visible. That customer is also highly sought after and others want to take that business away. How is that done? Deliver technology options earlier, provide better service, offer lower prices, or all of the above.

Intel is entering the foundry space as quietly as it can. As the largest semiconductor company with the most advanced fabs, it’s debatable whether Intel really can do anything quietly? With just a few foundry customers, the company is delivering products using its 22nm FinFET technology. Many speculated that Achronix would tape out its Speedster22i product at the end of 2011, at the same time as Intel, and then move to production in 2012. The company is one year behind that schedule, but it still holds the distinction of being the first fabless company with working silicon utilizing FinFET technology—and the first at 22nm. But is that enough to be successful?

Intel’s foundry customers jumped from the TSMC track and took the risk of skipping directly to a 22nm FInFET technology. TSMC’s early success was built on a few fabless customers who made it big in their niches. For now, Intel appears to be selectively picking customers and is willing to provide custom foundry solutions to fit the customer’s needs. Intel has the luxury of concentrating on a few small volume customers since it can use its own products to test the technology and fill its fabs.

At the most advanced technologies, there are numerous factors necessary to be successful. Samsung and Intel can offer a fully integrated foundry solution from design to fully packaged product. That’s pretty enticing when advanced packaging is so critical for an optimized solution. One thing is for sure—the Common Platform and GlobalFoundries seem to have the biggest ecosystem of anyone out there. If you can’t find what you need through one of GlobalFoundries’ partners, it’s probably not available.

But the bottom line is that for a foundry to be successful at the most advanced nodes it has to have successful foundry customers. That’s as true for Intel as it is for Samsung, and for TSMC, GlobalFoundries and UMC. Samsung won big with Apple, but how many companies like Apple are out there? Even Apple may not measure up to its past achievements in the future. And at each successive process node, as costs, complexity and risks increase, the unanswered question is which foundries will have enough successful customers to continue down this path.

—Joanne Itow is managing director of manufacturing at Semico Research.

By Joanne Itow
The industry’s quest to continue on the semiconductor roadmap defined by Moore’s Law has led to the adoption of a new transistor structure. Whether you call them finFETs, tri-gate or 3D transistors, building these new devices is difficult. But the technology is only half the challenge.

In 2002, Chen Ming Hu* spoke at the Semico Summit. The title of his presentation was “The Future of Semiconductor Scaling.” In 2002, Hu pointed out that TSMC already had fabricated 35nm CMOS FinFET transistors on TSMC production tools and expected the technology to be scalable to 9nm. Ten years has elapsed and Intel is the first to run 22nm Ivy Bridge products using tri-gate technology. From a technical perspective, FinFETs are now proven to be executable in volume production. But how should we look at this from a market perspective?

When will the foundries be ready to ramp FinFET technology in volume production? And more importantly, when will foundry customers realistically be ready to fully utilize the technology? Until recently, the most advanced dedicated foundries were planning to introduce FinFETs on their 14nm technology. It appears that schedule has been accelerated. TSMC is talking about a 16nm transition node with FinFETs. GlobalFoundries just announced the development of ARM low-power processor designs for 20nm and finFET process technologies targeting SoCs including graphics processors. This doesn’t commit GlobalFoundries to finFETs at 20nm, but it does offer some options.

Semiconductor units continue to grow, and wafer demand increases right along with it. Any new technology that enables improved or new electronic applications should be implemented. Unfortunately, it’s not that easy. New consumer markets are cost-sensitive. The introduction of innovative technologies must take into consideration market competition, cost to volume, and opportune applications.

The following graph presents wafer demand assuming finFET adoption at 20nm versus 14nm. Although this graph depicts an “all-or-nothing” scenario, if finFETs are adopted by the industry at 22nm/20nm it could mean a significant difference in the total wafers used to produce finFET products.

The following graph adds in a line comparing finFET adoption with the adoption of HKMG. Intel was also the first to implement HKMG at the 45nm node. GlobalFoundries and Samsung introduced HKMG at 32nm. TSMC began ramping HKMG with its 28nm production in 2011, almost four years after Intel.

The smartphone and tablet markets continue to grow and the ultra-mobile PC market is beginning to take off. When the foundries transition to 20nm, the capacity ramp will have to be much larger than the ramp for HKMG at 32nm/28nm. Filling those market needs will be challenging. When 28nm was introduced at TSMC, customers could opt for a poly/SiOn gate stack versus the HKMG. That had a dampening effect on the volumes of HKMG wafers during the early ramp of 28nm. As companies begin to roll out 20nm products, once again there are a few options. Some companies may switch to finFETs. TSMC plans to stick with a bulk planar structure. As an alternative to bulk planar, there is also a fully depleted SOI option. Whether finFETs are introduced at 20nm, 16nm or 14nm, Semico believes their ramp will prove to be another test of the foundry model.

For additional data on wafer demand by product by technology, please visit Semico’s website for a current list of manufacturing studies.

*Chen Ming Hu is currently Distinguished Professor of Microelectronics at the University of California, Berkeley. At the time of his presentation at the Semico Summit in 2002, he was the Chief Technology Officer for TSMC.

—Joanne Itow is managing director of manufacturing at Semico Research.