Summary

• GT Advanced Technologies continues to diversify by entering into the power electronics space.

• GT believes its Hyperion technology could overcome cost barriers and enable silicon carbide lamina deployment.

• The power electronics opportunity and others do not seem to be priced into GT's current valuation.

It was not that long ago that GT Advanced Technologies (NASDAQ:GTAT) operated exclusively in the solar space. In fact, up until August 2011, GTAT traded under the NASDAQ stock ticker symbol SOLR and was known as GT Solar International. Prior to 2011, GT Solar's business model consisted strictly of the manufacture and sale of directional solidification systems furnaces and chemical vapor deposition reactors to produce the low cost, high quality polysilicon, the key raw material used in the creation of photovoltaic wafers.

In 2012, the collapse of the polysilicon and photovoltaic equipment industry, which was largely caused by the oversupply of Chinese manufacturing capacity, nearly brought GTAT to its knees. But for its prescient acquisition of sapphire equipment manufacturer Crystal Systems in July 2010 and the company's considerable efforts to promote the use of sapphire beyond its traditional LED market, GTAT might have ceased to exist as an independent company. Today, however, despite the polysilicon and photovoltaic segments representing just 38.2% of total revenues combined through the first six months of fiscal year 2014 (down from 90.43% a year ago), GTAT's management team believes that the company's extensive research and development efforts should foster growth beyond the company's initial success in the sapphire materials segment.

While GTAT's relationship with Apple regarding the sapphire production in Mesa, Arizona seems to have received the bulk of most analysts' scrutiny in recent quarters, perhaps they should start taking note of GTAT's investments in other material segments. In fact, CEO Tom Gutierrez has reiterated this before. He stated during the most recent fourth-quarter earnings call:

As strong as our sapphire opportunities may be, the GTAT story is not just about our emerging sapphire business. In fact, our entry into the sapphire materials business may enable us to expand into other material segments whilst we have fully ramped the operation in Arizona. In addition, many of the diversification and investment fees that we have planted over the last several years in the LED, power electronics, advanced solar and industrial markets are expected to begin to bear fruits over the next 18 months.

Diversifying Through Innovation

Figure 1: GTAT's Growing Research and Development Since 2010

Despite photovoltaic and polysilicon revenues all but drying up in fiscal year 2013 (declining nearly 51% from the prior 12-month period), GTAT has made the calculated gamble to continue to innovate its way back to profitability. As evidenced by Figure 1 above, despite the sharp decline in the company's total revenues, it is clear from the consistent growth in the two R&D metrics highlighted above that GTAT remains dedicated to developing, as CEO Gutierrez describes, "a balanced company that is diversified." GTAT hopes the company's investment in alternative revenue streams will mitigate its exposure to the cyclicality of the of upstream solar equipment industry.

In addition to sapphire, the material heavily discussed in connection with GTAT's Apple engagement and activity in the LED space, one of the other potential revenue streams that could add significant value to GTAT in the future is silicon carbide. Silicon carbide could have immense implications in the power electronics space. In fact, but for the material's lack of commercial viability because of cost (which GTAT thinks it can help solve) and processing difficulty, silicon carbide is regarded as a superior semiconductor to the silicon used today in most power electronic semiconductors. The silicon carbide material has several advantages over silicon semiconductors employed in power electronics today. These advantages over silicon, which are identified in Ahmed Elasser's and T. Paul Chow's "Silicon Carbide Benefits and Advantages for Power Electronics Circuits and Systems," include enhanced heat dissipation, greater efficiency at higher temperatures, and improved switching frequency.

Despite the processing and cost-reducing challenges that limit silicon carbide's current adoption in power electronics, CEO Gutierrez has suggested that GTAT is positioned to make the superior semiconductor material commercially viable. This sounds familiar. In fact, similar to the approach it took with polysilicon and sapphire, GTAT believes that it can be an enabler of low cost and high volume silicon carbide.

The Potential of Silicon Carbide

To understand Gutierrez's efforts requires some appreciation of silicon carbide's potential and its chief pitfall. For at least two decades, silicon carbide has been considered the successor to silicon. The expectation has been that silicon carbide's thermal, band gap and efficiency advantages would cause it to be designed into higher voltage products. Researchers and designers have always been intrigued by silicon carbide because of its tenfold higher breakdown strength and threefold greater band gap and thermal conductivity over its less costly silicon cousin.

Figure 2: Potential Silicon Carbide End-Markets

Silicon carbide has become increasingly designed into higher efficiency solar inverters, but it has not yet gained much traction elsewhere. However, with silicon carbide increasingly undergoing testing in a variety of industrial and commercial settings (as identified above in Figure 2), there is a rapidly growing tide of silicon carbide success stories. Japanese industrial and consumer product giants particularly seem to be at the forefront of the move to displace conventional silicon with higher efficiency silicon carbide solutions. According to John Boyd's "Silicon Carbide Ready to Run the Rails,"in a trial of silicon carbide in underground Japanese rail cars, Mitsubishi was able to generate 38.6% in overall energy savings by displacing conventional silicon designs with more compact and efficient silicon carbide-based inverters and regenerative breaking systems. Mitsubishi is also developing silicon carbide inverter solutions for elevators and electric vehicles. Moreover, Toyota recently announced considerable progress in using silicon carbide as a replacement for silicon in hybrid and electric vehicle power control units. Per Green Car Congress, silicon carbide enabled Toyota to design a power control unit that was 80% smaller than the silicon version, improved fuel efficiency by 5% based on test trials, and dramatically reduced electrical power loss.

As reported in technical papers and elsewhere, the performance advantages of silicon carbide have never been in doubt. A 2011 article written by Burak Ozpineci and Leon Tolbert, aptly titled "Silicon Carbide: Smaller, Faster, Tougher," describes the advantages of silicon carbide and sets forth the case for silicon carbide supplanting silicon in hybrid cars and the electric grid. The authors explained:

Because electrons in silicon carbide require more energy to be pushed into the conduction band, the material can withstand much stronger electric fields, up to about 10 times the maximum for silicon. As a result, a silicon carbide-based device can have the same dimensions as a silicon device but withstand ten times the voltage. What's more, a silicon carbide device can be less than a tenth the thickness of a silicon device but carry the same voltage rating, because the voltage difference does not have to be spread across as much material. These thinner devices are faster and boast less resistance, which means less energy is lost to heat when a silicon carbide diode or transistor is conducting electricity.

With all the advantages that silicon carbide enjoys over silicon, one would expect it to be rapidly deployed. The primary problem is cost. Four-inch silicon wafers still sell for $1,000 or more. That cost is expected to drop over time, but perhaps not at rapid rates. According to "Efficiency Improvement with Silicon Carbide-Based Power Modules" by Zhang Xi, Daniel Domes, Roland Rupp, and Infineon Technologies, the industry anticipates a 10 percent decline in price may be possible over the next three years.

Despite the high costs, some market analysts expect that silicon carbide's deployment in semiconductors should soon accelerate. According to market research expert Yole Development, silicon carbide is expected to grow at a 39% rate so as to increase nearly ten fold (off of 2012 numbers) to a several billion dollar industry.

Figure 3: Growth of SiC Power Electronic Devices

GTAT's Strategy: Setting Up an Eco-System for Silicon Carbide Lamina

Although GTAT has been relatively quiet on its progress in the power electronics space, Gutierrez has repeatedly claimed in quarterly earnings calls that GTAT's ultimate goal is to help solve the cost barrier problem and enable new market participants. So far at least, GTAT has not reported any equipment sales into the space and its only offering is the SiClone 100, a furnace producing 4-inch silicon carbide wafers.

Despite the lack of tangible results thus far, GTAT believes that it can enable the substantial reduction in silicon carbide production costs. Its tool for accomplishing those cost reductions is the Hyperion ion implanter. That equipment, which GTAT has an extensive series of patents upon, enables ions to be implanted in a thin layer of silicon carbide or other materials like sapphire. It is then heated and exfoliated, or essentially peeled away. GTAT has expended extensive time and effort in perfecting the latest generation of the Hyperion, which it acquired from Twin Creek Technologies, Inc. To date, 200 man-years of research and development have been invested and 55 patents have been issued, with another 25 patents pending on technology related to Hyperion. GTAT claims the Hyperion, which will be commercialized in 2015, is a mature product that is ready to be deployed now. By pairing the Hyperion with silicon carbide, GTAT anticipates a near order of magnitude (10x) in cost reductions for the advanced material. In an August 13, 2014 investor event with Canaccord Genuity, Gutierrez suggested that silicon lamina wafers bonded to other less costly materials might permit the $1,000 wafer price to drop to as little as $250 to $300.

In its March 2014 "New Product and Technology Briefing," GTAT described in some detail how the Hyperion serves as the core tool for the producing super thin lamina (25 to 50 microns in the case of silicon carbide) that can then be bonded to cheaper materials. In the company's most recent earnings call, Gutierrez seems optimistic of the Hyperion technology's impact in the silicon carbide space. He stated, "We have exfoliated 32 micron think 2-inch [silicon carbide] foils. We are not aware of any freestanding silicon carbide lamina of this thickness been produced by anyone else in the world to-date." To put that in perspective, 32 microns is about one-half the thickness of average human hair. The goal of the lamina production system is to obtain the advantages of silicon carbide's superior physical properties but at reduced costs. GTAT suggested that the reduction in material use, waste and surface finishing costs promises to decrease costs dramatically. These efforts are far enough along that GTAT has disclosed that it is currently working with an unidentified tier 1 silicon carbide device manufacturer.

GTAT announced on April 30, 2014 certain initiatives designed to leverage its considerable investment in Hyperion. At that time, GTAT disclosed a memorandum of understanding in which GTAT will collaborate with EV Group, an Austrian based equipment manufacturer, to harness EV Group's specialty bonding expertise to design silicon carbide solutions for next generation consumer and industrial products. The goal is to enable high volume production processes and equipment to bond ultra thin sapphire and silicon carbide lamina to glass, silicon and plastics. The announcement may suggest that GTAT intends to put into place a more complete eco-system to leverage Hyperion in the sapphire and silicon carbide space.

It appears from a recently awarded patent that GTAT believes that its silicon carbide lamina strategy can enable expanded uses of silicon carbide. On January 30, 2014, GTAT was awarded patent WO 2014018462 A1 describing a method for exfoliating ultra thin silicon carbide lamina. The background section of the patent suggests a need for producing silicon carbide and diamond-like carbon materials cost efficiently. Although the Hyperion is not expressly mentioned, the patent describes a method for implanting ions and exfoliating thin layers of lamina from silicon carbide. The patent explains that this process avoids the time, cost and waste of having to grind down silicon carbide to the desired thinness. The invention is described as having application in the manufacture of LEDs and electronic devices. Notably, the patent description states that silicon carbide enables greater lumen production capacity in LEDs than standard sapphire solutions.

Interestingly enough, the current leader in the silicon carbide market is Cree, the North Carolina-based LED manufacturer. Despite silicon carbide's cost drawbacks, Cree is a proponent of the material's adoption in LED applications over sapphire, which currently dominates the market, based on silicon carbide's proven advantages in luminosity, useful life, and defect rate.

Should Gutierrez's assertion hold true that GTAT's Hyperion technology "enables thin wafer cost structures that are nearly an order of magnitude below today's wafering processes," concerns over the cost of silicon carbide may be significantly muted. If true, it seems reasonable to expect silicon carbide to capture much more than the 18% market share of all LEDs by 2020 that is projected by Carole Jacques in "Epi-Wafer Market to Grow to $4 Billion in 2020 as LED Lighting Zooms to $80 Billion."

Investor Takeaways

While it may still be too early for GTAT to provide much more detail on its power electronics roadmap, longer-term investors might be encouraged by the EV Group collaboration and the silicon lamina patent award. That evidence suggests that GTAT may be putting together a more comprehensive ecosystem that will support silicon carbide device manufacturers but also enable higher lumen LEDs. Given the claimed operational readiness of the Hyperion, the power electronics initiative that GTAT first spoke of in mid-2012 may be further along than the market generally appreciates and may bear fruit sooner than perhaps expected.

In fact, investors should anticipate that general market growth forecasts likely do not take into account GTAT's role in improving the commercial viability of silicon carbide. Based on the information provided in earnings calls and various reports, it seems reasonable to expect more robust growth rates should GTAT's Hyperion technology prove to materially drive down the costs of silicon carbide wafers.

Finally, GTAT's limited discussion of silicon carbide has assumed that the company will be only an equipment provider. The EV Group announcement did not dispel that notion. However, GTAT has stated publicly that it would not rule out other materials business initiatives apart from sapphire. Given the proprietary technology behind the Hyperion and its potential, it might not be terribly surprising if GTAT were to at least consider pursuing silicon carbide production itself. The fact that GTAT has disclosed that it is working with device makers (rather than wafer producers) might suggest that GTAT would consider funding its own materials business or entertain a partnership to do so.

All said, it is nearly impossible to quantify the value of GTAT's power electronics initiative. Goldman Sachs has a $20 price target on GTAT based almost entirely on its sapphire business, which it had viewed to be worth not less than $15 per share. Given GTAT has recently been trading in the $15 - $17 range and given the general under-appreciation of the potential of silicon carbide, it seems fair to assume most investors are ascribing little to no value to GTAT's power electronics business. Undoubtedly, the market is generally skeptical of the claim that Hyperion enables dramatic cost reductions. Indeed, GTAT faced rampant skepticism that it could cost reduce sapphire sufficiently to enable its use as a cover screen for phones. While some degree of investor skepticism is warranted when any new technology is claimed to be "game changing," the market currently seems to have completely discounted the Hyperion-enabled silicon carbide opportunity. As this article hopefully helps more patient and longer term investors appreciate, the price of GTAT simply does not reflect the inherent value of GTAT's silicon carbide enabling opportunity (and perhaps several others).