by James Montgomery, News Editor, Solid State Technology
June 8, 2008 – It’s something that many ASM International shareholders have been waiting for, after several years of persuasion and tactics to get the company to slough its struggling frontend business away from its profitable backend business — Applied Materials has made a verbal unsolicited $400M-$500M offer to buy its atomic layer deposition (ALD) and plasma-enhanced chemical vapor deposition (PECVD) businesses. ASM execs and board members said they would mull the offer and respond publicly by June 14. Applied declined to offer any comment to WaferNEWS other than what was in the brief press release.
In the meantime, WaferNEWS asked Gartner research VP Dean Freeman to help brainstorm the most likely reasons behind AMAT’s offer (and some unlikely-yet-interesting-but-good-luck-proving-it possibilities), and what it means for ASMI.
The most likely scenario is that Applied simply wants to pick up some additional technology for its portfolio: atomic layer deposition (ALD) and high-k technology, and a different method/chemistry of depositing low-k dielectrics. “One thing that Applied doesn’t have — or they haven’t confirmed that they have — is an ALD product,” Freeman told WaferNEWS. Picking up ASMI’s ALD technology could be a maneuver to avoid high-k licensing problems too, he added. From a PECVD and low-kstandpoint, the ASMI deal would give AMAT a different technology that it didn’t have, padding its 60%-70% marketshare in PECVD and 80% in low-k — maybe a $15M/year slice of a $1.1-$1.2B market, though, he noted, so not really a significant addition.
It’s possible but less likely that AMAT is also interested in ASMI’s new ALD processes for single-metal high-k gate stacks, he noted. “From what I understand, the metal gate is sputtered for the most part because the quality of ALD for metal gates is not as high,” Freeman said. “Device performance isn’t quite as good as you get from a sputtered film.”
Perhaps this move is not so much about technology, but about market consolidation. AMAT has made no secret in the past about its desire to be an industry consolidator — is this, then, a ploy to spark a new wave of market/industry consolidation and force competitors to react? (And/or simply coldly eliminate a competitor?) The ALD market for high-k, while not large now, has a projected ~14% CAGR as NAND flash chipmakers start using high-k gates, Freeman noted. “It could be Applied saying, ‘We’ll take a competitor out of the market,'” Freeman said. Though ASMI and these units are comparatively small, they do give AMAT enabling technology (high-k) and maybe some low-k patents to further those efforts going forward.
Or is there a customer play here? ASMI was widely reported to have won Intel’s ALD business for its 45nm HK+MG chips (though it probably would give AMAT all of Intel’s PECVD business). AMAT, of course, already does a good deal of business with Intel, but perhaps even more is better — adding the ASMI tech would probably would give AMAT all of Intel’s PECVD business, Freeman noted.
For the more adventurous industry handicappers, there may be other possible reasons for AMAT to make this deal. Freeman pointed out that in the past month, UK-based Hermes Focus Asset Management Europe Ltd., an 11% ASMI stakeholder, launched its own turnaround proposal for the company, including a suggested new CEO: Farhad Moghadam, formerly head of AMAT’s thin films product business group who left when Tom St. Dennis returned in July 2006. It’s possible AMAT wants to prevent one of its own strong former leaders taking the helm at a competitor, Freeman speculated.
Also, speaking of major customer Intel, it “has been known to create marriages in the past,” Freeman noted. ASMI’s frontend group has roughly €430M-€440M in sales, maybe €100M in ALD and low-k, and the bulk of that in furnace work, he pointed out. “It could be that Intel wants more of a focus on low- and high-k,” he said, and is leveraging another key supplier to reshape that emphasis.
Whatever the motivation, clearly some vendors will have to sit up and take notice. “In low-k the landscape doesn’t change that much, you still have Applied and Novellus,” Freeman noted. And with high-k, “ASMI really only served the logic market.” What may shake things up is if Applied leverages the added ASMI technology to crack into the memory segment, notably NAND flash — where Jusung serves top customer Hynix, and TEL is likely working with Toshiba (and may have a high-k logic metal gate system in the field), Freeman said. In fact, such a move for ASMI’s technology actually would make more sense for TEL, he said — reinforcing the notion that the deal is at least in part a strategic flanking maneuver.
What about the proposed deal from ASMI’s perspective? The company has been battling disgruntled shareholders for years over how to best deal with its two businesses: a slumping frontend vs. a thriving Singapore-based backend. In late April the company announced its latest plan to return its frontend business to profitability (and silence critics), by exiting the rapid-thermal processing (RTP) market and get the unit’s losses off the books by year’s end.
ASMI CEO Chuck del Prado is on record saying the company will entertain the offer at least as an obligation to investors, but also expressing pleasure that the AMAT offer suggests a valuation for the units that is “very different from what the market was assigning to them.” Investors clearly like this new angle of AMAT (or perhaps anyone with a serious intent) taking the profit-dragging frontend business, which has been losing money since 2005, off their hands — ASMI shares spiked as much as 24% on Friday after the announcements.
If ASMI agrees to sell its ALD and PECVD to AMAT, that leaves just its furnace business as a frontend business unit, which “is very competitive with low margins,” Freeman noted. “In epi, you’re up against AMAT which is very difficult.” If ASMI decides to complete a frontend separation and sell that final part too, who would take it? Possibly Hitachi Kokusai and TEL, and maybe even Aviza (who just got out of that market), Freeman noted. ASMI was #2 in that market, and had made good inroads into Taiwan memory companies. Hitachi Kokusai might also have an interest in the epitaxy side, too. — J.M.