Based on the annual revenues reported I summarized the situation going from 2015 through 2016 in the plot above. Please find a summary and additional comments from the article below going down the ranking fro all companies relevant to ALD:
Applied Materials grew 26.2%, due to growth in 3D NAND and leading edge logic devices (FinFET) and multiple patterning processes. However, reportedly not in ALD, which remains to be confirmed. Some of that huge growth must be ALD and possibly growth fro the Olympia ALD platform.
Lam Research made a solid growth of 7.5% gaining share against Applied Materials in deposition, as its 3D NAND shipments grew more than 80%. reportedly primarily due to growth in ALD.
Tokyo Electron was fourth, gaining 15%, primarily in CVD deposition processes. How much of this growth in CVD is actually ALD and the ratio singel/multi wafer vs. furnace was not reported. Others have claimed that Tokyo Electron is taking market share in PEALD from ASM with their NT 333 SiO2 ALD platform that relizes high speed ALD cycles to achieve system throughput of over 100 wafers per hour for e.g. SADP and SAQP. Please see a recent blog covering this tool and topic by Prof. Kessels (LINK).
TEL NT333 High quality SiO2 ALD tool (LINK)
ASM International dropped 6.6% in 2016 due to less investments by the fabs in single wafer ALD. Single wafer ALD is mainly used for the High-k / Metal Gate Stack which until today only comes ones per wafer. ASM single wafer is also deployed for spacers and multiple patterning using the ASMI high productivity XP8 platform with dual Chambers. Apparently, the other platforms with four, five or more wafers per chamber are more productive and used for multiple passes per wafer in multi pattering and cost sensitive memory products.
Hitachi-Kokusai dropped to but remained in Top 10 loosing market chares to its furnace competitor Tokyo Electron.
Korean ALD suppliers Jusung Engineering and Wonik IPS gained market shares in ALD - how much was not reported.
Clearly the growth 2016 was in multiwafer ALD driven by patterning and 3DNAND but still a lot of question marks on the ALD situation and I look forward to the outcome from the Gartner report due soon I think.