Demand for quality goods and PRC economic prospects
The demand for more expensive, as well as more reliable products could be encouraged by electronic technologies development slowdown. The slowdown means longest terms of moral depreciation of electronic devices and, therefore creates additional demand for durability and reliability.
According to the Moore’s law (ML), formulated by G. Moore from Intel decades ago, density on chips (or, in simple terms, computer performance) doubles each 2 years. Though several obstacles on the path of the above law were encountered during the decades that passed, until now the rumors about its close death have been from time to time greatly exaggerated. The following signs indicate, however, that the main candidate for the ML extension – EUVL (extreme ultraviolet lithography) – is not going to be ready for mass-production in 2014 as wished.
1) No reported progress on EUV sources since Jun 2010, when ASML NXE:3100 tool with 20W-power was introduced. This is in spite of the fact that more milestones (with source power up to 100-200W) were anticipated then (06.2010) till the end of 2011.
2) While 13.5nm EUV is far from being at hand, EUVL community speaks loudly about the next 6-7nm step. According to the author’s experience, this is a well-known and clearly alarming sign.
3) At the last EUV Sources’ workshop (Dublin Nov 07-09) a special session on alternative EUV sources (mainly Free Electron Lasers – FELs) was held. FELs are expensive, never proven in industrial conditions and demand change of the fab logistics (1 FEL source – multiple tools, instead of the present paradigm 1 source – 1 tool). The interest of industry in FEL shows IMHO their near-despair regarding the heavily-invested plasma sources.
4) The language of many titles and abstracts of the last International Symposium on EUVL (Miami, Oct 13-16 2011) is rather alarming E.g.:
EUV Masks: Ready or Not?
There are many challenges in EUV mask which needs innovative technologies.
While incremental progress has been made …, the rate of learning is not trending to meet the commonly perceived requirement in time for production.
Substantial progress, especially on source power, is still required before EUVL can be applied to high-volume manufacturing.
Dr. Yehoshua Socol www.FalconAnalytics.com