European players of the semiconductor were gathered at a symposium on technology in development. A debate on the value of investing in 450 mm wafers has greatly divided the members present. The big names were present. NXP, STMicroelectronics and even Globalfoundries who have factories in Ireland and Germany.
Remember, the 450 mm wafers are expected to increase yields and lower production costs to produce more chips at once. The problem is that this technology now costs very expensive. In addition to obtaining wafers that are more difficult to produce, requires expensive new equipment (see "Miniaturization of transistors and larger wafers: understanding technological issues") Many actors believe therefore that Europe does not to invest taxpayer money in research and development of an engraving on this type of wafer.
According to them it is better to put that money on MEMS or CMOS derivatives. The issue is of paramount importance, since a miscalculation grow semiconductor production outside of Europe, which will inevitably negative consequences for countries of the old continent and their founders. Jean-Marc Chery, a director at the Franco-Italian STMicroelectronics, seems to have summed up the situation.
According to him, one must first ask where does the market. If he moves towards mobile products, while the 10 nm on 300 mm wafers is sufficient to make the transistors necessary. Clearly, he favors the fine print that has a direct impact on the performance of the chip, to the detriment of the 450 mm which affects only the production costs, but today does not necessarily guarantee a lower price the transistor.
This position seems to be also held by Intel, which announced the construction of a new plant last month (see "Fab 42: Intel's new plant to $ 5 billion) that will burn in 14 nm, which is open in 2013, but only use 300 mm wafers. A manager of Intel Ireland moreover spoke, but remained cautious on the subject, as the representative of Globalfoundries.
Concretely, this movement of resistance seems to confirm the delay of 450 mm for the benefit of the new manufacturing process more complex. Intel is expected to be among the first on this technology, because its position allows some largesse, but we do not expect widespread adoption of 450 mm in the next five years. 
Remember, the 450 mm wafers are expected to increase yields and lower production costs to produce more chips at once. The problem is that this technology now costs very expensive. In addition to obtaining wafers that are more difficult to produce, requires expensive new equipment (see "Miniaturization of transistors and larger wafers: understanding technological issues") Many actors believe therefore that Europe does not to invest taxpayer money in research and development of an engraving on this type of wafer.
According to them it is better to put that money on MEMS or CMOS derivatives. The issue is of paramount importance, since a miscalculation grow semiconductor production outside of Europe, which will inevitably negative consequences for countries of the old continent and their founders. Jean-Marc Chery, a director at the Franco-Italian STMicroelectronics, seems to have summed up the situation.
According to him, one must first ask where does the market. If he moves towards mobile products, while the 10 nm on 300 mm wafers is sufficient to make the transistors necessary. Clearly, he favors the fine print that has a direct impact on the performance of the chip, to the detriment of the 450 mm which affects only the production costs, but today does not necessarily guarantee a lower price the transistor.
This position seems to be also held by Intel, which announced the construction of a new plant last month (see "Fab 42: Intel's new plant to $ 5 billion) that will burn in 14 nm, which is open in 2013, but only use 300 mm wafers. A manager of Intel Ireland moreover spoke, but remained cautious on the subject, as the representative of Globalfoundries.
Concretely, this movement of resistance seems to confirm the delay of 450 mm for the benefit of the new manufacturing process more complex. Intel is expected to be among the first on this technology, because its position allows some largesse, but we do not expect widespread adoption of 450 mm in the next five years.
- Shift to 450 mm wafers may wait two nodes for the 10 nm node in 2015-2017 according to Intel researcher (28/02/2011)
- EUV lithography needs at least 80 wafers per hour and is targeting a 250 Watt light source (03/03/2011)
- Crystal Solar: Vapor to Wafer to Cut Solar Costs (17/02/2011)
- Silicon Solar Cells Ditch the Wafers (22/02/2011)
- Wafer-level CNT growth on CMOS substrates (19/01/2011)
No comments:
Post a Comment