Robert Maire
Moderator
Industry Tone remains positive on 3D NAND & TSMC. H2 on track better than H1- 10NM build outs Small/Mid caps positive offsets missing headliners.
Despite the lack of analyst meetings from AMAT, KLAC & LRCX, there is more than enough data from other industry players to confirm the long held view of a better H2 for 2016. Though this comes as no surprise its always good to get further confirmation that there is no change in this outlook especially in light of macro unrest.
The data points are largely unchanged- Strong 3D NAND, TSMC ramping 10NM, DRAM dragging, Samsung somewhat subdued and Intel also light. Of these data points we would even say that 3D NAND may be even more enthusiastic in the near term.
It sounds like everyone has the quarter in the bag as most company management made cheery comments about the quarter while not specifically commenting on the numbers while in their blackout period.
TEL rebuilds....
The largest company at the show, Tokyo Electron, set some lofty goals as it is in the middle of a reorganization following the Applied breakup at the alter. The restructuring and management changes have delayed things more than Applied's similar recovery as Japanese companies tend to be a little slower and the added reorg further delayed that.
We think that it will be hard for TEL to regain momentum in etch and going into the AMAT hook up TEL had declining market share. We think it is most important for TEL to fix its financial model, much as Applied has been doing, to try to get earnings growth on weaker revenue growth through better margins and cost controls.
Value shift from front end to back end...
Our discussions at the show underscore our view that as Moore's law progress in the front end slows due to asymptotically reaching limits of lithography and materials science that the back end will increasingly take on more value as end users , such as Apple, continue their desire to jam more transistors into smaller spaces with the only way to achieve that goal is through back end processes.
This shift in value is similar to the shift in value from lithography to deposition and etch as lithography progress stalled due to EUV's delay. Much as dep and etch took up the slack in the font end we will now see back end processes take up the overall difficulties in the front end.
We will put out further reports on this after the show as this is an important topic.
10NM lite??....
It sounds a lot like 10NM is going to be one of those "lite" nodes much like 20NM where the industry blew through it without really putting down any roots. I continue to hear more and more buzz about 7NM plans and the feeling I get is that 7NM will be more solid than 10NM. GloFo may be the poster child of this strategy as they have decided to skip 10NM altogether. In their case, we think this is a wise decision because you can't make any money in this business by being a fast follower...the leaders are the money makers and GloFo can't wait for hand me downs from Samsung. So far spending on 10NM seems good but some of that spend may be more directed at 7NM capability.
We will watch for evidence of 7NM being a potentially bigger spending node.
Any hope for DRAM?.....
Though the industry would like to see s resurgence in tool demand for DRAM we can find no evidence of a spending rebound. Memory spending continues to be focused primarily on NAND as the demand is much more price elastic and the takeover of the disk drive market continues at full speed with a seemingly bottomless pit demand for more and cheaper flash storage
Retro is in! 200MM and 28NM rule...
Demand for old 8 inch tool sets sounds great as older fabs continue to do well with internet of things which uses trailing edge capacity. This is perhaps one of the best times we have seen in the industry for trailing edge capacity.
We think that aside from service and spare parts that companies will likely see further benefit from older tool sales that have seena resurgence.
28NM is the gift that keeps on giving..... this is the best node in recent times and perhaps the best node of all times as a lot of product continues to be manufactured at 28NM. Having the lowest cost per transistor helps. It is one of the last nodes before the industry got deeply mired in multi patterning.
We hear reports of continued good tool sales. Both 200MM demand and 28NM demand are overall positive for the industry as they help support the overall revenue stream of the company and are a bonus on top of leading edge sales.
Tower Jazz is the retro semi poster child for IOT
One of the companies that is the prime example of retro in the semi market is Tower Jazz. It is quickly becoming (and perhaps already is) the TSMC of trailing edge capacity. Their fabs are filled up with a very steady stream of very solid business and their costs are very low with cheap/almost free fabs and tools. Management has done a great job rolling up this part of the industry while no one was looking.
While IOT doesn't do anything for leading edge capacity it does a great job of sucking up older fab capacity for low cost processors, controllers, sensors and SOCs and all those things that fill up cars, drones, thermostats etc;.
While trailing edge may not be as sexy as leading edge 7NM , the profitability makes it quite attractive....
The Stocks...
The semi equipment stocks have been on a roll and if anything are likely a bit ahead of themselves at this point. Even the management of a number of companies we spoke with were surprised by the strength of the stocks. Many are at all time highs and continue to move higher on light news.
We might consider trimming some of holdings in some of the stocks that have outdone to market as we wonder what will support the strength post Semicon and going into August.
However we would want to make sure we have a good position in the names for the Iphone 7 rollout expected in late September as that will renew interest after a slower August
About Semiconductor Advisors LLC
Semiconductor Advisors provides this subscription based research newsletter, Semiwatch, about the semiconductor and semiconductor equipment industries. We also provide custom research and expert consulting services for both investors and industry participants on a wide range of topics from financial to technology and tactical to strategic projects. Please contact us for these services as well as for a subscription to Semiwatch
[table] style="background-color: #3194D6; margin: 0px 0px 0px 0px; width: auto !important"
|-
| rowspan="1" colspan="1" align="left" valign="top" style="color: #000000; font-family: Calibri,Helvetica,Arial,sans-serif; font-size: 11pt; padding: 8px 15px 8px 16px" | VISIT OUR WEBSITE
| rowspan="1" colspan="1" style="color: #000000" |
|-
[/table]
Despite the lack of analyst meetings from AMAT, KLAC & LRCX, there is more than enough data from other industry players to confirm the long held view of a better H2 for 2016. Though this comes as no surprise its always good to get further confirmation that there is no change in this outlook especially in light of macro unrest.
The data points are largely unchanged- Strong 3D NAND, TSMC ramping 10NM, DRAM dragging, Samsung somewhat subdued and Intel also light. Of these data points we would even say that 3D NAND may be even more enthusiastic in the near term.
It sounds like everyone has the quarter in the bag as most company management made cheery comments about the quarter while not specifically commenting on the numbers while in their blackout period.
TEL rebuilds....
The largest company at the show, Tokyo Electron, set some lofty goals as it is in the middle of a reorganization following the Applied breakup at the alter. The restructuring and management changes have delayed things more than Applied's similar recovery as Japanese companies tend to be a little slower and the added reorg further delayed that.
We think that it will be hard for TEL to regain momentum in etch and going into the AMAT hook up TEL had declining market share. We think it is most important for TEL to fix its financial model, much as Applied has been doing, to try to get earnings growth on weaker revenue growth through better margins and cost controls.
Value shift from front end to back end...
Our discussions at the show underscore our view that as Moore's law progress in the front end slows due to asymptotically reaching limits of lithography and materials science that the back end will increasingly take on more value as end users , such as Apple, continue their desire to jam more transistors into smaller spaces with the only way to achieve that goal is through back end processes.
This shift in value is similar to the shift in value from lithography to deposition and etch as lithography progress stalled due to EUV's delay. Much as dep and etch took up the slack in the font end we will now see back end processes take up the overall difficulties in the front end.
We will put out further reports on this after the show as this is an important topic.
10NM lite??....
It sounds a lot like 10NM is going to be one of those "lite" nodes much like 20NM where the industry blew through it without really putting down any roots. I continue to hear more and more buzz about 7NM plans and the feeling I get is that 7NM will be more solid than 10NM. GloFo may be the poster child of this strategy as they have decided to skip 10NM altogether. In their case, we think this is a wise decision because you can't make any money in this business by being a fast follower...the leaders are the money makers and GloFo can't wait for hand me downs from Samsung. So far spending on 10NM seems good but some of that spend may be more directed at 7NM capability.
We will watch for evidence of 7NM being a potentially bigger spending node.
Any hope for DRAM?.....
Though the industry would like to see s resurgence in tool demand for DRAM we can find no evidence of a spending rebound. Memory spending continues to be focused primarily on NAND as the demand is much more price elastic and the takeover of the disk drive market continues at full speed with a seemingly bottomless pit demand for more and cheaper flash storage
Retro is in! 200MM and 28NM rule...
Demand for old 8 inch tool sets sounds great as older fabs continue to do well with internet of things which uses trailing edge capacity. This is perhaps one of the best times we have seen in the industry for trailing edge capacity.
We think that aside from service and spare parts that companies will likely see further benefit from older tool sales that have seena resurgence.
28NM is the gift that keeps on giving..... this is the best node in recent times and perhaps the best node of all times as a lot of product continues to be manufactured at 28NM. Having the lowest cost per transistor helps. It is one of the last nodes before the industry got deeply mired in multi patterning.
We hear reports of continued good tool sales. Both 200MM demand and 28NM demand are overall positive for the industry as they help support the overall revenue stream of the company and are a bonus on top of leading edge sales.
Tower Jazz is the retro semi poster child for IOT
One of the companies that is the prime example of retro in the semi market is Tower Jazz. It is quickly becoming (and perhaps already is) the TSMC of trailing edge capacity. Their fabs are filled up with a very steady stream of very solid business and their costs are very low with cheap/almost free fabs and tools. Management has done a great job rolling up this part of the industry while no one was looking.
While IOT doesn't do anything for leading edge capacity it does a great job of sucking up older fab capacity for low cost processors, controllers, sensors and SOCs and all those things that fill up cars, drones, thermostats etc;.
While trailing edge may not be as sexy as leading edge 7NM , the profitability makes it quite attractive....
The Stocks...
The semi equipment stocks have been on a roll and if anything are likely a bit ahead of themselves at this point. Even the management of a number of companies we spoke with were surprised by the strength of the stocks. Many are at all time highs and continue to move higher on light news.
We might consider trimming some of holdings in some of the stocks that have outdone to market as we wonder what will support the strength post Semicon and going into August.
However we would want to make sure we have a good position in the names for the Iphone 7 rollout expected in late September as that will renew interest after a slower August
About Semiconductor Advisors LLC
Semiconductor Advisors provides this subscription based research newsletter, Semiwatch, about the semiconductor and semiconductor equipment industries. We also provide custom research and expert consulting services for both investors and industry participants on a wide range of topics from financial to technology and tactical to strategic projects. Please contact us for these services as well as for a subscription to Semiwatch
[table] style="background-color: #3194D6; margin: 0px 0px 0px 0px; width: auto !important"
|-
| rowspan="1" colspan="1" align="left" valign="top" style="color: #000000; font-family: Calibri,Helvetica,Arial,sans-serif; font-size: 11pt; padding: 8px 15px 8px 16px" | VISIT OUR WEBSITE
| rowspan="1" colspan="1" style="color: #000000" |
|-
[/table]
Last edited: