I know, we talked about that last week, here for example:
Message 28925095
They will be investing about $2.3b. Here is an excerpt from Lam's CC:
Martin B. Anstice - Chief Executive Officer, President and Director
No, I think having a conversation around NAND is just fine. I mean, I -- there's a decent component for the industry of conversions, right? I mean there's about 0.5 million wafer starts to conversion likely in NAND flash this year and that's a lot of spending, that's approximately $3 billion of spending.
Message 28905940
That number is for conversions, but it is more expensive to build capacity from scratch. That $2.3b won't buy a lot of capacity, and, IMHO, it will be needed for SSDs. There hasn't been an announcement from Samsung, as far as I know, about whether this fab will be used exclusively for NAND or some combination of NAND and DRAM. There are spot shortages in DRAM now as well as NAND. Listen to the Bedard talk I linked earlier, for views on 3D and on shortages. Or read some of the DRAMeXchange pieces I have posted over the past few weeks, here and on Don's thread.
And here is what Mike Splinter said in AMAT's CC:
Satya Kumar – Credit Suisse Securities ( USA) LLC
And Mike, I guess like on NAND, are you seeing much wafer capacity additions in the back half? Do you have an estimate for how much wafer capacity is, that you think is being in NAND Flash this year?
Michael R. Splinter
I don’t have a precise estimate, I think there’s still capacity expansion within existing fabs that’s everybody’s first move because 3D NAND isn’t ready yet and I don’t think we’ll see big expansions until we get to 3D NAND. And then as I said, we’ll see a build investment cycle in NAND, but I think most of that’s next year.
Message 28907390
Samsung and Hynix are behind the cost curve with 2D planar. My guess is that they are trying to move to 3D rather than go down to 1x nm because they think that they can catch up. Neither Micron nor Sandisk believe that the 3D that will be available in 2014 or, likely, even 2015 will be cost effective relative to their 2D production. That is the question that will be answered over the next 12-18 months.
And note--if Samsung does outfit the new fab with 3D equipment (which I am guessing that they will, at least in part), they won't, if it doesn't prove to be cost effective, just switch back to 2D easily. From AMAT's CC:
Ben Pang – B. Riley & Company
Okay. And then in terms of the 3D NAND, you mentioned some early shipments, second half of this year, is it easy for the customers to deploy that for planer if they want to or are there significantly different configurations for the tools that they’ve locked in, if they make a decision on 3D NAND right away?
Michael R. Splinter
Yeah, the customers that are making those early transitions are really focused on the 3D NAND technology, and they are very different. Some of these steps are very different than what you see in planer NAND. |
