To: Sam who wrote (55033 ) 5/31/2013 5:53:20 PM From: Bargain Hunter Read Replies (1) | Respond to True enough, but with effectively only two vendors left (Toshiba still assembles HDs, but not very many of them and they mostly just supply themselves), it will still be a profitable business for years to come. Predictions of the demise of tape have been around since about the late 80s, I think, and a lot of tape is still sold, albeit in dwindling amounts. Yes, disk is on something of the same trajectory as tape with a lag of a few decades. Back when (core) memory was expensive a lot of data processing involved reading active data from tape and writing it back. Eventually that role disappeared and its primary role was for archive storage. I'm not claiming that HDDs will lose out to the same degree, but SSDs will slowly erode at least the high-performance portion of the active data access role that HDDs have had.But the larger point for NAND was that the market opportunity for SSDs is so large that it will take quite a few fabs to exhaust it, and that kind of capital investment won't be happening any time soon. Which is a prime reason for believing that a healthy supply/demand balance will continue for an "extended period of time," as Sanjay said. Yes, the opportunity is large, but it is hard for me to get a sense of how quickly it will open up. To the degree that cases like the 34 server example can be found we can expect adoption to come very quickly after completion of pilot programs. Where the trade-off is less dramatic the change will come more slowly. Perhaps the length of the qualification/implementation cycle for SSDs in enterprise servers will prove similar enough to the amount of time required to plan and build fabs that the overall balance can be maintained. If SanDisk is correct about their competitive position the competition would be well-advised not to be aggressive about building new fabs.The question becomes, what does that do to Sandisk's and Micron's peak earnings capabilities, and what kind of PE will Mr. Market give them? I include Micron here because they are a similar albeit far from identical company that specializes in memory and has their own SSD division. Given the history of the NAND industry I think the market is going to be slow to believe that it is different this time. The participants will need to demonstrate discipline and show consistently good earnings. I also have a concern that the whole argument about conservative capital spending by the industry as NAND approaches its end of life could be counter-productive if people take that as "NAND has no long-term future". Who wants to invest in a capital-intensive industry with a 5-year lifetime? This concern will be reduced to the degree that SanDisk starts to demonstrate its capabilities with BiCS and 3D ReRAM while using 1Y and 1Z 2D NAND to compete effectively against other 3D NAND variants.
