Since last week, there has been an escalating story in the industry press around the availability of NAND flash, raising questions and concerns that there may be a shortage of NAND flash. Naturally, we at Smith have received questions and inquiries regarding the NAND flash situation. While we typically answer these questions directly to our customers, the mounting story pressure is one that warranted comment from our perspective.
The root of the NAND shortage in the industry press relates to earlier production cut backs by various NAND manufacturers. Among the more commented cut backs was Toshiba's 30% reduction in NAND production in late July, "due to market oversupply and price declines," as cited by EETimes.
There has also been speculation that there could be a NAND shortage as a result of Apple's widening of their suppliers for the anticipated announcements on Wednesday, Sept. 12. While a logical argument, we continue to see ample NAND and NAND related components available in the market. This is particularly true since SSD and PC end-products have not seen the demand that was anticipated for 3Q12, as was highlighted by Intel's announcement today on lowered sales (see this by Financial Times or this from Manufacturing Business Technology).
Previously, it was thought brisk SSD sales in 2012 would help grow NAND usage. That, however, has not come to fruition. And though NAND usage in the consumer electronics market helped offset that lesser SSD demand, it was not enough to grow the NAND needs to meet the forecasts. Coupled with production capacity increases from NAND manufacturing die shrinks, production easily outstripped overall NAND needs and contributed to excess inventory and lower ASPs.
This reduction in production by NAND manufacturers has provided some price stabilization – and, in fact, some price increases – for mid- to high-density chips. But the availability of these densities is relatively stable for most end users, and lead times have simply gone from immediate to standard production lead times. Where we do see tightness – not shortages, tightness – is in industrial applications using SLC NAND and in the secondary market of channel consumers, where the loss of production has also caused the loss of wafer availability for the third party flash product makers (SD Cards, Compact flash, thumb drives). Here, the prices have increased because NAND manufacturers prefer to support the larger, tiered, consumer OEMs with whom they have inter-related programs and builds before sending wafer capacity elsewhere.
What we also see is that the cutbacks have made it more difficult for some NAND purchasers to negotiate pricing for reduced ASPs. They are not yet willing to pay current market prices for product that was cheaper earlier this year. Maybe, rather than saying that there is a shortage of NAND, a better statement might be that there is a shortage of low priced NAND in the market right now.