The transition to 3D NAND technology has slowed to snail’s pace due to unexpected steps in the manufacturing process. This has had financial implications, with vendors struggling to produce the capital needed to meet increasing demand. Major tech companies like Intel and Toshiba have been feeding planar NAND to users in an attempt to catch up, but they will have to begin relying on 3D NAND by the end of the year when 2D offerings become virtually obsolete.
DRAM demand has been projected to grow to 20%, but supply is at a mere 15%. The NAND market currently represents 30% of the memory market, with all players racing towards the third quarter of 2017, when 3D technology will come to represent over half of the industry’s output.
Micron projects that the situation will only worsen. It expects demand to grow by 45% over the next four years, much of which will be fuelled by robust purchasing of solid state drives and mobile devices.
The transition has been slowed by several problems. Most players have misconstrued manufacturing costs. 3D NAND was expected to be cheaper than its 2D ancestor, but fabrication has turned out to be three to five times more expensive.
Semiconductor technology has managed such glittering achievements over the last few decades by following Moore’s Law, which no longer applies to NAND. Vendors have generally chosen 3D NAND for its price per bit. On the surface, this makes sense. Dig underneath, though, and you find a different reality. Vendors still win under these conditions, however, as prices soar to reflect scarce supply.
To read more insights and view diagrams, read the article on Semiconductor Engineering: NAND Market Hits Speed Bumps