With almost no growth in DRAM wafer capacity, and slowed technology migration, we believe DRAM supply bit growth will hover around 20% in 2013 and move up to 28% in 2014.
DRAM demand from traditional PCs to decline in 2013-2015, due to PC unit decline and low DRAM content growth (5% per year). However, we believe the fast growing tablets and ultramobiles (PCtablet hybrids) will more than make up for the PC DRAM demand decline. High-end tablets and ultramobiles use mobile DRAM, but we expect the mid- to low-end segment to use PC DRAM instead on cost concerns.
We expect smartphones to consume more DRAM than PCs in 2014. Not only is their unit growth fast (40% in 2013E and 26% in 2014E), but their DRAM content growth is also fast (27% & 35%) thanks to hardware spec upgrade competition amongst smartphone makers. As we saw during the rapid growth phase of PCs, we expect the smartphone industry to compete fiercely on hardware upgrades. We expect displays to be a key area of hardware upgrades, as they are such an important part of the user experience.
CAPEX in DRAM
We believe DRAM makers will refrain from making aggressive investments in new capacity and technology as they approach the limit of current lithography tools (ArF immersion scanners) at around 25nm. Moving to 20-22nm and beyond is physically possible with current lithography tool, but may not be economically practical due to upwards spike in process costs. As such, we believe DRAM makers will be hesitant to invest in a major way until next generation lithography tools (EUV scanners) are commercially ready.
The key issue delaying EUV scanners has been light source. High-volume commercial production EUV systems need to operate at >200W (and run 24hours/day). It took five years to increase power from 2W in 2009 to 55W in 2013, at which fab productivity is 43 wafers per hour (wph). Hopefully, they find light at the end of the spectrum to help DRAM Industry invest and keep pace with the surging demand of SmartPhones and Tablets.