TrendForce：LPDDR2 to Account for 70% Mobile DRAM Shipment during 2013; Maintaining Cost Advantage a Key to Success

According to DRAMeXchange, a research division of TrendForce, mobile DRAM growth is continuing to increase following the releases of smartphones, tablet PCs, and other mobile devices. It is predicted that within the DRAM industry, the proportion of mobile DRAM will increase from 11% during 2010 to 21% in 2012 out of total DRAM supply, and rise to approximately 24% during 2013. LPDDR2 product demand will also experience an uptrend following core chip enhancements and upgrades. With LPDDR2 prices breaking even with those of LPDDR1 during 2Q, shipments for LPDDR2 grew over 60% of total mobile DRAM supply, in effect ensuring its status as a mainstream within the market. In 2013, it is expected that LPDDR2 will account for over 70% of total mobile DRAM shipments. LPDDR3, on the other hand, are anticipated to debut during the same year and may begin appearing in a variety of high-end products. For DRAM manufacturers, the key to success will depend on how well the production costs of mainstream products are assessed and grasped, and how effectively different product mixes are utilized.   

From the demand standpoint, following the continuous SoC updates, the chips used in smartphones next year will mainly be of the dual and quad core variety. Since these chips will need LPDDR2 or higher in order to perform up to its full capacity, for 2013, demand for LPDDR1 is expected to exhibit a gradual decline, and may potentially account for less than 20% of mobile DRAM production. With memory DRAM manufacturers making the transition from LPDDR1 to LPDDR2 production, the dip in the former’s supply amount is likely to increase stocking-related pressure for manufacturers producing LPDDR1 smartphones. While LPDDR3 is set to be designed for high-end mobile phones, tablets, and ultrabooks, the shipments for such products are unlikely to be sufficient to propel overall LPDDR3 demand. As such, LPDDR2 will likely remain the market mainstream in 2013. The earliest year by which LPDDR3 can become mainstream is predicted to be 2H14.   

On the supply side, the four manufacturers currently known to provide LPDDR2 chips are Samsung, SK Hynix, Elpida, and Micron. With regards to its manufacturing-based technology, other than producing LPDDR1 chips, Samsung has already begun to switch to using the 35nm process. SK Hynix, on the other hand, uses the 38nm process, whereas Elpida and Micron both use the 30nm process for the production of LPDDR2. Given that the aforementioned manufacturers have similar technological expertise and production efficiency, and considering how, with uncertain demand, all maintain a conservative attitude towards transitioning to the 2x nm manufacturing process, the 3x nm process is likely to remain mainstream for 1H13. On the subject of mono die density, the 2013 mainstream LPDDR2 mono die provided by the four major manufacturers should be 4Gb. With the utilized capacities being highly similar to one another, how to create a differentiated or unique product combination like the MCP products will become a major issue for each manufacturer to consider.   

Looking forward to the 2013 mobile DRAM market, although the release of smartphones and mobile products is expected to lead to steady growth, with suppliers making a transition from commodity to mobile DRAM production, achieving a supply-demand balance continues to be a difficult task. The fact that market oversupply has been affecting mobile DRAM prices from quarter to quarter, likewise, will also make the process of maintaining cost efficiency and balance challenging. As manufacturers achieve similar technological capabilities, the key to successfully attaining profit and earning a place in the competitive DRAM market will largely depend on how well each DRAM maker controls manufacturing cost, enhances production efficiency, and utilizes product mix.