1HJan. contract price stays flat as concluded transaction prices show signs of uptrend; increase likely for 2HJan. or Feb.

TAIWAN: According to DRAMeXchange, a research division of TrendForce, 1HJan. contract price stayed flat. Average DDR3 4GB and DDR3 2GB prices held at $16.5 and $9.25, respectively. However, concluded transaction prices are gradually transitioning from low to average figures.

Not only will the capacity cuts from 2H11 take effect in 1H12, but current contract prices are still at historic lows. Thus, some PC OEMs have slowed inventory restocking, and are agreeing to higher contract prices. Furthermore, as the HDD shortage continues to be alleviated in 1H12, PC shipments should return to normal after February. TrendForce believes that as long as makers do not prematurely increase capacity because of the recent spot price stabilization, it is highly likely that DRAM price will increase in February.

From the market perspective, DDR3 2Gb spot price has increased from $0.72 in 1HDec’11 to the current $0.97. While chip price is approaching $1, representing a 35 percent increase, it is an undeniable fact that the spot market is gradually shrinking. This is attributable to the fact that notebooks currently have sufficient memory - the decline of the upgrade market in recent years is also a contributing factor.

The value of the spot market has fallen from encompassing nearly 50 percent of the DRAM market in 2004 to only around 15 percent in 2011. However, spot price still has an effect on the DRAM market and will also benefit contract price increases.

New product and core technology development key to survival in post-commodity DRAM era
The key to DRAM profitability lies almost entirely in capacity expansion and technology migration. From the early stages of 90nm and 60nm-node processes to the current 30nm and 20nm-node technology, shrinking technology has continually reduced production costs. However, it has also increased the number of chips per wafer.

As PC shipment growth has slowed significantly in recent years and content per box no longer shows strong growth momentum, profitability is becoming increasingly difficult to achieve with such weak demand momentum in the DRAM industry.

TrendForce estimates that in 2012, the DRAM market will see major developmental changes. First, DRAM capacity expansion plans will be more rational. As migration to the 20nm-node process is incredibly difficult, requiring a large capex for purchasing EUV and other advanced machinery. This leaves even leader Samsung with no choice but to slow technology migration plans. Thus, TrendForce forecasts the supply bit growth for the year at a mere 22 percent, down significantly from the 50 percent growth of previous years.

Unfortunately, the bit supply decrease does not indicate that the DRAM industry’s oversupply situation will soon see relief – instead, it is a sign that DRAM makers must turn to the development of core technology to hold onto profits. The rise of smartphones and tablet PCs has stimulated demand for mobile DRAM such as LPDDR2, which became the mainstream mobile memory specification in 4Q11, LPDDR3, which will enter mass production in 2H12, and even future specifications like Wide-IO and LPDDR4. DRAM makers with the core technology to develop these products will be the winners in terms of profitability.

Additionally, rising demand from cloud computing has benefitted DRAM makers focused on server DRAM production, and DDR3 4Gb-Mono chips have become another profitability index. Even in terms of commodity DRAM, this year many products will see improvements based on DDR3. For instance, in consideration of the rising popularity of ultrabooks, makers have already developed low-voltage DDR3L in order to meet power saving and extended battery life needs.

DDR3M is also being developed to maximize power efficiency, and some makers are even considering phasing-in GDDR as a possible PC memory solution, as it has higher performance than commodity DRAM in terms of graphics and gaming. DRAM makers have a fierce battle ahead of them, and those that cannot keep up with the pack will have to gradually pull capacity – only makers that possess the necessary core technology will have a chance to stay profitable and survive.