Taipei, October 29, 2008 --- Nanya and Inotera have also announced production cuts in addition to previous announcements from PSC, Elpida, Hynix, and ProMOS. For Nanya, in addition to delaying expansion plans for both Mei-ya (a joint venture with Micron) and Fab3 Phase 2, its Fab3 Phase1 will also cut production from 30K to 15K from now to Q1 next year as Fab3 Phase1 needs to transfer to Micron’s stacked process. Inotera has cut production by 20% from 130K to 104K. Combined with announcement from other DRAM makers previously, DRAMeXchange estimates total worldwide production cut of 12-13%.
Last week, despite the production cuts, spot price for DDR2 1Gb eTT declined another 3.8% last week (10/21-10/27) from US$1.05 to US$1.01 as market pessimism took over when Samsung announced its revenue for memory has slipped. DDR2 667 1Gb declined from US$1.06 to US$0.99, marking a 6.6% decline. This marks both the DDR2 667 1Gb eTT and original brand maker die price falling closer to history lows of US$1.0 and far below DRAMeXchange's estimated cash cost of US$1.3-1.5.
With the continuing fall in DRAM price, production cut is no longer merely a tool to increase price but a tool to conserve cash in order to survive. The old method of transitioning to more advance process in order to cut cost and raise competitiveness will also not work anymore as average cost (material cost plus package and test) of DDR2 1Gb using 70nm process is roughly US$1.9 if upgrade to 6Xnm. Even if we use 5Xnm, the average cost will be around US$1.4 while current DRAM die price is roughly US$1.0; therefore, there is no way DRAM makers can turn a profit even if 5Xnm process if used. Moreover, 5Xnm process will require more cash invested when most DRAM makers are facing a cash crunch. Besides, 5Xnm will produce up to 30-40% more dies than 6Xnm process and further swamp the market with unsold dies given the weak demand.
DRAMeXchanges' analysts points out that although more advanced process can lower die cost and increase competitiveness against your competitors, continued production cuts is more urgent in order to restore market balance. From a market perspective, if DRAM industry reduces production by as much as 30% in the short run, DRAM makers will need to consider continuing to reduce production by 20% on a long term basis in order for DRAM industry to come back to health.