Bookmark and Share


As 2013 approaches, producers of computer memory are engaging in strategic planning for next year, however, nobody is sure how much to spend on new process technologies as well as manufacturing equipment as demand for dynamic random access memory (DRAM) is getting lower. As a result, it is likely that memory makers will likely reduce output using different methods.

Memory Market Faces Changes

In past years, due to PC upgrade cycles, DRAM suppliers had no choice but to continually advance technology and optimize cost. From the mainstream 60nm process in 2009 to the current 30nm process, every year manufacturers cycled through 1 to 1.5 generations. However, as a result of both a weak global economy and declining PC shipments, and more recently, cannibalization from smartphones and tablet PCs (which has extended the average PC upgrade cycle from 2 to 3 years to 4 to 5 years), DRAM makers have been unable to significantly improve sales. 

As for DRAM supply, technology migration has resulted in greater output, which in turn has created a continual oversupply situation on the market these last few years. DRAM industry value continues to shrink, the majority of manufacturers have suffered heavy losses, and capex figures are decreasing every year. Looking ahead into 2013, TrendForce believes that after the storm gradually settles, only the strongest suppliers will remain, which will help bit output closer to the levels of PC+ era demand.

DRAM Makers to Slowdown Transition to New Technologies, Decelerate Bit Growth

Market research company TrendForce indicates that even industry leader Samsung is highly conservative towards next year’s capital expenditures and bit growth figures. In addition to slowing technology migration plans to the 28nm process, Samsung will only migrate to the 25nm process prior to the advancement to EUV technology. Priority will be placed on profit margins, and advanced process technology will be used primarily for mobile DRAM production.

The 30nm process is currently the mainstream technology for market share dominators Samsung and SK Hynix, while the 20nm process is not expected to see over 50% output until after the second quarter of 2013 due to design difficulties. As Samsung’s 20nm technology is being used mostly for the production of server and mobile memory, commodity DRAM output on the 40nm and 30nm processes will continue. Thus, next year’s bit supply growth forecast is a mere 19%, significantly lower than it has been in recent years. Micron and Elpida are currently migrating to the 30nm process; Elpida has finished client testing, and will gradually ramp up production in the coming months.

Taiwanese manufacturers, who have experienced the greatest losses, are slowing technology migration and decreasing PC DRAM production. Nanya plans to phase out commodity DRAM, turning instead to specialty DRAM production and foundry business. Powerchip’s commodity DRAM shipments have decreased significantly as well; capacity is currently down to 20K wafers per month. If manufacturers all slow output, TrendForce forecasts 2013 yearly bit supply growth will hit a low of 22.2%.

Despite of bankruptcies, mergers and acquisitions, the DRAM industry is still a perfectly competitive market. With acute oversupply, buyers have the upper hand in price negotiations. Both contract and spot prices for 4GB modules have fallen to historical lows, but neither market has picked up. There is a limit to how much lower price can fall, hence capacity cuts must be made to shrink supply. In the long term, manufacturers must all make capacity adjustments and lower capex if the DRAM industry is to return to a healthy state in 2013

Tags: DRAM, Business, DDR3, 2Gb, 4Gb, Samsung, Micron, Hynix, Elpida, Nanya, Winbod, Rexchip


Comments currently: 3
Discussion started: 10/18/12 11:40:18 AM
Latest comment: 10/18/12 03:07:24 PM
Expand all threads | Collapse all threads


Ironic that they want to slow progress, when the only time they make money is when a new form of DDR comes out and can be priced extremly high initially.
1 1 [Posted by: cashkennedy  | Date: 10/18/12 11:40:18 AM]
- collapse thread

ROFL. You shouldn't be worried about their revenue & profits .... they'll payout themselves in years to come.

Maybe you're too young to recall. In 2007, when intel released their enthusiast X48 platform and in early 2008 after P45 was released as first chipsets supporting ddr3, only ddr3 modules that early adopters could buy were in fact overpriced ddr3 chips which were working on same voltages as ddr2 (1.90V and above) which was marketed as Overclocking Friendly Memory Modules. And after Nehalem based Core i7 were released they proved to be unfriendly with its IMC that liked only ddr3 modules which operated with max voltage of 1.65V. And the only who eat most of cream selling those kind of modules were only memory module makers and not the memory chip makers because these kind of ddr3 shipments 1H 2008 were below 1% of total sdram sales.

So the early adopters of ddr4 modules previewed in Q3 2012 probably wont have better luck. And overall shipments of ddr4 even in 2014 maybe will be below 5%-10% of total sdram market.

Most of 8Gb ddr3 dies are more than satisfied with production on 40nm process, and i didnt see to many 8Gb or 16Gb dies that would be offered on either 16GB or 32GB PC modules. At the same time 8GB modules are vastly produced featuring cheapo 4Gb chips, and 4GB moduls with 2Gb chips which are nicely produced on far obsoleted 70nm not 60/50nm that most of them are produced in last two years.

So as long we dont need 32GB modules in our PCs they shouldnt rush anywhere And intel is more than conservative with this, even chips on its top end platform SB-E doesnt support more than 8GB modules. And IB is just shrink of SB and beside memory speed they offer no improvement on controller side.

0 1 [Posted by: OmegaHuman  | Date: 10/18/12 02:10:54 PM]

Lower node size really isn't need at this time. ULV DDR3 is plenty good enough for 90% of consumers. 30Nm RAM is just fine and cheap to make and sell.

New industry specifications like DDR4 are rolled out years before they are needed to allow a cost effective transition when the time comes that they show some advantage. Servers are about the only place where DDR4 will show any advantage and most enterprise is unwilling to pay premo prices for new tech unless there is substantial benefit.

Since DRAM makers continue to over produce, keeping prices low, then they can't really be in all that much need to transition to higher profit new tech. They seem to be making plenty of money on DDR3 even at fire sale prices, though they cry poor-mouth.
0 1 [Posted by: beenthere  | Date: 10/18/12 03:07:24 PM]


Add your Comment

Related news

Latest News

Wednesday, November 5, 2014

10:48 pm | LG’s Unique Ultra-Wide Curved 34” Display Finally Hits the Market. LG 34UC97 Available in the U.S. and the U.K.

Wednesday, October 8, 2014

12:52 pm | Lisa Su Appointed as New CEO of Advanced Micro Devices. Rory Read Steps Down, Lisa Su Becomes New CEO of AMD

Thursday, August 28, 2014

4:22 am | AMD Has No Plans to Reconsider Recommended Prices of Radeon R9 Graphics Cards. AMD Will Not Lower Recommended Prices of Radeon R9 Graphics Solutions

Wednesday, August 27, 2014

1:09 pm | Samsung Begins to Produce 2.13GHz 64GB DDR4 Memory Modules. Samsung Uses TSV DRAMs for 64GB DDR4 RDIMMs

Tuesday, August 26, 2014

10:41 am | AMD Quietly Reveals Third Iteration of GCN Architecture with Tonga GPU. AMD Unleashes Radeon R9 285 Graphics Cards, Tonga GPU, GCN 1.2 Architecture