Hit by poor market conditions in 2008, A-Data Technology saw its first annual loss since it was founded in 2001. However, the memory module house made a turnaround in 2009, and saw profits for the year reach an all-time high.
A-Data is relatively young compared to rivals, such as Kingston Technology, which was established in 1987. However, it rapidly advanced to become one of the world's major DRAM module suppliers. According to report, A-Data broke into the top-three of DRAM module suppliers for the first time in 2004 and four years later, it moved to second place, lagging only long-time market leader Kingston.
"I do believe in the long run, there'll be only two major players in this dedicated DRAM module field," said A-Data chairman Simon Chen said. "More small- to medium-size peers will turn their focuses away from DRAM modules."
Chen told more about how A-Data went through the DRAM industry downturn between 2007 and 2008, and about its future outlook.

Q: A-Data swung back to profit in 2009, and saw revenues return to a growth track. How did your business recover from the previous bad year?
A: A-Data was established in May 2001, and generated profits equivalent to its paid-in capital in 2002. It enjoyed profit growth from 2003 through 2006, when annual profits expanded to NT$1.6 billion.
In 2007 when the DRAM industry entered its unprofitable years, A-Data made a small profit. Conditions appeared to be worse than estimated in 2008, when the company suffered its first annual loss.
A-Data managed to make a turnaround in 2009. In addition to improved macroeconomic conditions, we also explored new ways to grow the business.
A-Data supports its chip partners when necessary. But we may have given too much support during the previous industry downturn, which became a huge burden on our profitability. We have made adjustments to the company's DRAM purchasing strategy to avoid a repeat if the next industry downturn hits.
A-Data also began to work on a system to rebuild its corporate identity in the second half of 2007, in a move to further enhance the company's presence and brand awareness in the international market place. The efforts are bearing fruits. We have expanded our brand visibility, which is important for our channel sales in emerging markets. Distributors are more willing to promote our products, as the brand-new corporate logo is very attractive.
In addition, A-Data stepped up hiring of talents with many years of experience in the memory industry between 2008 and 2009, including Dr Il Ung Kim, who formerly worked for Samsung Electronics' chip division. We find it necessary to build up a pool of talents for the company's future.

Q: You mentioned that A-Data is expanding its presence in emerging markets through its sales channel. Could you give a few more details?
A: We have put more resources into localization services since 2008. Though A-Data is a latecomer to this area, we have worked our way to becoming one of the world's leading module houses, and flash-based memory card and USB drive vendors.
Since 2008, A-Data has rapidly increased the number of its sales spots worldwide. In 2009, we opened new branch offices in Mexico, Brazil, South Korea and Russia. This year, we plan to add new offices in India, the UK, Chengdu (China) and Japan. At the end of 2010, A-Data will have a total of 17 sales offices worldwide, and we look to add four more next year.

Q: A-Data was quoted in a recent article saying you would generate record revenues and profits this year, as the markets for smartphones, touchscreen PCs and all-in-one PCs are set to boost demand for your DRAM and NAND flash products. From the supply side, how would you achieve the goal?
A: First I'd like to point out the new or maybe killer applications that we think will boost demand for memory products over the next couple of years.
I see many opportunities related to cloud computing. Some may think that the market for standalone PCs will shrink as hardware is not to be considered as cloud computing applications, and therefore DRAM modules will miss out on the opportunity. But in fact, the concept of cloud computing - in which applications and data are hosted on centrally-located servers - requires newer power-efficient systems, which will be the next dominant consumers of DRAM modules.
Notebooks, all-in-one PCs and servers are considered as hardware for cloud computing services, and modules used in the devices carry a higher margin compared to the ones in standard PCs. In other words, the DRAM module business is moving toward more profitable areas.
As for NAND flash, many market watchers and research firms have identified many applications that will boost demand. The market for higher-density memory cards continues to be mainly buoyed by smartphones. Memory cards are also seeing growing adoption in GPS devices.
The market for solid-state drives (SSD) is no doubt a driving force for NAND flash memory, and I believe it will show robust growth from 2013 onward. NAND demand in tablets, such as Apple's iPad and e-book readers, also has large growth potential.
As I mentioned, A-Data is expanding its worldwide presence, through our closer cooperation with local channel distributors and sales teams stationed at our selected regional offices. We are committed to our localization strategy especially in emerging markets.
Meanwhile, A-Data is constantly developing its IP and patents for NAND flash products, to pave the way for the take-off of new markets.
In addition, capacity at our production sites in Taiwan and China will expand this year. We are upgrading existing SMT lines to high-speed ones at plants both in Taiwan and Suzhou, China, and will also install new SMT lines at the Suzhou facility. Currently, A-Data has a total of 30 SMT lines - 20 in China and 10 in Taiwan.

Q: Do you have any preferred DRAM suppliers? How about NAND flash?
A: A-Data now buys DRAM from almost all of the leading chip vendors including Samsung, Hynix, Elpida and Micron, and Taiwan's Powerchip, Nanya, ProMOS and Winbond. We have sourced less chips from Winbond since the latter half of 2009, as it now focuses more on NOR flash and niche DRAM businesses.
For NAND flash, we work with Samsung, Hynix, Micron and intel. Toshiba contributes a minor share to our procurement.

Q: How much do DDR3 modules contribute to your DRAM sales?
A: A-Data has seen the sales ratio of its DDR3 modules grow fast since March this year, and now half of our DRAM modules are DDR3 products. We also have a certain group of customers demanding DDR products.

Q: Previous reports cited you as saying prices for DDR2 chips may exceed DDR3 parts during the second half of 2010, as demand may not go along with the supply side's reduction pace. Are you still keeping this estimate?
A: There will be a short-term rise in DDR2 prices before the memory type is categorized as niche DRAM. We see more chip vendors aggressively migrate their production to DDR3, and the supply of DDR2 will shrink further in the second half of this year. However, DDR2 chips will still be demanded in the channel through the end of the year.
The same happened when DDR was about to move to the niche market. Prices for DDR chips at a time surpassed DDR2 quotes.

Q: Can you explain A-Data's strategy with regard to products other than DRAM modules? Are you aiming for product diversification?
A: DRAM modules and NAND flash products now account for more than 90% of our total revenues, and I think the proportion will remain almost unchanged for the next three years. Of our core memory products, the sales ratio between DRAM and NAND flash products has been maintained at 50:50.
In 2010, we may see sales of our DRAM module business outperform the NAND flash segment due to higher ASPs of DRAM memory.
A-Data is also developing non-memory products, including the new power supply product line that will be unveiled during the upcoming Computex. Sales of our non-memory products are growing.

Q: Taiwan's DRAM makers are speeding up efforts to shift to sub-50nm processes. What are your views on their long-term competitiveness?
A: Taiwan's DRAM chipmakers suffered huge losses from 2007, 2008 through the former part of last year. In 2010 and 2011 they are expected to earn sufficiently to offset the cumulative losses, meaning that reducing the debt ratio is the top priority.
At the same time, the once financially-troubled players need considerable capex to procure immersion lithography equipment, which costs over NT$1 billion (US$30.19 million) per set, in order to shift to more advanced processes. But time is limited. If we say demand will be very brisk this year and 2011, they only have less than two years' time to advance process technology while also improving their financial structure.

Q: In early 2009, you urged the Taiwan government to play an active role supporting the island's DRAM industry. But later that year, the government's mission to restructure Taiwan's capacity-driven DRAM industry apparently failed. What are your views on such development?
A: I have to say that it's a pity the Taiwan government missed the crucial time for restructuring the local DRAM industry. Though the local players have moved to cooperate with their technology partners overseas, they are still lagging far behind their larger Korean rivals.
Samsung and Hynix have a combined market share of almost 60%. Although two alliances have been formed to compete against the two leaders, I still feel it's tough for them to challenge their dominance. Obviously the competition in the DRAM chip market is now between the Korean and non-Korean camps.
The same problem with Taiwan's panel makers. Their Korean rivals are expanding at a fast pace. I do think our government should act now, immediately carrying out plans to support Taiwan's two largest panel makers. The Taiwan government can consider encouraging the local panel suppliers to form alliances with peers in China or Japan, as they have the same goal of reducing the threat posed by Korean rivals.

Q: What is your outlook for the DRAM module market? Do you expect the market to grow? Will it be attractive for newcomers? Or will you expect some companies to quit the market?
A: I believe more small- to medium-size players will leave the DRAM module business, or turn to niche markets. A-Data will be one of the two major third-party DRAM module houses, and remain profitable in the market.
I expect the top-2 module houses' combined share will exceed 60% in 2013, compared to less than 50% in 2009.