Should we bail out the DRAM makers?
Struggling Qimonda is getting a financial booster shoot, with announcements over the weekend that it will take loans from the German free state of Saxony, a financial institution in Portugal, and from parent company Infineon totaling about $450 million to ramp its Buried Wordline technology.
As described by Qimonda, Buried Wordline technology allows the wordline of a chip design to be built into the silicon substrate, while the bitline is built on top of the substrate, which produces a leaner technology.
With the loan announcements, Qimonda’s execs are no doubt breathing deep sighs of relief. The announcements end weeks of discussions that could have led to less-favorable packages for the Germany-based DRAM maker, including ones that UK-based EDN sister publication Electronics Weekly said "potentially had Qimonda’s long-term independence hanging in the balance." (See "Qimonda saved from the brink with government bailout.")
I’m personally not a big fan of all of these government-sponsored bailouts. It’s my opinion that if you can’t properly manage your business, that business should go by the wayside, be that automotive, memory, or whatever. Yes, this is unfair to many of the employees who become layoff victims of the bad management, and, yes, this could hamper innovation because some of the tech industry’s best technologies have come from companies with financial issues, but when it comes to an industry like DRAM — one that has had multiple opportunities to correct itself — we’re just giving a man a fish, not teaching him to fish.
Clearly Qimonda’s backers disagree with my thinking and perhaps see some highly innovative approaches in the company’s Buried Wordline technology, or at least enough to plunk down a sizable chunk of change in this credit-crunched global economy of ours. In doing so, they are saving about 3,000 Qimonda jobs in Germany, according to reports. But is baling out a company in this manner the right thing to do in these economic times, especially in the memory industry?
IC Insights recently reported that there is a price surge coming for DRAM. This will come about because of decreased capital spending by memory makers in reaction to the currently slowing demand. As a well-know fact, overproduction has brought about an oversupply of DRAM, which has driven the memory’s ASP (average selling price) down, which, in turn, has driven revenue at DRAM suppliers down.
The report, while remaining professional, rightfully had an air of skepticism to it with regard to how long this pricing "recovery" would last. Glance at the last five years or so of the DRAM market’s pricing history and you’ll see increased volatility.
Some analysts have gone so far as to say that if the DRAM industry had better leveled itself off, the total market revenue declines expected for 2008 and 2009 would be to much lesser extents. And to be sure, Qimonda isn’t alone in its struggles. Elpida, Hynix, Powerchip, and ProMOS aren’t exactly flying high nowadays, either.
A collapse within the DRAM industry, one that drives some smaller players out and consolidates the segment some, could be just what it needs. Mind you, I’m not encouraging a monopoly here. Far from it. Recognize that if DRAM ASPs grow too high, the ripple effect could damage many other semiconductor industries, like PCs and consumer electronics.
However, trimming supply and competition would lead to a more stable pricing structure and a stronger market segment longer-term. For these reasons, I supported Micron when it in October snatched up Qimonda’s Inotera Memories stake for $400 million (leaving Qimonda with an almost equal investment loss) and watched carefully as Samsung made a move fro SanDisk.
What do you think? Share your thoughts on the DRAM industry below. Should its suppliers be bailed out or is a bailout the worst thing for the industry right now?
Tom commented:
Mike commented:
Robert commented:
Husky commented:
Israel Beinglass commented:
huh commented:
Suzanne Deffree, EDN commented:
Old Man in the PC commented:
ChrisPE commented:















