IBM exits DRAM, PCs still (probably) deep in the red

On the surface IBM's Q2 financials (IBM turns in strong Q2) seem somewhat straighter than Microsoft's nicely massaged Q4s, but a couple of nasty questions occur. Was that not, for example, chairman Lou finessing an exit from the memory market? And exactly how much did the PC operation's numbers improve? IBM made various restructuring moves in the DRAM manufacture area, and as the raw numbers are generally what Wall Street in interested in, Lou Gerstner's words on the subject haven't been widely reported so far. The overall picture is fine, and IBM is taking the necessary corrective actions to maintain shareholder value (Lou's speciality). But Lou's words on the subject were as follows: "The actions we're taking in our microelectronics and storage areas are intended to strengthen our technology segment substantially over the long term. With these actions, we are well down the path of exiting the high-volume manufacturing of DRAM chips, while shifting our resources toward the faster-growth, higher-margin custom chip area. In addition, we are taking various steps to integrate development and manufacturing activities in our hard disk drive business while reducing expenses. These steps will not only improve the competitiveness of our technology segment but will further strengthen IBM's overall business portfolio." A little analysis of the above yields tasty red meat. IBM has been a committed long-term player in memory manufacture, and during the crises of the 80s was, apart from Micron, just about the only significant US survivor. Its joint venture with Toshiba (which it also recently dumped) was intended to maintain its position in DRAM. Gerstner has now made the exit strategy clear, and it's therefore worth recording that an area once seen as strategically vital is now basically toast. Check out the nature of the exit strategy, and one of the associated partners, too. Infineon Technologies (aka Siemens) will be co-operating in a joint venture turning the Essonnes, France plant into a custom chip operation. Siemens (a major European DRAM holdout from the 80s, so a mirror image of IBM) has been muttering the 'custom chip' incantation a lot itself recently. Bottom line: we've got all this DRAM fab capability that's desperately unprofitable, and we've lost the will to keep going. 'Custom chip' is an implausible bolt-hole being put forward by outfits like Siemens and IBM as a valid route they can take to keep their old DRAM fab churning. But custom chips are not the same as DRAM - they're custom, and you need custom customers to go with them. It's a dubious fig-leaf, says The Register. Now take a little look at that storage comment. Lou intends to "integrate development and manufacturing activities in our hard disk drive business while reducing expenses." Rather than quizzing him as to what he means by this, we should just check with our nearest discount supplier. Here in Europe extremely tasty hard drives produced by the IBM plant in Hungary tend to be extremely cheap, to the extent that we've been convinced for some considerable time that IBM cannot be making money out of this business. Lou is obviously trying to slash costs here, QED IBM is probably losing a bundle on the business. And PCs? We'll switch to CFO Douglas Maine. Earlier this year IBM admitted the PC business had lost a sum not unadjacent to $1 billion in 1998. Now Maine is making hopeful noises. Overall operations relating to the PC business were profitable in the first half, he said. Overall operations relating to? What do you reckon that's supposed to mean, if it's not a bit of accounting creativity that incorporates a deal of business that isn't directly related to the business of selling PCs? More clues: Maine says the PC business is on track to deliver "attractive returns" - so it's not profitable, and as an "attractive return" isn't necessarily a profit, it's quite possibly not going to be. And then we've got the optimistic upside that moves attention away from the bit that's bleeding. Maine said that demand for some ThinkPad models exceeded supply during the quarter. Which makes you wonder about demand for PCs in general, or even for ThinkPads (the bit of the business that works) in general. ®