Re: SANDISK--Apollo asked me to comment on SanDisk's IP and related subjects. I made some recent comments on the SNDK thread. To summarize, it appears to me that, despite some key patents held by SNDK, flash memory is rapidly becoming a commodity in the field of non-volatile removable or embedded storage. About four years ago, SNDK clearly had proprietary technology in removable storage devices, based on its "'987" patent and others. Since that time, a number of other companies have become involved in flash memory to the point where there is no one single type of flash memory even in similar applications.
Among the key developers of flash memory was INTEL, which designed the "double density" chips that allowed twice as much storage capacity. In return for this technology, SanDisk cross licensed its own technology to Intel, making Intel a very substantial player. Intel, however, appears to be concentrating on non-removable memory, such as the 128 mb chip in its MP3 player. SanDisk is the leader in removable memory, but again, it has various agreements with other companies, such as Toshiba and Hitachi, providing SanDisk with additional manufacturing capacity, but providing those companies in return with certain flash technology.
Certain companies have attempted to produce or cell flash memory that may make use of technology patented by SanDisk. Lexar tried to do this but got sued and lost. As part of the settlement, Lexar must now pay royalties to SanDisk for anything it makes or sells that uses the '987 patent. A number of smaller companies may be using SNDK proprietary technology, but lawsuits may not be the best way to deal with these problems. One other aspect in the current situation--a temporary drop in demand--will probably result in smaller flash card makers dropping out of the market altogether, thereby broadening the market for SanDisk.
I think that the decision by SNDK to join with Toshiba in building and operating the FlashVision facility in Manassas, Virginia is the key to evaluating SNDK. For the first time (later this year), SNDK will own a part of its own fabrication plant. Given Toshiba's up-to-date chip technology (.13 and.16 micron) and also Toshiba's record of efficient plant operation, this plant will probably become the low cost producer of high capacity chips (256 mb and higher). Even though SNDK is not the only player in this field, it should succeed as the low cost producer, in addition to which it has a constant new stream of patents that make the chips work better. An example is a recent patent that makes sure each sector gets erased and reused with approximately the same frequency as other sectors. This extends the useful life of the chip.
The bottom line on removable flash cards is that, while turning into a commodity type business, SNDK still has leadership in terms of performance, reliability, and low cost, high volume production.
As for embedded flash memory, the field largely belongs to Intel, AMD, SSTI, and possibly Tower Semi. AMD is particularly strong, but its chips are generally low capacity (under 4 mb) for cell phones, set top TV boxes, and other appliances. Intel recently made an agreement with Tower to build a new plant in Israel. Tower has an agreement on embedded technology with Saifun, another Israeli company, and it is likely that this plant, currently under construction, will give SNDK a reasonably strong foothold in this part of the flash memory market.
In trying to determine the future prospects for SNDK, one cannot ignore its magnificent financial record. Cash balances have been maintained at very high levels, and the company is virtually debt free. This means that there is little likelihood of diluting earnings of the existing shares. It also means that in periods of slack demand, such as exists now, SNDK is better able to weather the storm than other smaller players with less cash on hand.
There is no doubt in my mind that the demand for flash memory continues to grow rapidly. Besides the obvious consumer oriented appliances, such as cameras and MP3 players, there are indications that removable flash cards could become a key component in health care, permitting a patient to carry a small, almost indestructible card containing pertinent health information, and able to be accessed only with the consent of the patient. The current "P-Tag" project that SanDisk is working on with a Dallas based firm could turn into a very large application that would be pretty much immune to consumer ups and downs.
On balance, it still appears to be the case that SanDisk dominates at least some parts of the flash memory market. That, combined with the robust increase in demand that is estimated for the next three or four years at least, should keep a company with good financial fundamentals on the right track. As far as the price of the stock goes, the prospects for profits are better now than when the stock was selling at more than twice its present price. The price fell mainly because SanDisk is considered just another semiconductor stock by many analysts, and the price fluctuates pretty much with the SOXX or the ML Semiconductor Holding Trust, trading on AMEX. Even though one can expect SNDK to fluctuate with the other semiconductors, the one thing that is clear is that flash memory is a market very different from, say, microprocessors or DRAM.
Art Bechhoefer |
