I can understand your feelings towards SMOD. Todays rally should help reduce some animosity towards SMOD. Here are my thoughts on SMOD:
1. CPQ is currently sole sourced with SMOD for memory. CPQ is looking at another source not becuase SMOD is not performing to expectations, rather because it makes business sense. Else they would have dropped them by now.
2. Today in the memory module market, relationships with key technology enabler OEMs are key (i.e Intel, Rambus etc). SMOD has been recognized by Intel for enabling SDRAM technology. SMOD is one of the few companies that can build, test and manufacture Rambus modules (called RIMMs). So their relationships with key technology enabler OEMs are good.
3. The memory module industry will expreience consolidation. The current key players are SMOD, Kingston, Simple Technology, Viking, VisionTek, PNY & Celestica. Only SMOD is OEM focussed. Why is this good, you ask - well with decreases in module prices, the trend amongst the channel is to buy the original memory for a few bucks more rather than risk buying the third party memory. In the days when 16Mb chips were going for $60, even a 10% delta was significant in absolute dollor amount. Today with 16Mb chips at less than $1.50, a 20% delta between OEM and third party is digestable. So the third party module guys are scrambling. Look at the recent set of events - Simple Technology is selling off its aftermarket memory business unit, PNY shut down their module facility in west cost. Kignston is loosing market share to OEMs. All of this will benefit the OEM, which in turn will benefit OEM focussed company, i.e Smart
4. There has been a shift towards SDRAM in the industry. At an average an SDRAM module has 3 times as many components as older technology. So the company has to expand manufacturing by 3X, just to make the same units. With decreasing revenues, most 3rd party module companies are not investing for the future in test equipment (which can cost upwards of $1M each), and design infrastructure. SMOD is. Look at who the top customers for HP8300 (test equipment for SDRAM modules) for 1997 and 1998 are - SMOD is among one of them. So SMOD's competition will spiral down, to SMOD's advantage.
5. SMOD has a killer engineering group. Curse AJ all you want. The fact is he has put toghter a company with unbelievable technical talent. Want proof - Look at who is first to market new technology - Who came out first with PC100 SDRAM DIMMs - SMOD, Rambus RIMMs - SMOD etc.OEMs take notice of such things and prospective OEM customers will line up with such a supplier.
6. SMOD is expanding. They have design and manufacturing operations in US, Europe and Asia. Which one of their competitors do - NONE except Celestica. Why is this good - well short response time to customer orders is key to the sucess of a large OEM. If the OEMs supplier is not global, then the time to react increases. OEMs tend to choose global suppliers.
7. SMOD is well positioned with DRAM chip suppliers. This helps them 2 ways. The first is that they build memory modules for the chip suppliers (DRAM suppliers such as Samsung, Hitachi, Mitsubishi etc typically contract out their module manufacturing), second SMOD will benefit from their relationship during an allocation market. This is bound to happen soon - else DRAM guys will drop out of the business.
8. If CPQ selects SMOD, and have kept them for nearly 2 yrs., they must see value in SMOD. Else they would have dropped them ages ago.
My recommendation is wait for SMOD to hit $12-$13 and buy like hell. Hold until Intel announces Rambus capable controllers (late 1999). SMOD stock will be at unbelivable heights. |
