I listened to the CITI conference yesterday, and this is mostly from memory, so may not be all correct.
Anyway, Dario said that SPSN was very comfortable with the current NOR market. Specifically, he said that SPSN and Numonyx now had 35% of the market each while Samsung had 15% and everyone else had the remaining 15%. During the last 4 Q's SPSN has been able to take business from Numonyx while Numonyx has gone to a saner(~6%) per Q reduction in ASP's.
NOR market is expected to continue to contract as NAND influence continues to grow (from $8 B currently to $6.5 B in ~ 2012). SPSN is moving to a more NAND based model that should give SPSN's MirrorBit certain advantages as strataflash scaling problems begin to appear at 32nm. SPSN isn't going directly at the NAND players, but incorporating NAND into products where it has an advantage. ORNand-2 is first true MirrorBit NAND and will be produced by SMIC in 1-4Gb low-densities for high value products.
Big reductions in CAPEX, perhaps to the $200 M (or less) range in 09. Dario mentioned $40 M Capex costs for q4. Currently, ASP's are going down at about 6% to 7% a Q while costs are going down in the 10% to 12% range. SPSN advantages in node migration and in-house production and a 75% reduction in testing costs due to BIST(65nm on board memory controller) are mostly responsible for new competitive advantages. SPSN is still spending about $100M a Q on outside foundries which should continue to go down. $130m inventory build in 07 wont happen in 08.
New products being introduced to ex Numonyx customers and others will help q3 and q4 revenues. SPSN continues to be a major force in the mid-range handset market with products that are more energy effecient than the NAND/DRAM solutions. 75% to 80% of the handset market remains NOR based and that market continues to grow.
First EcoRam solutions should appear in q4, but that's essentially an 09 product. The market for EcoRam could approach $2B/Y by 2012. Not much said about 45nm, but Dario did mention the movement of products from JV1/2 to more advanced processing plants as one of the methods SPSN is using to reduce costs.
Overall, the biggest worry I have is the lack of demand for SPSN's products. Although it appears that SPSN isn't currently being affected by the world economic slow down, things could get worse. Currently, SP1 is at 2k wafers per week, but not all that capacity will be used until the ex-Numonyx wireless customers finish their trials and start selling MirrorBit products(soon). 2k WSPW limits SPSN's GMs to about 25% while a full fab(4k WSPW) would raise GMs into the 30% range. So true profitability, not EBITDA or cash-flow, still remains a distant goal. Still, it sounds like SPSN has quite a few options/things that could improve profitability greatly over the next year or so. We'll see. |
