Macro picture: NOR flash is projected to grow slowly as it's a mature market. You reacted to this comment so it's a bit weird for me to see your current claim that I focused only on the micro picture. More about the macro picture below the next paragraph.
NOR market (revs) is going to grow slowly not just because its a mature business but also because a significant part of its business is under attack from MCPs which mimic NORs functionality, without using NOR, at a much lower cost. There is significant unit growth as more and more CE devices are showing up which can and do use NOR, but the industry revs are stagnant because ASPs are under pressure and some of the business is being lost to other newer solutions.
Micro picture: Spansion needs relative stable ASP's to make BE in H2, and there's some reason to believe that's possible with Numonyx loosing the incentive to initiate a new price war, and Samsung as ruthless as they are not positioned that well for the Nokia's of the world; they are better positioned to take share from Numonyx than from Spansion.
Having stable ASPs is beyond any company's control, in a highly competitive landscape. If Numonyx doesn't start a price war someone else will. The bottom line is capacity is out there and part of the unit growth is eroding fast and that's the reason for no pricing power!
In this day and age companies have a supplier/Customer relationships while fiercely competing in their end products. Samsung is a supplier of OneNAND to Nokia in one of their high end model despite the fact that they compete fiercely in their cell phone business. There may be a slight bias if the prices offered by a competitor/supplier are the same but when there are price differences, competitor does get that business and that's the right approach anyway.
To your comment: "do you guys know the price differences between NAND and NOR?", here's hoping it isn't meant to be patronizing. Content wise, the rather big price differences are caused so far by cost differences, and to a lesser extent by different performance characteristics that defined the original market spaces. The picture of cost differences is changing because of the introduction in the NOR space of Quadbit, 300mm, faster than usual introduction of next node (45nm). In the NAND space the picture changes because floating gate doesn't scale nearly as easy anymore as it did. (Macro comment).
I keep saying that you should not compare NAND with NOR. In the code execution market, NAND with some SRAM can mimic NOR at a much lower cost but NOR and this is the part of "traditionally" NOR business which is under attack. Even with 45nm/300mm o/p SPSN cannot compete with newer, lower cost solutions. SPSN is already seeing an erosion of revs from this business and it will get more apparent in the coming quarters and years.
FG will have problems down the road but that is not the only avenue left for NAND. They are already working on many alternatives. MirrorBit Quad, as much as you guys think can lower the cost/bit is not even listed at Spansion's web site. I tried to lookup the specs and couldn't even find it! I am curious as to what kind of Write performance does it have? I am pretty sure it is a lot slower than what NAND has been delivering.
Now back to your comment that some here are "silly" to überhaupt compare NOR with NAND. My reply was they're already competing in high end cell phone market so it's rather normal to make the comparison, to which you then replied "you ain't seen nothing yet". Seeing how you now emphasize that NAND will compete with NOR even more, what does your original comment mean to you? ;-) ;-)
Cell phones were 100% NOR market when it came to Boot capabilities and this is what is changing and it is going to irreversible because more and more of these devices that need to be booted are also looking for significant amounts of data storage. So yes, SPSN is competing but their revs are declining in this segment of their market and it is just going to get worse once Toshiba also becomes a Supplier of OneNAND. As for data storage market, NAND obviously will rule for the foreseeable future. MirrorBit and Quadbit have no chance in high density applications even if it is manufactured on a leading edge node on 300mm wafers.
Basically I agree with you that Samsung will be a fierce competitor. I always have. There's a very good sized buffer though between Samsung and Spansion: Numonyx. Numonyx can't cause a price war, and because of many reasons discussed here it will be eaten by both Spansion and Samsung. Hence to survive and get money to equip their empty Catania fab they will need to focus on cost custting, and focus on the more profitable products. This will mean a planned reduction in outsourcing to Intel and STM (Numonyx only got one functioning 200mm fab, that's all) which means a planned decrease in market share. I think there's reason to believe that everyone around the table already 'agreed' to their respective mid term positions because there's little alternative. I think they only still can disagree on long term positions. Here's where the macro picture comes into play again: There's probably good sized places for Samsung/Tosh and Spansion for various reasons, but not for Numonyx as they're too late in about everything and phase change mem is in a down right ugly shape cost wise for a very long period at least. The NOR market isn't likely to disappear as it is projected to grow slowly for years. I add to that that I think that the NOR market is going to be serviced by newer technologies and that Spansion already mastered the by far most advanced and best accepted and evolving interim technology, except for high end where Samsung/Tosh have the more cost effective and evolving NAND based solutions. So long term (e.g. four years from now) Samsung/Tosh will compete way more directly with Spansion. Samsung will need to find a new technology to replace floating gate + they'll remain a competitor for the Nokia's of this world. I think they'll find a coexistance modus then, each with different emphasis, each with a large base in cell phones too. For next two or more years however the buffer is big and ready to get smaller.
Where you and me disagree is that you give it a higher chance that Samsung/Tosh will spark a price war, and that you see a bit more opportunity for Samsung to compete directly with Spansion towards mid range cell phones than what I'm seeing. This results in your more dire picture for Spansion: You see Spansion loose (not a good chance to make BE in H2, not a good chance to survive long term). While I see more chance of Spansion growing nicely at the cost of Numonyx for years, make BE in H2 as a result, and have a good chance long term consequently.
Well, I have never said SPSN has no chance of surviving in the longer term. Spansion is trying to evolve and diversify away from the traditional NOR business and that is a good thing for them because NOR market is stagnant and may even erode. SPSNs future is not just tied to being BE or not in 2H08. SPSN stock has come down a lot and it offers a good Value but this Value also comes with significant risks! Market is clearly recognizing this and that is the reason the stock is where it is today. It say's it all, whether you like it or not. Markets are not always efficient but they see significant risks right now and when they see a clearer competitive landscape emerge in the NOR markets, they will reevaluate it's stock price. |
