Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : WDC/Sandisk Corporation -- Ignore unavailable to you. Want to Upgrade?

To: Allegoria who wrote (12465)	7/3/2000 10:40:12 AM
From: Ausdauer	Read Replies (1) \| Respond to of 60323

Eric, You have done your homework and I don't think anyone will contest the information you have organized and presented here. I have a few questions for you. ___________________________________________________________________________ You said... "A variation of the two transistor cell is also used by manufacturers for sequential access approaches, e.g., NAND (SNDK)... and later ...according to SSTI's IR, this allows SSTI to move up the food chain into higher density / data storage flash. Any competitor's sequential access NAND flash is inappropriate for code storage, so in a sense it is a one-way street. SSTI is already beginning to move up, claiming the fastest read/write data transfer rates for their 96 Mbyte CompactFlash Card." First, SanDisk's claim to fame is its NOR flash cells, not NAND. Are you saying that SNDK's products are unacceptable for code storage? Second, are you certain that SSTI manufactures all components of their 96MB CF card? During the last c.c. the SSTI CEO stated that the highest density product they will manufacture is 16Mbit. Has this changed? [Perhaps I misunderstood or am in error on this.] If not, it is clear that SSTI buys components for the CF product. Also, it pays card assembly royalties to SanDisk for the controller technology. I don't see SSTI as a threat to SNDK in the CF business. ___________________________________________________________________________ You stated... "Most industry observers believe that in the near future, 164K is all that is going to be required for flash code storage. Therefore unless Intel or AMD invent a new disruptive manufacturing process, SSTI will own the low-density (not lower end!) flash market. IMO" The argument you make suggests that the requirements of electronics manufacturers will remain static in this ultra-low density range. That would seem to be a bit unusual. I am not saying it is not possible, just that as devices become more sophisticated is it not equally likely that the requirements for these small, insular "bits" of code storage could increase or that a larger density chip would assume a central role that eliminates the need for a single dedicated 164K chip (thus simplifying the assembly process)? ___________________________________________________________________________ Do you believe that SSTI can easily enter moderate and high density flash applications using the Superflash technology they are licensing? I still believe that 16Mbit is a far cry for 256 or 512Mbit. SanDisk, Toshiba, Hitachi and Intel are pushing the envelope on the higher end. I don't see SSTI as capable of competing in this race. Further, SSTI could feel pressure from above as established semiconductor manufactures place pressure on the lower density manufactures. As I said before, if margins are fat and there is a profit to be had, competitors will come. Is the converse true? Will SSTI be able to make a play at moderate or high density applications or is this a "one-way street"? ___________________________________________________________________________ I appreciate your fervor and diligence in presenting the SSTI case. Owning both SNDK and SSTI would seem to be a good fit as they don't overlap. I personally remain doubtful that SSTI can be as well diversified as SNDK can be. Concentrating on low density applications and hoping that the technology doesn't migrate away from you seems a bit risky, but if you are assured that this is the case then SSTI should flourish. I look forward to your responses to my queries. Best, Aus

To: Allegoria who wrote (12465)	7/3/2000 11:11:34 AM
From: wily	Read Replies (1) \| Respond to of 60323

Eric, I found your link to the SSTI white paper very informative regarding Flash construction and operation differences. My question would be, why hasn't SSTI, if they have such a powerful advantage with their design, attacked the card market more aggressively? The article claims that manufacturing efficiency is the number one determiner of cost -- but that's simply not true. Bit density can have an overwhelming effect on cost: I don't know quantitatively what the difference is, but NAND Flash is supposed to have a slightly smaller cell size than NOR Flash. I think that NOR is typically 12L^2, where L = photolithography limit, eg., .18um I'm guessing that for NAND cell size = 10L^2. Next, NAND producers are able to do 2 bits/cell at a 30% cost penalty, which gives them an effective cell size (factoring in the cost penalty) of about 7.5. That is a huge advantage over NOR if NOR can't do multi-level cells. So, that is a major omission from SSTI's analysis in their white paper. wily

To: Allegoria who wrote (12465)	7/3/2000 3:33:32 PM
From: KevRupert	Read Replies (1) \| Respond to of 60323

Eric, Thank you for your time-consuming analysis of SSTI and SNDK! It is due to your comprehensive discussion, that I am quite content with my SSTI/SNDK power play combo. Guaranteed to pay more than any daily double at Belmont in Elmont! Thank again. Your time & efforts are greatly appreciated by many.

To: Allegoria who wrote (12465)	7/3/2000 4:11:49 PM
From: wily	Read Replies (1) \| Respond to of 60323

SSTI/SNDK cost comparison - take 2 If you start by assuming a cell size of 12L^2 for SSTI and 10L^2 for SNDK and then factor in the cost advantages of each -- fewer processing steps for SSTI and multibit for SNDK you can get an approximate cost factor comparison. SSTI needs 14 mask steps vs 20 for SNDK, so give a .7X advantage to SSTI SNDK gets a .5X advantage for 2bits/cell. Plus a 10/12 advantage for smaller cell size (guess). But then you have to add on a .3X cost penalty for the extra cost of doing 2-bit. SSTI = .7 * 12 = 8.4 SNDK = (.5 * 10)*1.3 = 6.5 So, SNDK has a 23% cost advantage over SSTI. But then you have to consider that the SSTI product may have a reliability advantage that makes the product more desirable in some cases. Also, Sandisk says that they put extra bits on each chip for insurance against manufacturing defects, and also to replace bad bits that occur during use. I don't know if SSTI does the same thing, and if they do whether the cost penalty is as great, since they probably don't have to allow for as many defects and failures. Bottom line is margins and the comparison doesn't matter much unless they are going after the same market. SSTI can't touch the highest end part of the Flash card market because they don't have MLC technology (I don't think). Or can they? Hitachi is making a 448MB card without using MLC, by stacking die in the package. So, presumably SSTI could do this too. Still, I would think a 23% cost disadvantage would be a significant hurdle. All approximations and guesses... wily