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 : Spansion Inc. -- Ignore unavailable to you. Want to Upgrade?

To: Joe NYC who wrote (3228)	1/29/2008 2:25:09 PM
From: BUGGI-WO	Respond to of 4590

@Joe " I think Buggi's post is pretty much consistent with what I wrote recently: 4bit ORNAND < 2 bit (MLC) NAND < 2 bit (Mirrorbit) NOR < 1 bit NAND < 1 bit NOR " To make it more precise: DIE size - 4bit ORNAND node + 1/2 node = 2 bit (MLC) NAND node - e.g. 65nm NOR = 53-55nm NAND - 2bit ORNAND node - 1/2 node = 2 bit (MLC) NAND node - e.g. 45nm NOR = 53-55nm NAND or: NAND 2b/c needs 37-38nm to equal Quad 45nm SPSN-NOR or NAND 2b/c needs 53-55nm to equal Quad 65nm SPSN-NOR or NAND 2b/c needs 71-73nm to equal Quad 90nm SPSN-NOR BUGGI

To: Joe NYC who wrote (3228)	1/29/2008 2:59:04 PM
From: Pam	Read Replies (1) \| Respond to of 4590

I think Buggi's post is pretty much consistent with what I wrote recently: 4bit ORNAND < 2 bit (MLC) NAND < 2 bit (Mirrorbit) NOR < 1 bit NAND < 1 bit NOR Message 24227254 4bit ORNAND is still not really in full production (with full functionality, at the latest technology), but when it is, that's what the die sizes would look like. Joe, at the same node, for the same density, it is a given that 4 b/c chip will be smaller than 2 b/c chip, which will be smaller than a 1 b/c chip. It gets a bit difficult to compare chips across different nodes and across different technologies because the peripheral circuitry involved in the two technologies may not shrink as much as the cells itself and that may result in unusual die size penalties. If the purpose of this comparison is to determine which has the lowest cost-structure (which is what Buggi is trying to do) it is useless to do such a comparison unless those products are available or will be available within a quarter or two. If one tries to compare 1Gb 65nm MirrorBit Quad with 16Gb 2 b/c MLC NAND at 43nm (both these products are currently sampling), it is not of much value because there is no competition when it comes to cost/bit. However, one should revisit the topic later, either when MirrorBit Quad catches up with NAND on nodes or at densities or both. That would make the comparison more meaningful. Also, as we discussed there is significant difference between the performance of Quadbit and 2 b/c MLC which will play an important role in decision making. But again, it is not a fair comparison because it is obvious that as you have more bits/cell the performance goes down significantly. Yesterday, Sandisk had their CC and they are planning to introduce 12Gb 3 b/c at 56nm in March/April. Some are suspecting that they used these chips in the latest 12GB microSD cards they showed-off at CES 2008 but I suspect they were using 16Gb 43nm chips in those sampled cards. It is being rumored the performance of these 3 b/c chips is comparable to 2 b/c from previous node and if that is the case, they should be fairly decent. The final verdict will have to wait.