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, NAND, NVM, enterprise storage systems, etc. -- Ignore unavailable to you. Want to Upgrade?

To: SiliconAlley who wrote (3720)	5/11/2018 5:28:54 PM
From: storage_savant	3 Recommendations Recommended By alanrs Bruno Cipolla Unwelcomeguest Read Replies (1) \| Respond to of 4828

Higher capacity points actually produce less revenue per bit than lower capacity points. One only has to look at DRAMexchange to see this. Today's session averages for 16Gb is 2.895, 32Gb is 2.644, and 64Gb is 3.664. Selling 4 16Gb chips produces over 3 times the revenue of a single 64Gb. So the notion that higher capacity points result in more revenue per unit basis is pure nonsense. Cost per bit is lower on a per unit basis, but the overall revenue is higher if the demand shifts from lower to higher capacity. I'm not sure why the middle capacity point is cheaper on an absolute per unit basis, I'll chalk it up to fluctuations in the demand for a specific capacity point. I agree that Art's original comment was not 100% correct, but you missed the overall point he was making (you seem to do this regularly). Memory (and logic) companies pay a fixed cost per wafer, not per die. The number of dies per wafer is variable, based on the total area of the chip. Because the area occupied by control logic and circuitry of each chip is similar regardless of size, higher capacities end up being more cost efficient on a per bit basis (the wasted space is a smaller percentage of the total area of the die/wafer). Therefore, memory companies can sell a larger capacity at a cheaper cost per bit and still make the same (or even greater) margins. It is reasonable to assume that: 1. Production cost to the company for higher capacities is lower on a per bit basis. Without knowing the exact difference between higher and lower capacity points on a node we can't tell if they are making more or less per unit, but typically higher capacities are better on an revenue and margin basis. 2. Of course costs will drop from node to node, but we can also expect production costs per bit to drop from node to node as well. Costs dropping on an absolute basis doesn't tell us anything without also considering changes to production cost as well.