4 ms·
4T is a pretty large amount of data though. I think what will happen is that more and more funds will move to SSD research and improvements, pretty much the sa
by tisme 14y ago
4T is a pretty large amount of data though.
I think what will happen is that more and more funds will move to SSD research and improvements, pretty much the same way it happened around the time that the GMR was discovered and first marketed.
There was an apparent slowdown in drive capacity increase and then, when the GMR based drives hit the market we were suddenly swimming in drives of 100's of megabytes.
That tech is getting older and the whole solid state movement seems to have the future. So this may be the last generation of storage media based on spinning magnetic platters. SSD still has a bit of catching up to do, but the writing is on the wall for hard drive makers.
Imagine to have to sign off on a mulit-year, multi-billion plan to create new factories that will have no application in a future iteration of the technology. That would be a very tough business decision. I'm happy they got as much production capacity online as they do.
On the plus side of all this: the last generation magnetic media, at 4T or even beyond will be sold at rockbottom prices to keep those factories alive as long as possible.
- Klinky 14y agoSeagate has demoed HAMR a tech which might allow them to get to 60TB and may start showing up in 6TB consumer drives within a few years. Other research is being done on adding sodium chloride to platters and IBM is researching magnetic storage at the atomic level. So long as magnetic storage can stay ahead of semiconductor fabrication then it will have it's place. As far as someone making a decision to build a factory for commodity hardware manufacturing, that same issue occurs with NAND which is quickly seeing profit margins plummet. E.g. a 50nm NAND plant 4 years ago is no longer competitive today. This is probably why Intel jettisoned their NAND manufacturing.