Replacement for HDD and SSD looks promising

from GalCiv2 Forums

IBM is working on this new tech to replace HDD and SDD tech. Looks promising:

http://www.engadget.com/2014/09/04/ibm-wants-to-kill-the-hard-drive/?ncid=edpicks&google_editors_picks=true

76,617 views 9 replies
Reply #1 from WinCustomize Forums Top

Certainly sounds good....;)

Reply #2 from Sins Forums Top

Basically, this is just a new form of an SSD with dramatically increased information density and without the drawbacks of flash.

  • Infinitely rewritable
  • More data dense than flash memory by two orders of magnitude
  • Read/write 1E6 times faster than HDD memory
  • Cheaper than existing HDDs

This sounds too good to be true, but hey...  If they make it work, this'll be awesome.

Galactic Civilizations IV
Reply #3 from JoeUser Forums Top

If they can pull it off, what's not to like?

Reply #4 from GalCiv2 Forums Top

Of course, to take advantage of the speed, the downside is that you'll need a disk controller and data bandwidth capable of transferring that quickly. ;)

 

Reply #5 from Stardock Forums Top

They achieved a storage capacity of less than 8 bytes. Sadly sounds like another memory tech that will be 'five years away' for decades.

Reply #6 from Sins Forums Top

Killjoy!

 

The controller and bandwidth isn't that big a deal.  They'd need to redesign the architecture to take advantage of so much speed, but when you look at bandwidth in other applications, it's already staggeringly high compared to drive utilization.  The R9 290X has 320GB/s, several hundred times the bandwidth of current SSD's.  Terrible bandwidth for drives is a product of terrible drives more than anything.  There's no point in making a SATA controller cost you more money just so you can not utilize more of it's bandwidth.

Reply #7 from WinCustomize Forums Top

Yes, what psychoak said....;)

Reply #8 from Sins Forums Top

Quoting kryo, reply 5

They achieved a storage capacity of less than 8 bytes. Sadly sounds like another memory tech that will be 'five years away' for decades.
End of kryo's quote
Normally I'd fully agree there, but in the video, he discussed pricing which I thought was odd if it wasn't moving forward at a swift pace.

Reply #9 from Stardock Forums Top

Manufacturing cost was a SWAG on his part, and he didn't say anything about the time needed to produce commercial products using the technology. So I'll file this one in the same category as MRAM (which we've been waiting 25 years for) until proven otherwise. :)

Edit: If you look into this a bit further you'll see it too has been in the works for 10 years already.