HP Memristors Will Reinvent Computer Memory ‘by 2014

[dawp]

according to this wired article:

http://www.wired.com/wiredenterprise/2012/07/hp-memristors/

looks promising for ssd especially

 

[ShintaiDK]

Anual HP press release.

Look at the previous press releases. And you would see memristors should have been in consumer devices by now

So 2014? Needs a miracle.

 

[sm625]

Well we are halfway through 2012 and we still dont have on-die SSD controllers and SSD DIMMs. So even if we were to get Memristor based DRAM in DIMM form in two years, we have no easy way to scale up the memory support. That would set us back a couple years. And windows has no native support for installing and running from nonvolatile RAM.

I am assuming that memristors will replace DRAM and our OS will simply be installed onto this RAM. Right now windows would just stupidly handle this form of memory, it has not even begun to be optimized for non-volatile memory. Which means it will be wasting huge amounts of memory for basically no reason. Remember, there is no reason to load a program or its libraries into RAM if this RAM is the same RAM the programs are already stored on. Everything would essentially be already preloaded on bootup. It will takes years for a company like microsuck to adjust to this new paradigm, and they havent even begun. Nor have intel/amd, since they have no on-die high bandwidth nonvolatile memory controller.

 

[ShintaiDK]

The evil ghost of slowpoke HDs haunt us

 

[Pariah]

The title of the article is misleading. It implies we will see the technology no later than 2014. If you read the article, it says we will see the techology by late 2014 at the earliest.

Article also says HP doesn't appear to have any actual products in mind yet, which implies we won't be seeing anything using this technology in 2014.

 

[bryanW1995]

I think that I started hearing about this around 1994...

 

[Kusnierek]

And we'll all have microwaves in our cars by 2000.

 

[hal2kilo]

Still waiting for magnetic bubble memory.

Seriously, I had invested in NVE which make electronic devices which use the spin of the electrons for sensors and supposedly memory devices. I sold it because it just seemed to be going sideways mostly and not up. They sold patent rights to Toshiba I think to develope spintronic based memory but it seems like it's flowndering.

I also noticed at the time that HP was working of this memrister memory. Seems like some of this stuff just doesn't make it out of developement.

IBM is also working on nanowire memory.
http://www.dailytech.com/IBMs+New+R...ut+Only+has+One+Car+Per+Lane/article23442.htm

Bottomline is that NAND based storage in just an interim solution for storage.

Seriously, it will be obsolete within ten years.