The future of Storage

[seanp789]

I've seen several news posts recently which Ill post here. Now you dont need to read all the articles to join the discussion but should at least familar with the topics mentioned.

Holographic Media
http://www.dailytech.com/Article.aspx?newsid=1478

Samsung 70nm OneNAND
http://www.dailytech.com/article.aspx?newsid=1606

IBM Atomic magnetics
http://www.extremetech.com/article2/0,1697,1944455,00.asp

Perpendicular Storage
http://arstechnica.com/news.ars/post/20060116-5986.html

Flash is improving by leaps and bounds and gets much cheaper every 6 months, a trend I elieve will continue. It still has a limited number of write cycles which companies seem to be ignoring commercially for now. Write speeds are still less than 10MB/s.

I think IBM's discovery with magnetics will essentially extend the life of magnetic HDD. If magnetics bits could be stored at the [sub]atomic level combined with holographic or perpendicular recording methods, superparamagnetism would not be an issue any time soon.

In the near future flash hybrid HDDs with perpedicular recording will likely be the first major change we see. I'm curious to know where people think we will go from there...

 

[Matthias99]

Originally posted by: seanp789
I've seen several news posts recently which Ill post here. Now you dont need to read all the articles to join the discussion but should at least familar with the topics mentioned.

Holographic Media
http://www.dailytech.com/Article.aspx?newsid=1478

Long-term, holographic storage looks the most promising in terms of 'bulk storage', eventually displacing magnetic tape, MO disks, and magnetic disks for both nearline and archival storage. Extremely high density combined with potentially low cost/GB would be hard to beat, although it's just not there yet in terms of technology. This is where I'd bet in the 10-year range.

Samsung 70nm OneNAND
http://www.dailytech.com/article.aspx?newsid=1606

Basically, cheaper, faster flash. It's gonna have to get cheaper and faster than that to displace magnetic hard disks, and the limited write cycles are a big problem in some areas. You get the cost down to about 1/10 of what it is now and you have a market for this stuff, at least in high-performance computers.

IBM Atomic magnetics
http://www.extremetech.com/article2/0,1697,1944455,00.asp

IE, "Magnetic RAM". Actually not a new technology -- magnetic-core RAM was used as nonvolatile storage before magnetic disks and flash memory were available back in the 60s -- but IBM is trying to do it on a nanometer scale. Been hearing about these for a long time, and it doesn't seem like it's going anywhere real fast. Could eventually displace DRAM if they can solve the technological issues.

Perpendicular Storage
http://arstechnica.com/news.ars/post/20060116-5986.html

Near-term, this is going to be widely used in hard drives, easily increasing storage density by 50-100%, maybe more as it goes forwards. This technology has already reached the point of commercial viability, and will be on shelves within 6-12 months.

In the near future flash hybrid HDDs with perpedicular recording will likely be the first major change we see. I'm curious to know where people think we will go from there...

Hybrid flash drives *could* be a step up, but flash has to get cheaper (so you can put more than a few megs of cache on a drive) before these will really take off IMO. What happens with flash will largely hinge on how cheap they can make it.

 

[patentman]

Originally posted by: Matthias99

Originally posted by: seanp789

Perpendicular Storage
http://arstechnica.com/news.ars/post/20060116-5986.html

Near-term, this is going to be widely used in hard drives, easily increasing storage density by 50-100%, maybe more as it goes forwards. This technology has already reached the point of commercial viability, and will be on shelves within 6-12 months.

Actually, perpendicular recording technology has already reached the market....Seagate is selling a notebook hardrive using this tech right now.

 

[Matthias99]

Originally posted by: patentman

Originally posted by: Matthias99

Originally posted by: seanp789

Perpendicular Storage
http://arstechnica.com/news.ars/post/20060116-5986.html

Near-term, this is going to be widely used in hard drives, easily increasing storage density by 50-100%, maybe more as it goes forwards. This technology has already reached the point of commercial viability, and will be on shelves within 6-12 months.

Actually, perpendicular recording technology has already reached the market....Seagate is selling a notebook hardrive using this tech right now.

I had read about that, but I didn't think you could buy them in retail yet. In any case, perpendicular recording is here now, or at least in the very near future.

 

[patentman]

Yeah, its odd how long it took perpendicular to get to the market. I used to work at the patent office as an examiner and there were patent applications filed as early as 1990 or so drawn to perpendicular media. I imagine the delay had to do with the fact that its very hard to cope with the interferences between domains that are packed as close together as they are in perpendicular media. Would be interesting to see how seagate dealt with this issue.

As for the perp drive now on sale, I believe its a 150GB notebook drive....

 

[TheSource]

so what new patents have come in latley patentman? tell us what the future holds

 

[patentman]

Originally posted by: TheSource
so what new patents have come in latley patentman? tell us what the future holds

LOL, unfortunately I do not work at the patent office any more. I work in a patent law firm in downtown D.C. I liked being a Patent Examiner (benefits and flexibility in work schedule were unparalleled), but its hard to say no to a 45% raise, especially when I am doing essentially the same thing I did at the PTO, just from the opposite perspective (meaning I argue from the applicant's perspective against the PTO's position, whereas before I would argue from the PTO's perspective against the applicant's position. While I do work on some applications involving magnetic and optical storage, it would be a breach of client confidentiality for me to disclose what is in those applications to anyone outside of my firm, particulary if the appliation has not yet been published (publication of an application usually does not happen until 18 months after the application is filed).

With the above in mind, I can generally say that when I was an Examiner I examined a lot of applications drawn to various type of perpendicular media, some of which claimed to achieve storage densities as high as 200GB per square inch (note that an average hard drive platter is roughly 3.5 inches (usually 3.74 inches). You can do the math and figure out the capacity of one of those platters actually is, but it should be obvious that this is much higher then current media.

The major problem associated with perpendicular media, however, and the reason it has likely taken so long to come to fruition can be summed up in 1 word, "noise." With perpendicular media (and indeed, all thin film magnetic media), there is a probelm called "intergranular exchange coupling" that becomes increasingly problematic as the storage density of a magnetic thin film is increased. As you may know, in magnetic media, individual magnetic domains are generally isolated in the grains that are present in the thin film. Thus, reduction in grain size is one of the primary ways to increase the storage capacity of magnetic thin film. However, as grain size decreases, the relative distance between magnetic domains also decreases. Although magnetic field intensity decreases with the square of distance from the fields origin, when the grain size is reduced substantially, the domains are placed in such close proximity to one another that they start to interfere with one another. This interference is called intergranular exchange coupling.

Anyways, a lot of patents that I examined were directed at correcting or lowering the amount of intergranular exchange coupling in magnetic thin films, even those having very tiny grains and high capacity. Usually this involved the use of various underlayers (layers which are deposited between the substrate and the magentic layer for the purpose of controlling the crystalline structure or grain size of the magnetic layer) or the use of a "soft" magnetic layer underneath the magnetic recording layer. A soft magentic layer is a layer of magentic material that has relatively low coercivity, meaning that the direction of the magnetic domains in the layer changes relatively easily in response to an external magnetic field. These layers have been used in various ways, e.g. to "pin" the magentic domains of a recording layer in a certain direction (improving coercivity of the recording layer); to counteract intergranular exchange coupling; and to improve signal to noise ratio (measured by a read head flying over the surface of the media)

I could go on and on, but basically it comes down to this. All of the magentic recording media applications I examined were concerned with at least one of the following: 1) Increased storage density; 2) increased signal to noise ratio 3) increased coercivity; and 4) reduced production costs. The ways in which each of these is accomplished are widely varied, and so I will leacve them for another discussion.