Quite a number of you have written in with the
news about
Sony and Nippon Zeon Co Ltd announcing plastic hard drives. The current estimate is that it will cut production costs by 30-40%, on 5 gig drives. No word, however, on when production/testing will start.
Slashdot Mirror
1 - Solid State HDs aren't new. I saw a review of one of the first ones in Boot about a year ago.
2 - EEPROM is expensive and slow. That's why it's not used for solid state hard drives. Usually, PC100 SDRAM is used, with a battery pack and a regular magnetic hard drive. When the power goes out, the battery pack keeps the data in the RAM while the hard drive makes a copy of the data. Protection of data in a power outage is essential for these things because you generally only can afford and only need a solid state hard drive for a server. The ones Quantum makes can do sustained 30 MB/sec off the RAM and average seek times of 3 - The ones Quantum make (the only ones I've seen, I don't work for Quantum) go up to 3.2 GB, although I'm sure the prices are all 5 digits.
4 - Heat? The way hard drives are manufactured now, the aluminum is just the actual platter and doesn't carry any data. All the data is on an iron oxide coating. (that's why hard drive platters are yellow-orange) My understanding is that these new drives would be the same on the outside (metal casing) but use plastic instead of aluminum - plastic platters coated also with iron oxide. Even so, the main cause of heat in hard drives is from air resistance to the platter/heads over the platter, so the plastic might still be affected. Might be interesting to see what happens. I run 7200 and 10000 RPM hard drives and I've never had a problem with heat though. Warm to the touch, but not hot.
Jonathan Wang
Okay, the whole big thing about SCSI is *not* about transfer rates...unless you're a pro video editor, you won't saturate a bus anyway...the hard drive is the limiting factor (you *need* a RAID array for this to become an issue).
SCSI can re-order commands in the queue so that they work more efficiently. Imagine a file server. Jim Smith is uploading his analysis of Red Hat's stock performance. Sally Jones is trying to get an index of reports. The two are reading and writing (and to two different places on the drive).
(Each chunk is the result of a single request)
Here's ATA:
-shove chunk of data to disk for Jim
-get chunk for Sally
-shove 2 chunks for Jim
-get chunk for Sally
-shove chunk for Jim
...and so on
Here's SCSI:
-Shove 10 chunks for Jim
-Get 10 chunks for Sally
-Shove 10 chunks for Jim
-Get 10 chunks for sally
even if the requests are coming in in an intertwined manner. Seek is the big factor here. In this scenerio, SCSI is going to beat the pants off ATA. If we just have a single-user workstation, where Bill White is reading an mp3 from his hard drive, as long as the ATA hacks are enabled (DMA and whatnot), the two are going to do about the same. Remember, though, that VM and background processes may be doing some writing. On an average Win95 user's system, the two will be about the same. On a fileserver with lots of simultaneous requests, there's a big difference. On a Linux box with swap chewing away and a file or two writing, I can only guess that there would be a difference.
_The Basics of SCSI_ is a good handbook to get if you're interested in the way SCSI works (dunno if they updated it for Ultra, though).
I'm a fan of SCSI, but ever since RAM prices dropped, with Linux's really good caching (well, compared to the Mac or Windows), I can only say that you're better off getting more RAM instead of SCSI most of the time. Pop 512 more megs in instead of a SCSI card and a big SCSI drive...you'll probably get more benefits.
Incidently, Firewire is supposed to use the SCSI command set, so I guess it has all the features of SCSI. If it becomes a consumer item, maybe Firewire drives will be cheaper than their SCSI counterparts, and I'll be able to pick a good one up.
How many different fruity colors will these come in?
---
Have a Sloppy night!
As copyright owner of this comment, I authorize everyone to defeat any technological measure which limits access to it.
Um. Didn't Benjamin Franklin do experiments with fuzz and glass? Seems really nasty if a bit of fuzz or something gets in there...you'd probably build up charge like crazy. 'course, hard drives are supposed to be sealed, right? (Wonder if they're waterproof...)
Anthrax spores can last quite awhile without special solutions. Though, they aren't actually doing anything in spore form.
Reading an interview of Syed Iftikar founder of Syquest and Castlewood, he mentions that Castlewood are looking into replacing aluminium for the orb platters.
http://www.smsmag.com/archi ve/1998/wintermag98/5929.htm
But as with all science projects it's very secret until a product is announced and all patents protected. He says something about Comdex '99, but my bet is that this is something for the next generation of Orb's.
Patrik
hahaha funny jo... hey wait, you're right... and i have one too!
oh, does no one remember "My First Sony"?
matguy
Net. Admin.
matguy(.com)
I'm glad someone got the reference in the subject...
All the IBM 2.5" drives Ive ever opened up (going back to '96 or so) were glass. (Now..why I opened them up..gee..they break real easy..who woulda thunk?)
Well, it would spin up faster for the same input of energy. OTOH, it would have less inertia holding the RPMs steady. Friction might need to be significantly decreased. OTOH, with all the really small disks being announced these days, they *must* have solved that problem.
I think we've pushed this "anyone can grow up to be president" thing too far.
Will they have to pay royalties to Fisher-Price for this one?
Luckily, right now that bacteria can't survive without a lot of *help*. You know, special solutions, temperatures, other stuff. And even then, they don't last long.
Would the reduced mass (and hence lower wattage motors) of a plastic spindle mechanims off-set (to a limited extent) the poor heat disipation?
Perhapse using a magnesium space frame skeleton that could double as a heat sink.....
Glass platters are used in notebook harddrives, at least those from IBM. And glass platters are starting to make their way into hard drives for other markets to like the IBM Ultrastar 36ZX and 18LZX for servers.
Ever seen a floppy disk?
I'm a loner Dottie, a Rebel.
This will be great if two things happen. First, more platters in the same sized hard drive, which means an increase of capacity without signifigant cost. Second would be higher RPM speeds since the drives weigh less than their metal equivilents. I wouldn't mind a 10,000rpm EIDE hard drive. In UDMA 33/66 that would really fly.
I'm a loner Dottie, a Rebel.
I doubt it. Hard drives get hot, as I understand it, because of friction between the platters and the air inside the drive. For a given speed, I don't think it matters whether the platters are aluminium or plastic or glass or plutonium. If the plastic platters can be SMOOTHER than the metal ones then maybe that'd help, but there are all sorts of funny turbulence and Tesla Turbine-esque issues here that I don't understand.
I guess I was a little lacking in description, or I wasn't expressing what I had in mind very well.
I was not thinking of curent technology.
I was not thinking of EEPROM as such, but a device with EEPROM like qualities (Able to store data with no power, electronically eraseable, fast (at least compared to the magneto-elecromechanical sloths of today).
What I had in mind was something that was plastic (cheap and durable),
fast (near RAM speeds),
logically huge (entire digitized feature length films, say all of the Indiana Jones trilogy on a single cartridge),
physically small (small enough to easily fit into a pocket, but not so small that my clumsy fingers can't manipulate it. About DAT sized).
For example, the data stored on a CD or a DVD is (practically speaking) stored in plastic. But we still need to spin it around and move a laser back and forth over it. What if we set up an entire array of microscopic lasers and sensors that could read the entire surface of the disk in one shot?
Latency would be merely the time it takes to address a particular sensor (negligible). Data transfer rate would be merely the rate at which it took to step through the sensor array (Blazing fast).
I'm thinking something that is rugged enough to be able to withstand dropping on the floor, getting coke spilled on it, blah, blah, blah.
The perfect companion to the ChiaMac I would say... :)
I've finally had it: until slashdot gets article moderation, I am not coming back.
My only concern with plastic platters is heat stability. If the platters get hot enough, it's quite possible the heat will bake the magnetic patterns off the surface (try cooking a magnet). It's probably a good guess that plastic conducts heat poorly (unlike aluminum), so that might be a problem (expecially if you discover your power supply fan died).
Heh. Add beside "My hard drive crashed," "My hard drive melted" to the list of late homework excuses }}:-)
What ever happened to glass hard drives? They're supposed to be heat stable (and create a very smooth surface - less heat).
AFAIK the battery is only for the CMOS memory where the BIOS keeps config info; I've never lost the BIOS by pulling the battery.
:) ), and put them back in to find that everything is OK. Unless there's an onboard battery (something that I doubt in the little 8-pin EEPROM for storing the program for a PLC I was playing with recently) the EEPROM retained data without power.
I've pulled EEPROM chips, carried them around, chewed on them (not hard enough to damage the case
I'm going to assume that platters are already made out of hi-tec plastics, but sony is just talking about putting a plastic shell around the electronics instaid a mettal one
"Subtle mind control? Why do all these HTML buttons say 'Submit' ?"
ReadThe ReflectionEngine, a cyberpunk style n
I wonder how many of these things would melt.
Anyone knows that when plastic burns it emits a toxic gas. Perhaps it's deliberate.
I recall stories of glass platers floating around a couple years ago. They were to be lighter, faster and cheaper to produce. I even saw some proto-types in a materials engineering course. Whatever happened to this idea? Perhaps they didn't make enough noise about "market share" and "stock prices". Too bad development and investment go hand in hand.
--Let's hack root on 127.0.0.1 --panZ
Anthrax is not one of the bacteria that have this property.
Sure, you were right, but the information was useless.
Last I heard, they have found some strains of some Pseudomonas species that could break down some plastics and use them as food. Pseudomonas is a rather fun genus like this. There are species that can also use common soap, wood fiber (paper?), and even some detergents (with phosphorous) as food sources.
However, they are also opportunistically pathogenic. This is not so much a problem because of the ability of causing infections in this case, but more because it means it is adapted to living inside the human body, ie: it requires a high degree of moisture and 37 degrees Celsius (98 degrees for us silly americans). As a side note, they probably would not react well to the iron on the surface of the disks.
However, I suppose if you would keep the inside of your case at 37 C and 80% humidity, you might get a bit worried about this.
Summary: I think there are better chances that your aluminum hard drive will spontaneously combust than a plastic hard drive getting eaten by bacteria.
EEPROMs do need electricity to keep their data stored, but they only need 1.1 volts and require very very low wattage so the little lithium battery you see on the motherboard (open it up and look, I promise you it's there somewhere) is enough for the BIOS et al to retain their memory.
I'm a loner Dottie, a Rebel.
So why not create a near-vacuum in the drive? Shouldn't be too difficult, and I can't think of any problems it might cause.
Seeing how plastic is more of an insolator than a conductor (of heat along with electricity).. how do they plan to use these is systems where the harddrive is active alot? Metal hard drives have the advantage of what they are made of.. metal, whereas the new harddrives would be sitting in basically a thermous. (unless they plan on using plastic for only the platters and not the whole shabang?)
Make up your own joke about plastic hard drives to go along with toy operating systems from Redmond.... What next? Nerf monitors? Play-do mice? Lego keyboards? (That would be kinda cool actually...)
Would I be able to put a magnet up to it?
I know that the cost of production of these plastic drives will be much cheaper as plasti c injection molding produces products much faster than metal. But how reliable will these drives be and how quick?
Plastic is a lousy conductor of heat. Without metal HD cases to transfer the heat to the PC case and keep the drive cool, I suspect these drives will heat up a lot more. Or worse, warp and melt down. :)
From the story: ``Sony has been promoting various types of hard disks with several partners, and it won't start investments until it becomes clear which technology is the best,'' said Masashi Kubota, an analyst at ING Barings. This seems to suggest to me that this is fairly early on, and it will be some time before we see these. Oh well, I have enough storage to last me quite some time. I find denser media more exciting than cheaper media.
Trees can't go dancing
So do them a big favor
Pretend dancing stinks!
I think either most of you are confused or I am (it's probably me.) I think that the hard drive case etc is plastic. Not the actuall platter. That wouldn't make any sense, because then platter needs to be metal in order for the heads to read/write data to it since it works by electromagntism. Maybe their just going to coat the plastic platters with some sort of metal coating. Would some one please set me straight here.
Check out Compubrick for some lego keyboards, among other things :-).
----
----
Open mind, insert foot.
Will the heat generated cause degredation in the platter's ability to hold data? Will this also mean that drives will use less energy? Great for ploptops, eh? And since the platters could be lighter, does this mean faster rotation? I wish this article was a bit more complete...
-
ping -f 255.255.255.255 # if only
Yes, the article is talking about making the actual platters out of plastic rather than aluminum. They will coat the plastic platters with a magnetic coating, just like they do now with the (non-magnetic) aluminum ones.
I've got one already. It works with Linux too. I call it /dev/fd0
The article mentioned talks between Sony and Castlewood systems. Castlewood Systems make the ORB drive. This could mean *much* cheaper removable media for the ORB or quite possibly ORB variants.
I wonder if the current ORB drive could read a plastic disk?
Also wouldn't platic platters be more durable?
Couldn't we focus on storage media that didn't have moving parts? Even if the platters were moving at 30,000 RPM, the cpu could chew through 100,000 instructions by the time the platter spun around once. Mechanical devices are just too slow. How about a 200G plastic EEPROM cartridge about the size of a DAT tape?
How are they going to make long chain polymers have a smoother surface than a crystaline aluminum alloy? If the mass of the platter is less, can they get higher spindle speeds? Faster spin up and down times? Can they also start making cheaper housing assemblies, to cut the weight and price even more? Boy, I hate contentless stories :-)
That's right! Pick your size! I'll hand craft each one of them to your liking. I will soon be mass-producing them with a cutter and conveyor belt. Guaranteed to fit inside any computer!
HAH! I can use ICE and make sure the drives don't get too hot. They won't crumble to bits AND they never get dusty nor chip. Scratch your platter? Just shave off the top layer of ice and boom, you've reformated and made a new disc!
-
ping -f 255.255.255.255 # if only
No, see comment #14.
There are bacteria that consume oil derivatives, right? I wonder if there's something that would eat the plastic in these drives. Heh.
That would restore meanings to a "bug"...
That might provide an interesting, if silly and slow, way to destroy one's data if being raided: smear it with a colony already growing on agar, and figure that they probably won't autoclave the disk...
Only the dead have seen the end of war.