Analysts Predict Acceleration in HDD Storage Density Growth

[CaedenV]

[citation][nom]magicandy[/nom]Put on those tin foil hats here, but have you ever pondered that maybe HDD density/platter/read head advancement is being intentionally slowed so their speed doesn't catch up with SSDs and make the billions of dollars memory companies are investing irrelevant? For all we know we should already be at 20-30 TB drives which at least match the read/write speed of the larger SSDs but have much higher storage space and far greater reliability.[/citation]
OCZ isn't exactly paying Seagate to slow down their tech. mechanical storage is great, but it is nearing the end of the line. It is like moving from floppy to CD, and then CD to DVD, and then DVD to flash drives. It is just a progression, and while there are still some advantages to the old formats, the gains in the new formats simply make the old ones disappear in time. No conspiracy, just the end of one era, and the beginning of the next.

 

[CaedenV]

[citation][nom]blazorthon[/nom]I doubt that there is no way to do it without increasing form factor. Even if there wasn't, then slim, but ultrafast and very high capacity 5.25" HDDs would be viable. Regardless, I am very sure that it could be done within a 3.5" drive form factor.[/citation]
more parts=more failure points and extra expense. The last thing we need are drives that fail more often, and cost 2-3x as much as they currently do. Besides, we already have the tech in SCSI/SAS drives which have an actuator per platter, and can work much better than consumer drives... but nobody buys them... because they cost $300ea at the entry level.

 

[TeraMedia]

I can't even imagine rebuilding a RAID 5 or RAID 6 array made out of 6 TB SATA drives.
Step 1: Click start
Step 2: Go get cup of coffee
Step 2a: Fly to Colombia
Step 2b: Travel to coffee bean plantation
Step 2c: Harvest coffee beans...

We are reaching the pragmatic end of 3.5" HDD size growth already. The cost of a failed 3.5" drive is simply too high. If you have redundancy or a backup, a restore or rebuild will take hours or days for a 2-4 TB drive. If you don't, then that's an awful lot of data to lose...

You can fit up to five 3.5" drives in 3 adjacent 5.25" bays with removable trays. In a RAID 5 array, that's up to 16 TB. In the same 3 bays (but yet with less depth), you can fit up to eighteen 2.5" drives. If you used RAID 6 to add a bit more redundancy, that would give you the same 16 TB, but with much higher throughput and probably lower heat output and power consumption. Or you could split the array in half and improve latency too. My point: increasing areal density is fine if speeds improve and reliability stays at least constant. But unless you're the library of congress and rebuild / recovery times don't matter, you probably don't need a higher-capacity 3.5" drive than what is already available.

 

[blazorthon]

[citation][nom]CaedenV[/nom]more parts=more failure points and extra expense. The last thing we need are drives that fail more often, and cost 2-3x as much as they currently do. Besides, we already have the tech in SCSI/SAS drives which have an actuator per platter, and can work much better than consumer drives... but nobody buys them... because they cost $300ea at the entry level.[/citation]

$300 for what capacity? And really, we both know that consumer versions would be cheaper than non-consumer versions anyway, especially if they switched SCSI and SAS with SATA.

 

[danwat1234]

[citation][nom]CaedenV[/nom]more parts=more failure points and extra expense. The last thing we need are drives that fail more often, and cost 2-3x as much as they currently do. Besides, we already have the tech in SCSI/SAS drives which have an actuator per platter, and can work much better than consumer drives... but nobody buys them... because they cost $300ea at the entry level.[/citation]

I have not heard of this. What drive has independant actuators?