There has been some confusion about whether Western Digital's IntelliPower technology varies a drive's rotation speed or not. Well, it doesn't. A WD spokesperson said that IntelliPower does not vary a drive's spin speed while it is in use. The drives have fixed spin speeds and these may vary with individual drive capacity points …
The mystery is solved!
I'm glad to see this particular tech myth put to bed. You have to give it to WD for a clever marketing trick though...
Not an easy task
It would be incredibly difficult to vary the spin speed of a modern hard drive while it's in use. I can't imagine the mammoth engineering effort would be worth the little power it would save.
I have a nice selection of 400Gig RE2 drives here that vary their spin speed... They spin up... Click... Spin down... spin up... Click... Spin down...
Does spinning speed make a difference?
Why should it really take more juice to spin at a faster speed? Once you're up to speed you're just fighting friction and no acceleration is involved.
I can understand seek speed making a difference because massive accelerations are involved and thus faster seeking means more power is needed. You could likely make power gains by reducing seek speed.
Numbers for morons
Morons will always assume bigger numbers mean better performance.
Why listen to a CD on a 1x CD player when you can listen on a 52x? That way it will only take a minute to listen to your favourite albums.
Want one racing stripe on your fiat scetiento? no, two will make it go twice as quick!
Want to get data from a 5400rpm HDD with 100MB read speed? no, it will come off much faster with a 7200rpm disk with 100MB read speed.......
On the other side of the coin are the marketing folks who come up with this bolony. Intellipower? Does it intelligently control the power? No, it is a fancy name that indicates the engineers have come to some kind of compromise with the design.....
"Why should it really take more juice to spin at a faster speed? Once you're up to speed you're just fighting friction and no acceleration is involved."
Energy = force * distance.
Friction is a force.
The thing is spinning.
@Charles and AC 10:35
Friction of any kind is proportional to the square of velocity, hence a 10k disc will require four times as much power to spin it as a 5000 RPM disc of the same design.
Seek speed makes no difference - this is mostly down to how long the head takes to move across the disc, not waiting for the right sector to come around.
Generally speaking, a faster rotation speed means faster data transfers, whatever the interface speed - think of everything as a bottle-neck, and you will get the picture.
And here I thought that Micro$oft still had a monopoly on "intelli-[fill in the blank]" for hardware...
Mine's the insulated one with the earmuffs- I'm off to the server room.
Power and discs
A lot of the power used to keep a disc spinning is used to overcome air drag. The faster the disc spins, the greater the amount of drag which goes up as a cube of the speed although flat surfaces and discs are a bit amonalous in that respect.
As for using variable-speed discs, controlling the head height above the surface requires a particular disc speed to maintain the aircushion at the correct level for optimum data reading and writing, making a variable-speed drive very difficult to design and manufacture (it could be done with multiple hR/W heads designed for different rotational speeds but it wouldn't be cost-effective).
It's pretty much moot anyway as the move towards solid-state mass storage is growing apace and the time of rotating mass storage media is just about over.
Rob, you were right on the money until you talked about SSD taking over rotating media. They will both exist in parallel for many many years. The price performance per GB and reliability will keep spinning media around (yes you are right I said reliability!).
The big money will be in pulling these storage techs together with the controllers and everything else up the storage chain.
the difference is that you could be waiting longer for that 100MB to come round under the heads while the drive spins slowly.. rather than having it start to read much faster of a drive with a faster spin speed
especially since that 100mb is unlikely to be in one nice big continuous file
go look up rotational latency