Question I can't be the first person to think of this idea?

[ch33r]

So to make hard drives speed up to compete with SSDs, why don't they add more than one needle to read the platters, just on opposite sides of eachother so they don't interfere with eachother? Why is this not a thing?

 

[ch33r]

About what?

I don't see what the issue is. You aren't really changing how the hard drive works. You're just adding more readers. Kind of like when CPUs hit the clock speed wall so we added more cores Dual-core hard drive... Lol... No?

 

[dudio]

I kind of want to buy 4 15000 RPM drives in RAID 1.
See how they compare to a dirt cheap ssd.

Completely stupid idea but 60k RPM. Holy crap.

It would still be maxed at 15000 rpm..

 

[USAFRet]

I don't see what the issue is. You aren't really changing how the hard drive works. You're just adding more readers. Kind of like when CPUs hit the clock speed wall so we added more cores Dual-core hard drive... Lol... No?

You're not the first person to think of this.
It has almost certainly be tossed around, and discarded due to the above reasons, and probably many more.

Get a couple of engineering degrees, and propose your idea to Hitachi.

 

[ch33r]

You're not the first person to think of this.
It has almost certainly be tossed around, and discarded due to the above reasons, and probably many more.

Get a couple of engineering degrees, and propose your idea to Hitachi.

Ok, so then my next question is this, why don't they make it so that the tip of the needle, the tiny little component that actually reads... why don't they make it bigger so it covered more area of the platter and then it can fetch more data in one rotation????

 

[USAFRet]

Ok, so then my next question is this, why don't they make it so that the tip of the needle, the tiny little component that actually reads... why don't they make it bigger so it covered more area of the platter and then it can fetch more data in one rotation????

Because that's not how the drive works. A head cannot read multiple 'tracks' at once.

In addition, a larger head weighs more, and would move slower.

The current design for the typical hard drive has been refined over several decades. It is what it is.

 

[ch33r]

Because that's not how the drive works. A head cannot read multiple 'tracks' at once.

In addition, a larger head weighs more, and would move slower.

The current design for the typical hard drive has been refined over several decades. It is what it is.

Ok, so why don't they design a new head, something completely new, a piece that doesn't move at all, and stays stationary, and covers the entire width of the platter. Then an entire 500GB single file can be fetched in one rotation. I mean.... rhe actual piece at the end of the head that reads the track is very tiny. So put a bunch of those on one mechanism, all lined up side by side across the platter, stationary, doesn't move. Each one covers one track, side by side. The platter just spins and is covered all across so the "head" (or whatever it would be) doesn't have to move at all

I bet what's going through your mind right now is: "My god you're so stupid, obviously if that was possible, they would have done that"

Yes... I know that.... I get that.... but Im trying to figure out like.... why hasn't this happened?

 

[AllanGH]

Because the complexity of reading all of the parallel heads is ridiculous to engineer, exorbitantly expensive, and is completely unnecessary.

I recommend some physics, electronics, and engineering courses. Having that set of knowledge would make all of these answers unnecessary.

 

[ex_bubblehead]

And that's what I was asking. Why is this not a thing?

Ok, answer this. How would you keep all 2,3,4 sets of heads aligned? Just one write in the wrong place and you can kiss your data goodbye. It's difficult enough with the one set used now.

In addition, how would you expect to keep this in the current form factor?

 

[ch33r]

Ok, answer this. How would you keep all 2,3,4 sets of heads aligned? Just one write in the wrong place and you can kiss your data goodbye. It's difficult enough with the one set used now.

In addition, how would you expect to keep this in the current form factor?

Well, the form factor would have to be expanded a bit, so yes the standard would change. Barring that, I don't know

 

[ex_bubblehead]

Well, the form factor would have to be expanded a bit, so yes the standard would change. Barring that, I don't know

So, everyone wanting to install one of these monstrosities would have to buy a new case, power supply (or 2) to accomodate it. Likely to cost a couple of thousand for a usable sized drive as well. Not gonna happen. You've been given many, valid reasons as to exactly why this isn't a thing.

 

[ch33r]

So, everyone wanting to install one of these monstrosities would have to buy a new case, power supply (or 2) to accomodate it. Likely to cost a couple of thousand for a usable sized drive as well. Not gonna happen. You've been given many, valid reasons as to exactly why this isn't a thing.

Ok so what about the other idea I proposed aside from that?

 

[ex_bubblehead]

You haven't proposed anything practical so far.

 

[ch33r]

You haven't proposed anything practical so far.

Well there was the one about having 4 heads, one on each side.... but what about this one....

Why don't they design a new head, something completely new, a piece that doesn't move at all, and stays stationary, and covers the entire width of the platter. Then an entire 500GB single file can be fetched in one rotation. I mean.... the actual piece at the end of the head that reads the track is very tiny. So put a bunch of those on one mechanism, all lined up side by side across the platter, stationary, doesn't move. Each one covers one track, side by side. The platter just spins and is covered all across so the "head" (or whatever it would be) doesn't have to move at all

 

[AllanGH]

You haven't proposed anything practical so far.

The tracks on a platter are far smaller than you can imagine. You can't fit enough physical heads in, side by side, to take advantage of the areal density of the magnetic storage media.

 

[ex_bubblehead]

The tracks on a platter are far smaller than you can imagine. You can't fit enough physical heads in, side by side, to take advantage of the areal density of the magnetic storage media.

Not to mention the almost impossibility of producing usable quantities as 100% of those sub micron size heads must be fully operational. A single head failure trashes the entire array. Rough guess $1000 per assembly. Back to 5-6 figure prices for mass storage.

 

[ch33r]

Not to mention the almost impossibility of producing usable quantities as 100% of those sub micron size heads must be fully operational. A single head failure trashes the entire array. Rough guess $1000 per assembly. Back to 5-6 figure prices for mass storage.

Uggh… well theres gotta be some kind of solution here. I mean... mechanical drives have been functionally the same for 50 years. The only difference is now we have SATA speed instead of what we had to start. Something needs to change, because if not, the way I see it.... Hard Drives are basically done because SSDs have dropped in price enough

 

[AllanGH]

well theres gotta be some kind of solution here.

A solution to a nonexistent problem....

Whatever.

 

[jsmithepa]

Quick, call Seagate!

Mechanical drives will never beat solid state.

 

[Darkbreeze]

Im confused

Clearly.

 

[Darkbreeze]

the way I see it.... Hard Drives are basically done

Finally, you get it. This, is exactly correct.

Hard drives ARE done. To loosely quote John Hurt, "They might not be being shown the door, yet, but they are certainly being handed their hat".

 

[CountMike]

Those are not needles, those are heads similar to tape decks. They cover several tracks at once unlike what needle and grooves do in gramophone but more like stereo and quadraphonic and audio tapes. It's possible for one head to read different streams from different tracks. That's what modern HDDs made faster than old ones. That has more influence on actual data speed than rotational speed.

 

[DSzymborski]

I don't see what the issue is. You aren't really changing how the hard drive works. You're just adding more readers. Kind of like when CPUs hit the clock speed wall so we added more cores Dual-core hard drive... Lol... No?

Sorry to be blunt, but none of this makes any sense. One has physically moving parts, the other has logic gates. Multi-core solutions exist because they scale well and don't dramatically increase the failure rate of CPUs. Your solution for a hard drive doesn't scale well and dramatically increases the failure rate of hard drives.

 

[fiasse1]

RAID0 is a thing...

 

[DSzymborski]

RAID0 is a thing...

Unfortunately.

 

[USAFRet]

Ok, so why don't they design a new head, something completely new, a piece that doesn't move at all, and stays stationary,

They DID.

It's called a Solid State Drive.
No moving parts.
Really fast data access.

You might want to look them up.