Multi-Arm Hard Drives???

[techfuzz]

How come we haven't seen these types of devices yet? Wouldn't it be logical to place actuator arms on opposite sides of the platters and allowing data to be read at 2x the speed of current drives? Discuss...

techfuzz

 

[SagaLore]

The way the binary stream is written to the platters, I don't think another arm on the opposite angle would function properly. You'd need even more complicated circuitry to sync it up - and then you'd end up with the "left hand doesn't know what the right hand is doing" problem, which involves even more circuitry. In the end, you add a ton of overhead/latency.

 

[fuzzybabybunny]

I can't wait for mass produced cheap RAM drives

 

[JeffreyLebowski]

Why come up with ways of making mechanical drives when you can spend money on making solid sate drives that are many times faster, use less power and create less heat?

 

[Demon-Xanth]

There was one. It was one of the Seagate Barracuda 2 models, but the dual head configuration wasn't compatible with DOS. It was a 2GB 50 pin SCSI drive.

 

[Demon-Xanth]

http://www.seagate.com/support/kb/disc/tp/cuda2.html

 

[BoberFett]

Originally posted by: JeffreyLebowski
Why come up with ways of making mechanical drives when you can spend money on making solid sate drives that are many times faster, use less power and create less heat?

Maybe because with the current level of solid state technology, putting 250+ GB in a solid state drive would be larger, use more power, and create more heat than a mechanical drive?

 

[andy04]

I feel it would only be helpful of long sequential read/write not for ramdom access and small r/w

 

[techfuzz]

Ok, so what if the arms were independent of each other and could read/write wherever they chose (as long as there was no conflict)? How much performance could you gain then?

techfuzz

 

[spidey07]

Originally posted by: techfuzz
Ok, so what if the arms were independent of each other and could read/write wherever they chose (as long as there was no conflict)? How much performance could you gain then?

techfuzz

Multiple platters. There's your answer. Been around for a LONG time.

 

[techfuzz]

Originally posted by: spidey07
Multiple platters. There's your answer. Been around for a LONG time.

The issue isn't multiple platters, it's multiple-arms reading those platters. I think you could gain almost 2x the performance of current drives with a second arm (and heads). Allowing the arms to work independently could permit tracks on the inside and outside of platters to be read at virtually the same time.

techfuzz

 

[acole1]

Originally posted by: spidey07

Originally posted by: techfuzz
Ok, so what if the arms were independent of each other and could read/write wherever they chose (as long as there was no conflict)? How much performance could you gain then?

techfuzz

Multiple platters. There's your answer. Been around for a LONG time.

The thing with the current multiple platters configuration is that both/all heads are attached to the same arm.

So, in theory you could use a separate arm, per head, per platter to speed up random and non-sequential access... but then I think you would increase your risk for damage and corruption to the disk exponentially.

If you could have each arm independent, and each platter spinning independently, it would practically be RAID, if you think about it; so based on that you can get a rough estimate on the performance gains possible.

 

[LegendKiller]

Originally posted by: fuzzybabybunny
I can't wait for mass produced cheap RAM drives

When you can buy a 750GB drive for $300 let me know. OK?

You guys forget that consumer and enthusiast usage is a very small amount of total drive production and sales. Storage, mainframes, and servers, need huge drives and it's becoming a bigger need as time goes on, people want to store everything. With the way magnetic recording is increasing, it's going to be around a very long time.

 

[jimbob200521]

Originally posted by: BoberFett

Originally posted by: JeffreyLebowski
Why come up with ways of making mechanical drives when you can spend money on making solid sate drives that are many times faster, use less power and create less heat?

Maybe because with the current level of solid state technology, putting 250+ GB in a solid state drive would be larger, use more power, and create more heat than a mechanical drive?

Not to mention cost about a bazillion times more...

I think the idea of of dual arms is a good one, but who knows how much work it would be to get those arms to sync up and not read the same data twice, as well as the pure mechanics of fitting double the hardware (arms, gears to move them, control circuits, etc) under an already very cramped drive interior. Maybe in the future it would be possible, but I don't know that it could be done with today's current tech and not increasing the drive size.

 

[OS]

files are mostly in one spot on the platter, there's not really any benefit to having lots of arms moving around independently when the data you need is only at one spot on the disk.

lots of arms might be useful for like a server/multiple simultaneous access use, but since the arms are fixed, how do you prioritize which one gets priority spinning the disk. I guess you could have some kind of arm moving assembly around the platter also, but that's diminishing returns.

 

[randay]

R A I D

 

[MichaelD]

I can't wait until hard drives....


...wait...


I can't wait until people learn how to post in the correct forum. This is just blatantly the WRONG FORUM. How about General Hardware? Maybe even Highly Technical?

Off Topic is for girls/booze/money/bitching about work.

 

[OrganizedChaos]

it took the manufacturers this long to build drives with one arm that can't crash into a platter, now you want them to try for 2?