[Dailytech] Hard Rectangular Drive May Replace Current Spinning Discs - Page 4 - Overclock.net

oliverw92 · 06-19-09

Quote:

Originally Posted by 003

Yes, there have actually been audio transducers designed on that principle.

piezo buzzers

they can be made tiny

OmegaNemesis28 · 06-19-09

Noooo - I rather have width then length on the hard drives because of space reasons.

denrocks2 · 06-19-09

I'm not going to pay thousands of dollars for one of these.

But we all know they will start in the GB's and slowly move up to make their money.

DuckieHo · 06-19-09

Quote:

Originally Posted by noldevin

I think multi-head standard HDDs are a better idea. I mean really, why hasn't anyone tried it?

Where are you going to stick the additional heads, arms, and motor?

Quote:

Originally Posted by OmegaNemesis28

Noooo - I rather have width then length on the hard drives because of space reasons.

These can be configured to be any size....

superon · 06-19-09

Quote:

Originally Posted by denrocks2

I'm not going to pay thousands of dollars for one of these.

But we all know they will start in the GB's and slowly move up to make their money.

Storage density doesn't increase incrementally on consumer devices because the company thinks they can make more money that way. It's actually technological limitations that limit the growth of storage density.

Kirmie · 06-19-09

Quote:

Originally Posted by Madman340

I don't see how you think a huge spinning disk would be more efficient. (What we have now really)

The low power consumption is from the properties these heads (p-whatever crystals) are made out of. Really all they have to do is have the heads in exactly the same areas at the perfect distance from eachother... They would reach all areas of the rectangular disk.

Right now we have a single, moving head per platter, I proposed that they use mulitple non-moving heads per platter. The rectangle would allow all spots to be accessed, but some spots would take longer to access than others do to the motion.

Lets say you have the digit 1,2,3,4,5 on the ground in a hallway and one person had to go to each number by going across the entire hall each time if they wanted to read it. Would it take the person longer to read these numbers in order 1,5,1,5,2,4,1,5 or would it take longer to read 2,4,2,4,3,2,3,4? The further those numbers are space out, the longer it takes to reach from one end to the other, similar to going from the outside to the inside part of a platter and vice versa. This could be reduced by having a nearby head overlap the reading area. In the same situation above, this would work like having a person who could read any number as long as there were exactly two other numbers between you. If this only happened on the right it would involve two people (heads) or three people for both sides. Then if you read the number 1 it would only take you two moves to read the number 5. This would not be possible on the outside of the platter since only the outside head would be able to reach it.

My proposal would have the number 1,2,3,4,5 placed on a merry-go-round with an arbitrary number of people, lets go with two for giggles. If the two were spaced apart so that if one was reading the number 1 the other would be reading the number 5. They would read the numbers by having the merry-go-round spin so that if a person read the number 1 then they would read the number 2 next. So lets do the math with the scenarios with a starting position of the original person at 3 and going left or lower first, always left for merry-go-round (second person is two numbers right or higher, third person two numbers left or lower). Also, only one number can be read per stop though the starting spot can be read.

Order of 1,5,1,5,2,4,1,5:
Back and forth 1 person: 2 + 4 + 4 + 4 + 3 + 4 + 5 + 4 = 30
Back and forth 2 person: 2 + 2 + 6 + 2 + 5 + 2 + 5 + 2 = 26
Back and forth 3 person: 0 + 4 + 4 + 4 + 1 + 2 + 1 + 1 = 17
Merry-go-Round 1 person: 2 + 1 + 4 + 1 + 3 + 3 + 3 + 1 = 18
Merry-go-Round 2 person: 2 + 1 + 1 + 1 + 1 + 3 + 3 + 1 = 13
Merry-go-Round 3 person: 0 + 1 + 1 + 1 + 1 + 2 + 1 + 1 = 08

Order of 2,4,2,4,3,2,3,4:
Back and forth 1 person: 1 + 4 + 4 + 4 + 3 + 1 + 3 + 1 = 21
Back and forth 2 person: 1 + 2 + 6 + 2 + 1 + 5 + 1 + 1 = 19
Back and forth 3 person: 1 + 2 + 2 + 2 + 1 + 1 + 1 + 1 = 11
Merry-go-Round 1 person: 1 + 3 + 2 + 3 + 1 + 1 + 4 + 4 = 19
Merry-go-Round 2 person: 1 + 3 + 2 + 3 + 1 + 1 + 1 + 2 = 14
Merry-go-Round 3 person: 1 + 1 + 1 + 1 + 1 + 1 + 1 + 2 = 09

In both situations the reading in a mery-go-round (spin) motion requires the least movement and thus the least seek time. Another way to put it would be that the maximum number of movements for the original head to reach a number in the back and forth method is eight and occurs when you need to read the number 1 or 5 twice in a row. The maximum number of movements needed to reach a number in the spin is 4 and this occurs whenever you have to read any number twice in a row. The only number with an equal or lower max than that in the back and forth method is reaching the number 3 which has a max of 4 as well. Either could be faster in a specific situation but the mery-go-round method will be the fastest on average.

Also, if somebody is good at math and not too lazy could they put that into a formula for me?

Edit: Sorry if some of that is confusing, wrote it out in chunks while I'm working.

NickOulton · 06-19-09

Quote:

Originally Posted by Kirmie

Right now we have a single, moving head per platter, I proposed that they use mulitple non-moving heads per platter. The rectangle would allow all spots to be accessed, but some spots would take longer to access than others do to the motion.

Lets say you have the digit 1,2,3,4,5 on the ground in a hallway and one person had to go to each number by going across the entire hall each time if they wanted to read it. Would it take the person longer to read these numbers in order 1,5,1,5,2,4,1,5 or would it take longer to read 2,4,2,4,3,2,3,4? The further those numbers are space out, the longer it takes to reach from one end to the other, similar to going from the outside to the inside part of a platter and vice versa. This could be reduced by having a nearby head overlap the reading area. In the same situation above, this would work like having a person who could read any number as long as there were exactly two other numbers between you. If this only happened on the right it would involve two people (heads) or three people for both sides. Then if you read the number 1 it would only take you two moves to read the number 5. This would not be possible on the outside of the platter since only the outside head would be able to reach it.

My proposal would have the number 1,2,3,4,5 placed on a merry-go-round with an arbitrary number of people, lets go with two for giggles. If the two were spaced apart so that if one was reading the number 1 the other would be reading the number 5. They would read the numbers by having the merry-go-round spin so that if a person read the number 1 then they would read the number 2 next. So lets do the math with the scenarios with a starting position of the original person at 3 and going left or lower first, always left for merry-go-round (second person is two numbers right or higher, third person two numbers left or lower). Also, only one number can be read per stop though the starting spot can be read.

Order of 1,5,1,5,2,4,1,5:
Back and forth 1 person: 2 + 4 + 4 + 4 + 3 + 4 + 5 + 4 = 30
Back and forth 2 person: 2 + 2 + 6 + 2 + 5 + 2 + 5 + 2 = 26
Back and forth 3 person: 0 + 4 + 4 + 4 + 1 + 2 + 1 + 1 = 17
Merry-go-Round 1 person: 2 + 1 + 4 + 1 + 3 + 3 + 3 + 1 = 18
Merry-go-Round 2 person: 2 + 1 + 1 + 1 + 1 + 3 + 3 + 1 = 13
Merry-go-Round 3 person: 0 + 1 + 1 + 1 + 1 + 2 + 1 + 1 = 08

Order of 2,4,2,4,3,2,3,4:
Back and forth 1 person: 1 + 4 + 4 + 4 + 3 + 1 + 3 + 1 = 21
Back and forth 2 person: 1 + 2 + 6 + 2 + 1 + 5 + 1 + 1 = 19
Back and forth 3 person: 1 + 2 + 2 + 2 + 1 + 1 + 1 + 1 = 11
Merry-go-Round 1 person: 1 + 3 + 2 + 3 + 1 + 1 + 4 + 4 = 19
Merry-go-Round 2 person: 1 + 3 + 2 + 3 + 1 + 1 + 1 + 2 = 14
Merry-go-Round 3 person: 1 + 1 + 1 + 1 + 1 + 1 + 1 + 2 = 09

In both situations the reading in a mery-go-round (spin) motion requires the least movement and thus the least seek time. Another way to put it would be that the maximum number of movements for the original head to reach a number in the back and forth method is eight and occurs when you need to read the number 1 or 5 twice in a row. The maximum number of movements needed to reach a number in the spin is 4 and this occurs whenever you have to read any number twice in a row. The only number with an equal or lower max than that in the back and forth method is reaching the number 3 which has a max of 4 as well. Either could be faster in a specific situation but the mery-go-round method will be the fastest on average.

Also, if somebody is good at math and not too lazy could they put that into a formula for me?

Edit: Sorry if some of that is confusing, wrote it out in chunks while I'm working.

superon · 06-19-09

Quote:

Originally Posted by Kirmie

Right now we have a single, moving head per platter, I proposed that they use mulitple non-moving heads per platter. The rectangle would allow all spots to be accessed, but some spots would take longer to access than others do to the motion.

Lets say you have the digit 1,2,3,4,5 on the ground in a hallway and one person had to go to each number by going across the entire hall each time if they wanted to read it. Would it take the person longer to read these numbers in order 1,5,1,5,2,4,1,5 or would it take longer to read 2,4,2,4,3,2,3,4? The further those numbers are space out, the longer it takes to reach from one end to the other, similar to going from the outside to the inside part of a platter and vice versa. This could be reduced by having a nearby head overlap the reading area. In the same situation above, this would work like having a person who could read any number as long as there were exactly two other numbers between you. If this only happened on the right it would involve two people (heads) or three people for both sides. Then if you read the number 1 it would only take you two moves to read the number 5. This would not be possible on the outside of the platter since only the outside head would be able to reach it.

My proposal would have the number 1,2,3,4,5 placed on a merry-go-round with an arbitrary number of people, lets go with two for giggles. If the two were spaced apart so that if one was reading the number 1 the other would be reading the number 5. They would read the numbers by having the merry-go-round spin so that if a person read the number 1 then they would read the number 2 next. So lets do the math with the scenarios with a starting position of the original person at 3 and going left or lower first, always left for merry-go-round (second person is two numbers right or higher, third person two numbers left or lower). Also, only one number can be read per stop though the starting spot can be read.

Order of 1,5,1,5,2,4,1,5:
Back and forth 1 person: 2 + 4 + 4 + 4 + 3 + 4 + 5 + 4 = 30
Back and forth 2 person: 2 + 2 + 6 + 2 + 5 + 2 + 5 + 2 = 26
Back and forth 3 person: 0 + 4 + 4 + 4 + 1 + 2 + 1 + 1 = 17
Merry-go-Round 1 person: 2 + 1 + 4 + 1 + 3 + 3 + 3 + 1 = 18
Merry-go-Round 2 person: 2 + 1 + 1 + 1 + 1 + 3 + 3 + 1 = 13
Merry-go-Round 3 person: 0 + 1 + 1 + 1 + 1 + 2 + 1 + 1 = 08

Order of 2,4,2,4,3,2,3,4:
Back and forth 1 person: 1 + 4 + 4 + 4 + 3 + 1 + 3 + 1 = 21
Back and forth 2 person: 1 + 2 + 6 + 2 + 1 + 5 + 1 + 1 = 19
Back and forth 3 person: 1 + 2 + 2 + 2 + 1 + 1 + 1 + 1 = 11
Merry-go-Round 1 person: 1 + 3 + 2 + 3 + 1 + 1 + 4 + 4 = 19
Merry-go-Round 2 person: 1 + 3 + 2 + 3 + 1 + 1 + 1 + 2 = 14
Merry-go-Round 3 person: 1 + 1 + 1 + 1 + 1 + 1 + 1 + 2 = 09

In both situations the reading in a mery-go-round (spin) motion requires the least movement and thus the least seek time. Another way to put it would be that the maximum number of movements for the original head to reach a number in the back and forth method is eight and occurs when you need to read the number 1 or 5 twice in a row. The maximum number of movements needed to reach a number in the spin is 4 and this occurs whenever you have to read any number twice in a row. The only number with an equal or lower max than that in the back and forth method is reaching the number 3 which has a max of 4 as well. Either could be faster in a specific situation but the mery-go-round method will be the fastest on average.

Also, if somebody is good at math and not too lazy could they put that into a formula for me?

Edit: Sorry if some of that is confusing, wrote it out in chunks while I'm working.

It is my understanding that with this technology there is basically one head per sector -- so the point you raised wouldn't apply for this situation. Each head only has to worry about one sector.

[Teh Root] · 06-19-09

Peizoelectricity is not that hard of a concept. Ever have a Casio Digital Watch? There's a peice or quartz, technically silicon dioxide that bends to an electric current and field when it returns to its original shape is emits a small amount of electric field charge.

HellKat · 06-19-09

I actually made a post about piezoelectric crystals the other day. It might help some to understand it better.

http://www.overclock.net/amd-cpus/52...rclocking.html

		Overclock.net - Overclocking.net > Industry News > Hardware News
[Dailytech] Hard Rectangular Drive May Replace Current Spinning Discs

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)