Cost of DVD as data storage versus HDD (UK)

[half_pint]

half_pint wrote: [snip]
Few failures initially, reaching a maximum and then falling back to
zero once all have failed.

No its bath shaped you troll.

Ahhhh, I`d presume from your statement that you consider yourself more
qualified than me in the field of statistics. Please cite your
qualifications, and name the type of statistics used to describe the
lifetimes of components. No hints from anyoen else please, lets see if
half_wit has any clue at all about this.

You're an idiot if you think qualifications in statistics qualify you
to define the failure curve of electro-mechanical devices (apparently
without data or any understanding of engineering). Odds-on
electro-mechanical devices will fail within the first few months
(manufacturing defects) or after a relatively long period of time (a
few years, when stuff wears out). In simple terms, if it lasts six
months there is a very good chance it will last three years.

Not that your reply was in my direction - but the orignial point was
the belief this continually used 3gig drive would last 100 years.

Using stats or mechanics as your argument, the chances of that
happening are Zero. Or do you suggest otherwise?

Well its irrelevant anyway, as the drive will and indeed has "lasted
forever"
forever being "as long as I wanted it to" 2 and 3 gig drives are virtually
worthless these days, try buying one on Ebay, you will spent more on the
postage than you will on the drive, a bit like the 4 16meg simms I pulled
from
my machine, I cant be arsed to sell them because the £1 or £2 I would get
is just not worth the effort, it would involve a hour or so of effort and I
don't
work for below the minimum wage.

 

[Alex Fraser]

[snip]

Ok, so my drive spins at 7200 rpm, the same as yours. How big are your
platters? Lets be VERY generous, and say the full 5 gig capacity of your
drive is on a single platter. My smallest drive is 180 gigs - lets say
there are 3 platters there. My platters therefore hold 60 gigs each,
despite being the same physical size as your platters. Therefore the
data density on my platters is 12 times greater than on yours.

Therefore, for each revolution of the platter, my drive can read 12 times
more data. That`s 12 times the amount of data in the same amount of
time, making the data transfer rate 12 times greater.

Unless I'm mistaken, you're right about the relative data density (assuming
one platter for the 5GB drive), but that number is the product of two
others: relative bits per unit length (along the track), and a factor you
ignored, relative tracks per unit length (radially). The latter affects
capacity, while the former (along with spindle speed) affects transfer rate.

Alex

 

[Peter]

I have never heard about it either, but I usually search before posting. I
heard about Weibull distribution before, so here is the result:
http://www.weibull.com/hotwire/issue21/hottopics21.htm

Interestingly enough very small times and very long times are not plotted. I
guess you can figure out the reasons.

 

[J. Clarke]

Not what I said "Your drive spins at either 5400 or 7200, the *same* as
mine"

Miine spins at 5400 (5401 I think), which statisfies the the 5400 or 7200
clause.




However what you fail to realise is that data just behind the read head
requires one revolution for it to be read (unless it has multipule read
heads).

That's called "latency", and it's only one factor in drive performance.

 

[J. Clarke]

I`ve NEVER heard of it being described as a bathtub shape. How long ago
was
your education, out of interest? What type of statistics would you use to
describe the failure rates? Everytime I`ve ever seen the relevant type of
stats being used, it`s always been a bell shape.

The bell curve is called the "normal distribution" of data about a mean.
Failures in electronics exhibit something more akin to a skewed bimodal
distribution, with a peak early, a peak late, and a fairly wide separation
between them, which does if you look at it the right way kind of resemble a
bathtub.

Adjusting the parameters
could just about come up with a very weird bath-tub shape, but it`s
certainly nothing like a bath-tub as I know it Pretty much all the
time
I`ve ever seen it used, it`s given a nice bell shape. Sometimes short and
fat, sometimes tall and thin, but always a recognisable bell. ~66%
withing 1SD of the average failure time, ~66% of the remained between 1
and 2 SD of
the average and so on. This inevitably leads to a nice bell shape -
exponential decay and all that.

Well, you might have seen that in a statistics class, but statistics classes
deal in methods of calculation, not in actual performance of real-world
devices. If you're referring to the published MTBF numbers, they're not
based on service life but on probability of failure _during_ the service
life, which is why the numbers are so large, with the assumption being that
the device will be replaced due to obsolescence long before wear becomes an
issue. Their purpose is to allow an organization to estimate maintenance
requirements and maintain adequate quantities of spares when they have a
large number of devices of like kind in service.

 

[lysander]

I`ve NEVER heard of it being described as a bathtub shape. How long ago was
your education, out of interest? What type of statistics would you use to
describe the failure rates?

As I said I don't do statistics, only electronics. A description of
the bathtub curve is here:

http://www.itl.nist.gov/div898/handbook/apr/section1/apr124.htm

 

[half_pint]

Note to self #1: Read entire thread before posting.

Note to self #2: Don't post drunk.

Note to self #3: Remember #2.

I was refering to Simon Finnigan of course, not you, if that is any help.

 

[Rob Morley]

However what you fail to realise is that data just behind the read head
requires one revolution for it to be read (unless it has multipule read
heads).
So my 5400 is only about 33% slower than a 'modern' 7200 drive.

A fact which even the most persistant of trolls cannot deny.

That's the maximum time that the head will take to start reading the
data after it has been positioned over the track. You have not taken
account of the time for the head to position over the track or the speed
of data transfer once it has started reading.

 

[guv]

Not what I said "Your drive spins at either 5400 or 7200, the *same* as
mine"

Miine spins at 5400 (5401 I think), which statisfies the the 5400 or 7200
clause.

And which by definition, you would agree that a modern 7200rpm drive
will out perform a 5400rpm drive?

However what you fail to realise is that data just behind the read head
requires one revolution for it to be read (unless it has multipule read
heads).
So my 5400 is only about 33% slower than a 'modern' 7200 drive.

Seeing as the density of data is far more in a modern multi-platter
drive, the amount of data read in one revolution will be so much more,
the spin speed of your ancient drive becomes irelevant when trying to
suggest your drive reads and writes the same volume and speed as a
modern drive.

A fact which even the most persistant of trolls cannot deny.

Since I'm not a troll and you refuse to post as requested the drive
model number to PROVE what you are suggesting is nonsense, perhaps you
would like to prove this point incorrect and do so? Or will you just
continue to ignore this as you have previously?

 

[Alex Fraser]

[snip]

Seeing as the density of data is far more in a modern multi-platter
drive, the amount of data read in one revolution will be so much more,
[...]

If you look at a family of drives where the only difference is the number of
platters, or rather, heads, you will find no difference in performance. I
assume this is because only one head is used at a time, even though all
heads move in unison.

Alex

 

[Guest]

Since I'm not a troll and you refuse to post as requested the drive
model number to PROVE what you are suggesting is nonsense, perhaps you
would like to prove this point incorrect and do so? Or will you just
continue to ignore this as you have previously?

half_pint (and his previous incarnations) is deranged. He refuses to
comprehend or accept any sense people talk to him.

There is no point trying to enlighten him, just kill file him, life is too
short.

 

[Neil Maxwell]

I`ve NEVER heard of it being described as a bathtub shape. How long ago was
your education, out of interest? What type of statistics would you use to
describe the failure rates? Everytime I`ve ever seen the relevant type of
stats being used, it`s always been a bell shape. Adjusting the parameters
could just about come up with a very weird bath-tub shape, but it`s
certainly nothing like a bath-tub as I know it

He's obviously talking about the failure rates for the IBM Deskstar
drives...

 

[half_pint]

That's the maximum time that the head will take to start reading the
data after it has been positioned over the track. You have not taken
account of the time for the head to position over the track or the speed
of data transfer once it has started reading.

WEll they are not relevant to my point so obviously not.
But thanks for verifying that i am correct anyway.

 

[half_pint]

And which by definition, you would agree that a modern 7200rpm drive
will out perform a 5400rpm drive?

At what?

Seeing as the density of data is far more in a modern multi-platter
drive, the amount of data read in one revolution will be so much more,
the spin speed of your ancient drive becomes irelevant when trying to
suggest your drive reads and writes the same volume and speed as a
modern drive.

SO you agree that "my 5400 is only about 33% slower than a 'modern' 7200
drive."

Since I'm not a troll and you refuse to post as requested the drive
model number to PROVE what you are suggesting is nonsense, perhaps you
would like to prove this point incorrect and do so? Or will you just
continue to ignore this as you have previously?

I have a Samsung and noisy Western Digital I an not going to dismantal
may computer, post your drive model number first.

 

[half_pint]

[snip]

Seeing as the density of data is far more in a modern multi-platter
drive, the amount of data read in one revolution will be so much more,
[...]

If you look at a family of drives where the only difference is the number of
platters, or rather, heads, you will find no difference in performance. I
assume this is because only one head is used at a time, even though all
heads move in unison.

Good pooint Alex, I was just about to say that!!!

 

[half_pint]

half_pint (and his previous incarnations) is deranged. He refuses to
comprehend or accept any sense people talk to him.

There is no point trying to enlighten him, just kill file him, life is too
short.

Your life will be pretty short with that kind of attitude to wisdom.

Never ignore wisdom.

 

[Peter]

SO you agree that "my 5400 is only about 33% slower than a 'modern' 7200

drive."

You are absolutely right! "5400 drive is about 33% slower than a 7200 drive"
is TRUE!
With accuracy of more than 99.7%.
Let me repeat: "5400 drive disk rotation speed is about 33% slower than disk
rotation in a 7200 drive"

You can also argue that 7200 drives are not more than about 25% better than
5400 drives. ;-)

 

[J. Clarke]

Your life will be pretty short with that kind of attitude to wisdom.

Never ignore wisdom.

If you had any to offer then you might have a point. But you don't. You
take one of at least four different factors that contribute to the
performance of storage devices, and on the basis of that one factor,
without any test results or calculations or literature citations or
anything else to support your argument conclude that you have proven
something about disk performance.

That is not wisdom of any kind.

 

[Rob Morley]

WEll they are not relevant to my point so obviously not.
But thanks for verifying that i am correct anyway.

You originally wrote "I dont think new harddrives will be any faster
than mine ( speeds are basically the same 5400 or 7200 ) so I cant see
them writing any faster". So you are either stupid, trolling or
deranged.

 

[guv]

At what?

Do you know what a hard drive is for and what the definition
"performance" means?

SO you agree that "my 5400 is only about 33% slower than a 'modern' 7200
drive."

Your drive might spin at 33 less speed, but that has no relevance in
your claim your ancient 3 gig drives performance is equal to modern
drives. Are you trying to move the goalposts? Every modern 5400rpm
drive will outperform your ancient drive, even though the spin speed
is the same.

I have a Samsung and noisy Western Digital I an not going to dismantal
may computer, post your drive model number first.

I have several machines and even more hard drives. I dont need to
dismantle any machine to know what model number the drives are. Its
clearly shown in control panel, system devices. Perhaps if your
knowledge of PCs was better, you would know a few more facts than the
lack of knowledge you persist in displaying.

I just had a quick look online at specs of drive of slightly larger
and newer drives than your own (A massive 6 gig!). It says transfer
speeds are up to 5 meg per second. Now compare that with modern up to
100meg per second drives. Do you notice any difference in those
figures?

It would appear your argument is based on a flawed premise that
technology has not advanced. If you want to believe that, then be my
guest.