Physical Hard Drive - PHD - a new technology!

[SecurityTheatre]

One square inch is 645,160,000 square microns. This is 4.16231426+17

Hah!

I just realized where your 41 trillion came from.

You SQUARED 645,160,000.

A square inch is a measurement of area. It doesn't meant that you square the value.

FYI, there are 645.16 million microns in a square inch. (25,400 microns per linear inch)

http://lmgtfy.com/?q=1+sq+inch+in+sq+microns

And my nickname is cause I love math, math is awesome, and Einstein loved math.

Indeed it is, my friend.

 

[SecurityTheatre]

Wait... your patent isn't based on the calculations you presented in that previous post is it?



.

 

[Soundmanred]

magnetic "Pits"

Speechless.

One square inch is 645,160,000 square microns. This is 4.16231426+17

Somehow even more speechless.

 

[TheEinstein]

http://solidsmack.com/fabrication/the-tantalizing-tantillus-3d-printer-10-micron-printing/

Yeah a good review could earn, if you were a friend, a lot but your blacklisted from my friends list and thus sad

You keep looking at possible negatives, and using your biases wrongly.

Oh well I am getting noticed elsewhere.

One day soon someone will want to make millions of dollars by making me many millions of dollars.

 

[TheEinstein]

Hah!

I just realized where your 41 trillion came from.

You SQUARED 645,160,000.

A square inch is a measurement of area. It doesn't meant that you square the value.

FYI, there are 645.16 million microns in a square inch. (25,400 microns per linear inch)

I squared a square and this is due to low sleep, 3d and 2d confusions.

Microns is attainable to width and height of 100 with ease however.

Even 25 microns.

Now then off to get water on my face

However quick note... not possible to use 3d, and 2d conversion of millions of inches would be bad.

 

[SecurityTheatre]

I squared a square and this is due to low sleep, 3d and 2d confusions.

Microns is attainable to width and height of 100 with ease however.

Even 25 microns.

Now then off to get water on my face

However quick note... not possible to use 3d, and 2d conversion of millions of inches would be bad.

Listen, mistakes happen, but I would hope you had done these calculations before. My reply wouldn't have been NEARLY as snarky without your snide remarks about "keeping up" and the limey shit at the end. Putting that aside.... I frankly don't care. Lets actually talk about the technology some more.


I should be clear to say that a good review of an idea isn't from a bunch of "yes men". You need to answer the hard questions. I'm a skeptic by nature, but that doesn't make my opinion invalid. I have grilled several other people about their patent or dissertation in the past and they came back with solid responses to the criticism.

I expect the same from you, unless you don't believe you're correct anymore.


Assuming you can print 10-micron linear resolution (100 square microns in 2d) as this printer claims (actually not, that was 10 microns in depth, not linear, but meh, whatever), show me the density calculation that you claim to be able to beat magnetic media. Let's just assume we have a magical printer that can produce 10 micron linear resolution AND do it with.... say 16 shades of each color without halftoning. Such a thing doesn't exist, but if it did.... You have 6.4 million dots per inch with a storage density of 12 bits per dot. This gives you a very impractical and error prone 76.8 million bits per square inch.

Lets assume you can use "height" as a factor on your 3d printer. This is difficult to achieve, but we're making wild assumptions. Give yourself 16 levels of height. Now you have another 4 bits at 16 bits per dot.

Given these wildly generous assumptions, you can store 100 million bits per inch.

Sort of getting there. But even assuming 16.7 million colors of depth (which isn't possible to measure accurately), you get around 200 million bits per inch.

We're using exotic imaginary printers, with exotic, imaginary scanners that include exotic, imaginary height detectors of sub-micron precision. And we're an order of magnitude too low still.

Adding your "structure" claim, that's somewhere around 0.5 bit more depth, which is really insignificant.

*shrug*

I don't see it.

All I've ever said is PROVE IT.

 

[TheEinstein]

I will do a comprehensive write up on my 3 tier system. including math proofs and walkthroughs. give me a few days.

 

[velis]

Security, I think you're talking to a 10-year old or at most an early puberty one. I used to have such warped understanding of physics at about that age and I'm not even a decent engineer. In my defense, it took me about a week to research such stuff far enough for me to see it was not feasible. And I had no internet available at the time.

I think his "shapes" are supposed to be sub-pixel ones - there are many possible shapes enhancing each bit with many bits of additional information.
His colors stem from looking at a photo and marvelling the "full color" his biological eyes see. Then he read about light spectrum being much wider than that and he enhanced his single bit further by coloring his bit-sized shapes.

All the margins and other encription / compression helpers are his attempts at actually solving the original idea of a shaped and colored bit (which is thus no longer a bit, but at least several bits of information which even you acknowledge). In a while he will realize that in order to have these colored / shaped bits, he could much better use the space as an ordinary binary matrix, but in order to do that, he'd have to go to levels sub-atomic. Last I checked, plastic substrate was still plastic. Those molecules are not prefixed "poly" for nothing.

That said, thanks for bothering with him, I learned a lot about printer physics from your answers

 

[bryanl]

What are some of your other inventions?

 

[alkemyst]

What are some of your other inventions?

He won't answer in detail, but just laugh you away.

I passed by his ideas to my brother who is a storage engineer, in deeply into big data, taking Harvard grad classes in it as well as in the top 1% on Kaggle.

This guy claims to love math and problem solving yet is no where anywhere on the net doing that.

He claims to have patents, but won't share.

He keeps making substantial mistakes that he passes off as trivial (like HDD's using pits).

My thoughts are he believes some of the diagram reduction methods will work on computer data. It's easy to take a map of 50,000 end points and reduce them to a few hundred, but to reverse that you don't get back all 50,000 end points. You may be lucky to get back 80%. This wouldn't work for computer data.

The new color items he is adding would require additional inventions to do the reading and writing. As based on what he is explaining nothing exists that can do this reasonably in cost and/or size.

Another thing that he'd have to reinvent is making compression not expensive/slow to read from disk.

The final problem is anyone asking for more details is laughed at and told they basically lack even rudimentary understanding of this topic.

 

[alkemyst]

Here are some more thoughts: they can read magnetic disk platters at 15,000 RPM's/sec. Sure it might not be as "dense" as many other methods out there, but it can be read, reliably, and it doesn't decay.

I am not aware of colors that don't decay, which could be a huge issue. This has been discussed above.

Think of air spectrum, we have theoretically billions of frequencies. Yet in practice we can't use them all, because you can't assign something to 720.1254mhz. And if you
did, the receiver technology required to discriminate at that level would be outrageously expensive. So not only do you have to assign large amounts of bandwidth to
frequencies, but you must ensure guard-bands as a safety net between frequencies. The same goes for colors. To what degree does he think he can write color that can be read?

How fast can something like color, which requires optical imaging, can be read?

Disks aren't that expensive, and they are fast as hell right now, SSD's are the future. I don't care if his thing can do 5x the size of modern disks, if its slow, and
expensive, who the hell would want it?

 

[Sho'Nuff]

1. Optical and magneto optical recording media (with wavelength variation) is old. At least 10-15 years old.

Source- I was a patent examiner at the USPTO and examined magnetic recording media applications.

2. I hope your patent application does a better job of describing your application than the OP, else I fear there is a very good chance a US examiner will reject it under 35 USC 112(1) as failing to meet the written description and/or enablement requirements of US patent law.

Source - I am a patent attorney specializing in patent prosecution and licensing.

For the record, nothing in this post is legal advice. I am not your attorney.

 

[Sho'Nuff]

How fast can something like color, which requires optical imaging, can be read?

Disks aren't that expensive, and they are fast as hell right now, SSD's are the future. I don't care if his thing can do 5x the size of modern disks, if its slow, and
expensive, who the hell would want it?

nano dot (aka quantum dot) printer technology could put the color down (if I understand the invention correctly). As for read speed, we already have technology that can quickly read optical media(recognizing that optical media is substantially different from magnetic media). The principle here looks like the op is trying to merge recording tech with some form of optical spectroscopy. UV-VIS might or might not have the resolution required, but it is a fast measurement technique and might be adapted for the purpose articulated by the op. Not without a significant amount of effort, though.

Fwiw, the cost concerns you raise are basically the same concerns that the community raised 15 years ago when seagate devised how to grow perpendicularly oriented columnar magnetic crystals (a technology which is still primarily responsible for the high capacity magnetic drives we have today). When that tech was developed there were all sorts of resolution issues with the read/write head(s), which were eventually addressed in a cost effective manner.

According to my IT buddies, Tape is still the most cost effective and reliable solution for system wide backups. So even if the tech proposed in the op is slow, it might have a market if it can beat type in the $/Gb and reliability categories.

 

[Sho'Nuff]

I'll just leave these here:

http://onlinelibrary.wiley.com/doi/10.1002/pssc.201084049/abstract

http://mazur.harvard.edu/publications/Pub_95.pdf

http://arxiv.org/pdf/cond-mat/0602531.pdf


http://m.authorstream.com/presentation/prasantakumarbeh-1793286-3d-optical-data-storage/

 

[SecurityTheatre]

I'm going to respond in the spirit of this thread... take a deep breath.

Here are some more thoughts: they can read magnetic disk platters at 15,000 RPM's/sec.

Can you describe those units? I would like to buy some. Please write out what RPM means and restructure your sentence.

Sure it might not be as "dense" as many other methods out there, but it can be read, reliably, and it doesn't decay.

Pronoun fail. But regardless of which noun your pronoun is referencing, I think you're overstating.

I am not aware of colors that don't decay, which could be a huge issue. This has been discussed above.

Fair criticism. Using a 16-color scheme with archival dyes, however, could theoretically last for hundreds of years (longer than modern drives). But obviously, it wouldn't work with a high color scheme.

Think of air spectrum, we have theoretically billions of frequencies. Yet in practice we can't use them all, because you can't assign something to 720.1254mhz. And if you
did, the receiver technology required to discriminate at that level would be outrageously expensive. So not only do you have to assign large amounts of bandwidth to
frequencies, but you must ensure guard-bands as a safety net between frequencies.

You are sort of accurate, but for the wrong reason. We absolutely have the technology to build a very sharp bandpass filter. The real problem is that within a very fine channel, you have a very limited amount of bandwidth. You don't have room for FM modulation, or even phase information, so modulations like QAM and QPCM don't work as well, if at all.

There are a number of theorems about this, including the Shannon-Hartley Theorem based on some of the old Nyquist calculations.

http://en.wikipedia.org/wiki/Hartley's_law#Hartley.27s_law

In short, the width of the channel is proportional to the amount of data that can travel through it. In practice, modern encoding schemes can stretch this a little bit (maybe even double or more), but this is not unlimited. A 40Mhz channel will never give you approximately 40Mbps of bandwidth with simple encoding. We see this in practice. 2.4Ghz technologies with 20Mhz channels struggle to break 20Mbps even with QAM. 802.11ac uses some more dense encoding (256-QAM) and manages to get a useful maybe 40-50Mbps out of each 20Mhz channel at the cost of much higher noise sensitivity.

How fast can something like color, which requires optical imaging, can be read?

I am going to disagree here. There is no fundamental limit on the rate at which color can be read, beyond the round-trip time of the photon to/from the detector, which should be nanoseconds or even picoseconds on the scale we're talking about, as well as the time it takes for the photoreceptor to charge, which should also be nanoseconds. I'd bet, theoretically, that optical read technologies could be built to read large chunks at a time, rather than having to use racetrack formats on magnetic disks. I don't think read speed is an issue here. In fact, it's one of the few non-issues here.

Disks aren't that expensive, and they are fast as hell right now, SSD's are the future. I don't care if his thing can do 5x the size of modern disks, if its slow, and
expensive, who the hell would want it?

If it were possible to make as cheap as he claims AND as dense as he claims, it would be of marginal utility for some niche markets, especially if it could be "printed" on-site.

That said, it's both impractical and niche and is unlikely to do either dense storage, or cheap storage.

 

[jagec]

See, I thought this thread was going to be a joke about PhD students.

This guy sounds like a born architect: propose wildly fanciful ideas that are beyond everything that exists...but for the pesky problem that some poor, underappreciated engineer actually has to figure out how to BUILD them.

 

[alkemyst]

I'm going to respond in the spirit of this thread... take a deep breath.



Can you describe those units? I would like to buy some. Please write out what RPM means and restructure your sentence.




Pronoun fail. But regardless of which noun your pronoun is referencing, I think you're overstating.




Fair criticism. Using a 16-color scheme with archival dyes, however, could theoretically last for hundreds of years (longer than modern drives). But obviously, it wouldn't work with a high color scheme.



You are sort of accurate, but for the wrong reason. We absolutely have the technology to build a very sharp bandpass filter. The real problem is that within a very fine channel, you have a very limited amount of bandwidth. You don't have room for FM modulation, or even phase information, so modulations like QAM and QPCM don't work as well, if at all.

There are a number of theorems about this, including the Shannon-Hartley Theorem based on some of the old Nyquist calculations.

http://en.wikipedia.org/wiki/Hartley's_law#Hartley.27s_law

In short, the width of the channel is proportional to the amount of data that can travel through it. In practice, modern encoding schemes can stretch this a little bit (maybe even double or more), but this is not unlimited. A 40Mhz channel will never give you approximately 40Mbps of bandwidth with simple encoding. We see this in practice. 2.4Ghz technologies with 20Mhz channels struggle to break 20Mbps even with QAM. 802.11ac uses some more dense encoding (256-QAM) and manages to get a useful maybe 40-50Mbps out of each 20Mhz channel at the cost of much higher noise sensitivity.




I am going to disagree here. There is no fundamental limit on the rate at which color can be read, beyond the round-trip time of the photon to/from the detector, which should be nanoseconds or even picoseconds on the scale we're talking about, as well as the time it takes for the photoreceptor to charge, which should also be nanoseconds. I'd bet, theoretically, that optical read technologies could be built to read large chunks at a time, rather than having to use racetrack formats on magnetic disks. I don't think read speed is an issue here. In fact, it's one of the few non-issues here.




If it were possible to make as cheap as he claims AND as dense as he claims, it would be of marginal utility for some niche markets, especially if it could be "printed" on-site.

That said, it's both impractical and niche and is unlikely to do either dense storage, or cheap storage.

I am not the expert here, however; if you want to PM me your trivial stuff I will put in in the hands of those that do this.

Most of what you have suggested above can't fit in a compact form factor.

Read speed is a factor of not only the sensor, but getting it to the device.

Bandwidth filter <> FCC guidelines and still expensive.

Your 802.11ac claims are way behind what is already out there in labs.

It was just RPM...not RPM/sec

gg though and just messed up this thread.

Basically following your path, you just used GoogleFu.

Got to add the Grammar Nazi post in HT as well...not sure where my reference to "IT" was anything other than the drive. but obviously you had more English classes <-- please don't take apart the bold text, it's like a pun to you.
Gratz, now take a breath.

 

[SecurityTheatre]

Your 802.11ac claims are way behind what is already out there in labs.

Which claims? AC very seldom uses a 20Mhz channel (it can do up to 160Mhz).

It was just RPM...not RPM/sec

:whiste: Yeah, I know. I was just teasing.

Basically following your path, you just used GoogleFu.

That and a MSc Computer Engineering. And several graduate seminars on wireless networking. And participating in a couple of IEEE working groups on wireless security back in the late 1990s. But yeah, mostly Google.

please don't take apart the bold text, it's like a pun to you.
Gratz, now take a breath.

:hmm:

()