If admitting his mistakes would open him to a lawsuit and if his goal was to preserve the good name (?) of his company, the only winning move was not to play. The fact of giving this interview at all would have brought him into trouble from either or both sides. Which once more raises the suspicion that this guy doesn't understand what he's dealing with.
10m in diameter should read 20 um (that's micrometer) in diameter. I made a typo in the number, and the 'micron' sign was eaten on submission. I know I should have used that preview button...
This is a hoax, or at least it contains several fatal mistakes. Just do the math: 1E-1 m / 8E+10 bit = about 1E-12 m / bit Taken into account that the spacing between atoms in solid iron is 2.9E-10 m, this is blatantly impossible.
Moreover, if I remember correctly, the theoretical lower limit to the size of a ferromagnetic domain is about 100 atoms. On the basis of this, I estimate that it is not possible to write more than 8MB on a very thin wire 10cm in length. This is not very impressive to me, but at least it comes close to the '10 million divisions or cells' mentioned in the article.
Under ideal cicrcumstances, it might be possibly crank 10GB onto a 10cm wire 10m in diameter if one treats it as a cylinder and writes different bits around its circumference. However, this would require truely revolutionary technology, and render the data very vulnerable to torsion forces. Moreover, the article suggests that it is not the case: '"bamboo"-type structure of domains'.
Furthermore, it is full of fancy-sounding word combinations I can't make sense of, and contains other shameful mistakes, like the "magnetical CD", and the bits/bytes confusion. Taken together, it drives my bogometer into the red...
Slashdot Mirror
If admitting his mistakes would open him to a lawsuit and if his goal was to preserve the good name (?) of his company, the only winning move was not to play. The fact of giving this interview at all would have brought him into trouble from either or both sides. Which once more raises the suspicion that this guy doesn't understand what he's dealing with.
10m in diameter should read 20 um (that's micrometer) in diameter. I made a typo in the number, and the 'micron' sign was eaten on submission. I know I should have used that preview button...
This is a hoax, or at least it contains several fatal mistakes. Just do the math:
1E-1 m / 8E+10 bit = about 1E-12 m / bit
Taken into account that the spacing between atoms in solid iron is 2.9E-10 m, this is blatantly impossible.
Moreover, if I remember correctly, the theoretical lower limit to the size of a ferromagnetic domain is about 100 atoms. On the basis of this, I estimate that it is not possible to write more than 8MB on a very thin wire 10cm in length. This is not very impressive to me, but at least it comes close to the '10 million divisions or cells' mentioned in the article.
Under ideal cicrcumstances, it might be possibly crank 10GB onto a 10cm wire 10m in diameter if one treats it as a cylinder and writes different bits around its circumference. However, this would require truely revolutionary technology, and render the data very vulnerable to torsion forces. Moreover, the article suggests that it is not the case: '"bamboo"-type structure of domains'.
Furthermore, it is full of fancy-sounding word combinations I can't make sense of, and contains other shameful mistakes, like the "magnetical CD", and the bits/bytes confusion. Taken together, it drives my bogometer into the red...