Terminology Discrepancy (split from Breakthrough in nanotechnology)

[Enthalpy]

And you believed the tale of nanotube processor?

I regret that the MIT technology review prints such titles.

[swansont]

And you believed the tale of nanotube processor?

I regret that the MIT technology review prints such titles.

 

The following is a rhetorical question; the existence of a carbon nanotube processor is tangential to the discussion.

 

Is a Nature paper credible enough for you?

http://www.nature.com/nature/journal/v501/n7468/full/nature12502.html

[Enthalpy]

300 transistors don't make a processor. Not even a 4004.

 

I disapprove the hype in science papers, be it in Nature or elsewhere.

[swansont]

300 transistors don't make a processor. Not even a 4004.

 

I disapprove the hype in science papers, be it in Nature or elsewhere.

What is a device that processes data to perform a task, if not a processor?

 

The CNT computer runs an operating system that is capable of multitasking: as a demonstration, we perform counting and integer-sorting simultaneously. In addition, we implement 20 different instructions from the commercial MIPS instruction set to demonstrate the generality of our CNT computer.

[Enthalpy]

Obviously you've been fooled by the paper. That is why I dislike hype in science papers.

 

One bit of a full adder takes ~70 transistors, an Alu some more. With 300 transistors they can have made a 4-bit Alu.

 

A processor still demands a sequencer, instruction decoding, registers, memory addressing... which take very much more than the Alu. With 300 transistors they did not make a processor. They miss a factor of 15. Impossible. Sorry for that.

 

The logic may be that this team achieved to integrate 300 working nanotube transistors, which is a very nice achievement. Instead of publishing "we integrated 300 nanotube transistors", they chose to make something like a 4-bit Alu with them, which is more concrete and makes a milestone. Possibly, they added a sequencer and so on in silicon around - or they didn't. All right.

 

But then, someone (One author? The scientific journal? The general press?) described the 300 transistor chip (maybe a 4-bit Alu but nothing more) as a processor, and this is dishonest. This fools readers with the best scientific background who are not specialists of that area. This should not happen.

[Lucius E.E]

Obviously you've been fooled by the paper. That is why I dislike hype in science papers.

 

One bit of a full adder takes ~70 transistors, an Alu some more. With 300 transistors they can have made a 4-bit Alu.

 

A processor still demands a sequencer, instruction decoding, registers, memory addressing... which take very much more than the Alu. With 300 transistors they did not make a processor. They miss a factor of 15. Impossible. Sorry for that.

 

The logic may be that this team achieved to integrate 300 working nanotube transistors, which is a very nice achievement. Instead of publishing "we integrated 300 nanotube transistors", they chose to make something like a 4-bit Alu with them, which is more concrete and makes a milestone. Possibly, they added a sequencer and so on in silicon around - or they didn't. All right.

 

But then, someone (One author? The scientific journal? The general press?) described the 300 transistor chip (maybe a 4-bit Alu but nothing more) as a processor, and this is dishonest. This fools readers with the best scientific background who are not specialists of that area. This should not happen.

 

I applaud your critique of this, unfortunately ostensible article.

 

The header should read as follows:

"178 Nanotube transistors successfully integrated on a single chip wafer"

 

However I do find this to be a magnificent proof-of-concept. Ultimately I feel as though carbon nanotube processors are possible, and in the not so distant future we will see nano computers, and not long after quantum computers. The exponential rise in technology isn't slowing down; it is accelerating every year. Carbon nanotube processors are right around the corner; considering we can already compute basic numerical calculations.

Edited August 28, 2014 by Lucius E.E

[Phi for All]

!

Moderator Note

These posts were split from a different discussion. Rather than being related to the original topic, they turned out to be a discrepancy over some terms used. The original conversation can be found here.