« previous   misc   next »

Physicists Create a Working Transistor From a Single Atom


By Dan8267   Follow   Mon, 20 Feb 2012, 10:13am PST   3,842 views   30 comments
Watch (0)   Share   Quote   Permalink   Like   Dislike  

Australian and American physicists have built a working transistor from a single phosphorus atom embedded in a silicon crystal.

The group of physicists, based at the University of New South Wales and Purdue University, said they had laid the groundwork for a futuristic quantum computer that might one day function in a nanoscale world and would be orders of magnitude smaller and quicker than today’s silicon-based machines.

http://www.nytimes.com/2012/02/20/science/physicists-create-a-working-transistor-from-a-single-atom.html

So cool.

Comments 1-30 of 30     Last »

TPB   Mon, 20 Feb 2012, 10:51am PST   Share   Quote   Permalink   Like (1)   Dislike (2)     Comment 1

Jan Hendrik Schön had credible research and groundwork for a futuristic bio transistor. The problem was when asked for proof, he produced duplicate graph showing voltage drainage.

If nano technology processor ever becomes common, though the chip may be more powerful than anything we ever had in a typical OS computer to date. The devices they put them probably wont be as robust as we use them today. Most likely they will power things like a thick sheet of binder plastic, that can store and power calculations on a few embedded spreadsheets. For what that sheet of plastic is used for, that's all it will ever need. Or to power digital business cards, or post cards, or even postlet adverts.
And though the nano processor and memory embedded in the plastic sheet is thousands of times more powerful than the chips in our current notebooks, or desktops, for its size. It really wont have the same power as the desktops and laptops.
It will be more like an ant is a hundred times stronger than a human. But we can easily step on them.

Patrick   Mon, 20 Feb 2012, 11:37am PST   Share   Quote   Permalink   Like   Dislike (1)     Comment 2

Would definitely help Patrick.net site performance. A lot of the time is going to string manipulation, which cpu intensive.

But maybe wouldn't help network latency.

Dan8267   Mon, 20 Feb 2012, 11:48pm PST   Share   Quote   Permalink   Like   Dislike     Comment 3

Patrick says

But maybe wouldn't help network latency.

That's what quantum entanglement if for. I don't remember where the article I read was, but there was a experiment which claimed to transmit a Beethoven symphony faster than light using quantum entanglement. The idea was that the spin of two electrons are always opposite according to some conservation law. You change the spin of one and the other changed immediately. What Einstein called spookiness at a distance.

TPB   Tue, 21 Feb 2012, 7:52am PST   Share   Quote   Permalink   Like (1)   Dislike (1)     Comment 4

Patrick says

A lot of the time is going to string manipulation, which cpu intensive.

What are you doing, with strings, that's so cpu intensive?

Dan8267   Tue, 21 Feb 2012, 8:54am PST   Share   Quote   Permalink   Like   Dislike (2)     Comment 5

String manipulation and lexicographical processing is typically CPU expensive compared to common database queries, especially doing things like searching for substrings.

Patrick   Tue, 21 Feb 2012, 8:59am PST   Share   Quote   Permalink   Like (1)   Dislike (1)     Comment 6

TPB says

What are you doing, with strings, that's so cpu intensive?

Just generating the forum pages, like the one you're viewing right now. It's a whole bunch of db access and string formatting.

Can't cache them, because they have to be up to date with latest comment, customized for your user ID, etc.

TPB   Tue, 21 Feb 2012, 9:15am PST   Share   Quote   Permalink   Like   Dislike     Comment 7

Dan8267 says

String manipulation and lexicographical processing is typically CPU expensive compared to common database queries, especially doing things like searching for substrings.

I know what string manipulation is, I wanted to know what he was doing, to understand why he was doing what he's doing.

TPB   Tue, 21 Feb 2012, 9:24am PST   Share   Quote   Permalink   Like   Dislike     Comment 8

Are you using strong typed object classes, or verbose routines and inline functions on page load?
I mean Do you have class objects representing the threads, and one for each post, and another for the user?
Or is all of this constructed onload with a series of database calls, and then string manipulation is applied accordingly? Which is how a lot projects start out. But then become resource monsters, every time you go back in to inject a little more logic here, and a new check there.
This is why I love business layers, where you can build the logic to have data just save it as it should be saved in the first place. So you don't have to extrapolate data on the fly when you build the page.
But it can certainly be more efficient to load objects, than use business rules to manipulate your strings, in those classes, and have your data ready before you call it to populate a control.

Patrick   Tue, 21 Feb 2012, 11:41am PST   Share   Quote   Permalink   Like (1)   Dislike (1)     Comment 9

TPB says

Are you using strong typed object classes, or verbose routines and inline functions on page load?

God forbid! I have a well-justified hate of object-oriented anything, having worked on performance problems for years.

The site is just brutally simple plain PHP. Mostly derived from Wordpress, but I stripped out all objects and actually the majority of the supposed "functionality", so it's hardly Wordpress anymore.

TPB   Tue, 21 Feb 2012, 1:06pm PST   Share   Quote   Permalink   Like (1)   Dislike (1)     Comment 10

Patrick says

God forbid! I have a well-justified hate of object-oriented anything, having worked on performance problems for years.

The site is just brutally simple plain PHP. Mostly derived from Wordpress, but I stripped out all objects and actually the majority of the supposed "functionality", so it's hardly Wordpress anymore.

Right on, I love objects because because they suit my needs though of course if it wasn't for generics and LINQ I could never do what I'm able to do using collections or arrays, in the System.Collections namespace.

At the end of the day, your comments are like how I feel about most over used patterns. I just use what works where it works. I've done some crazy nested objects using nested generic objects with relational list(of T) children objects. That have stood up to rigorous loads and heavy user scrutiny. So I've learned to think of things in objects. It's more effort up front, building the schema and wiring in your code classes to match or represent the data.
But once that is done, the reusiblity of those objects are a cinch, as is calling them in the web application.

Then as soon as I think I'm getting good at it, every one is looking for MVC which best marries the biggest no nos of ASPX and ASP utilizing long brutal OnGet or OnPut routines then placing the data in the webpages like old school ASP like Alligator tags. So at the end of the day, no matter how far you go out or into developing a standard. At the end of the day, the code still has to do the work you're deferring for deeper layers, they're just adding more layers and complexity.

I've been programming OOP for so long, programming Android or MS MVC for that matter. Forces me to do other work arounds, to accomplish what would otherwise be simple solutions.
It's so pointless that sometimes, I feel the best solution would be just to write verbose inline routines and functions in all of the event calls.

Kevin   Tue, 21 Feb 2012, 1:58pm PST   Share   Quote   Permalink   Like   Dislike (1)     Comment 11

Dan8267 says

Patrick says

But maybe wouldn't help network latency.

That's what quantum entanglement if for. I don't remember where the article I read was, but there was a experiment which claimed to transmit a Beethoven symphony faster than light using quantum entanglement. The idea was that the spin of two electrons are always opposite according to some conservation law. You change the spin of one and the other changed immediately. What Einstein called spookiness at a distance.

You can't transmit information with quantum entanglement. This is quantum mechanics 101. As soon as you observe the state of an entangled particle, you change it.

This makes quantum entanglement useful for things like exchanging crypto keys and certain computations, but it can't actually transmit information.

Dan8267   Wed, 22 Feb 2012, 4:26am PST   Share   Quote   Permalink   Like   Dislike (2)     Comment 12

Been too long since I read it. I can't remember the details or find the article now. If any one else can...

Dan8267   Wed, 22 Feb 2012, 4:31am PST   Share   Quote   Permalink   Like (1)   Dislike (1)     Comment 13

Patrick says

Can't cache them, because they have to be up to date with latest comment, customized for your user ID, etc.

Yeah, you pretty much have to build the page from scratch on each request because of things like the ignore feature.

It's impossible to optimize code without seeing it, so the only thing I can suggest is to make sure you're code isn't copying the strings into a larger memory block whenever it runs out of room in the current block. Of course, if you are writing to a buffered network stream, that should be ok. You don't want to have to build the whole page in memory before sending it out.

Kevin   Wed, 22 Feb 2012, 1:17pm PST   Share   Quote   Permalink   Like   Dislike (1)     Comment 14

You can still cache things even if you have ignored users.

Just cache an intermediate representation. Omit the chunks that shouldn't be shown at render time.

Hell, just caching for users who aren't logged in will probably be a big win.

Hysteresis   Thu, 23 Feb 2012, 12:12am PST   Share   Quote   Permalink   Like   Dislike     Comment 15

Kevin says

You can still cache things even if you have ignored users.

Just cache an intermediate representation. Omit the chunks that shouldn't be shown at render time.

Hell, just caching for users who aren't logged in will probably be a big win.

+1

a big chunk of each web page can be cached because those parts don't change.

Patrick   Thu, 23 Feb 2012, 12:33am PST   Share   Quote   Permalink   Like   Dislike     Comment 16

Hysteresis says

+1

Did you notice the "Like" link? If not, how can I make it more noticeable?

Kevin says

Hell, just caching for users who aren't logged in will probably be a big win.

Still has to have uncached things like "minutes ago" on the home page, and the comment excerpt.

Hysteresis says

a big chunk of each web page can be cached because those parts don't change.

Yes, I should do that. It's work and added complexity, and the potential for caching errors (not updating cache correctly) but probably worthwhile.

Dan8267   Thu, 23 Feb 2012, 3:35am PST   Share   Quote   Permalink   Like (2)   Dislike     Comment 17

Patrick says

Hysteresis says

+1

Did you notice the "Like" link? If not, how can I make it more noticeable?

One word: boobies.

Kevin   Thu, 23 Feb 2012, 7:24am PST   Share   Quote   Permalink   Like (1)   Dislike     Comment 18

Just update the time snippet. You can even do it in JavaScript.

Patrick   Thu, 23 Feb 2012, 8:32am PST   Share   Quote   Permalink   Like (1)   Dislike     Comment 19

Thanks! I had not thought of the JavaScript solution for that.

But I still can't cache the home page very long, because each new comment causes that thread to pop to the top of its forum, and people want to see it.

TPB   Thu, 23 Feb 2012, 8:59am PST   Share   Quote   Permalink   Like (1)   Dislike (1)     Comment 20

This why I love objects, there doesn't seem to be that much data in the forum summary lists. I would keep a global collection going that I updated when posts were made or edited. But then that could cause greater problems than would be worth it, for the milliseconds I might save. That's if there was a serious issue, where something had to be done at all. I get the feeling you're just splitting hairs with your self, in a quest for the most optimization as possible. But if this project was for a client, you would be more like...
"Dude what in the hell is wrong you, this is the best it's going get, with out getting crazy and creating more problems. "

TPB   Fri, 24 Feb 2012, 12:46am PST   Share   Quote   Permalink   Like   Dislike (1)     Comment 21

John Bailo says

Another example is the yes/no/yes faster than light experiments. Now the word is its a faulty cable.

Which proves my point, half of it, is not science at all, it's just a matter of getting your model or contraption as it were. Produce the results you want. I can paint Elvis behind a drum set, doesn't mean he was a drummer.

Kevin   Sat, 25 Feb 2012, 11:09am PST   Share   Quote   Permalink   Like (2)   Dislike     Comment 22

Oh, yeah, science is a big industry full of people just out to get rich. Why, I hear some of these researchers even make six figure salaries after 20 years or so!

TPB   Sun, 26 Feb 2012, 1:24am PST   Share   Quote   Permalink   Like (1)   Dislike (1)     Comment 23

Kevin says

Oh, yeah, science is a big industry full of people just out to get rich.

No they've been drafted by Politicians, to be the all knowing unquestionable Wizard behind the Curtain. So much so, that I have to take the position, that places me on the receiving end of an inquiry, about my favorite subject.

Why don't Scientist do what Scientist do best, make my Intel Processor small enough to fit in the space of a wrist watch, with the power of a IBM enterprise blade server? And let me worry about the Assholes in Washington, and wade through the B.S. and chose my own positions in the issues they invent.

Kevin   Sun, 26 Feb 2012, 7:01am PST   Share   Quote   Permalink   Like (1)   Dislike     Comment 24

TPB says

No they've been drafted by Politicians, to be the all knowing unquestionable Wizard behind the Curtain. So much so, that I have to take the position, that places me on the receiving end of an inquiry, about my favorite subject.

Other than that being a complete lie, you're absolutely correct.

TPB says

Why don't Scientist do what Scientist do best, make my Intel Processor small enough to fit in the space of a wrist watch, with the power of a IBM enterprise blade server?

You're confusing scientists with engineers.

TPB   Sun, 26 Feb 2012, 9:00am PST   Share   Quote   Permalink   Like (2)   Dislike     Comment 25

And You're confusing Doctors with God.

bdrasin   Mon, 27 Feb 2012, 1:07pm PST   Share   Quote   Permalink   Like (1)   Dislike     Comment 26

Hail Purdue! That's my Alma Mater - good to hear her making the news for something other than for underachieving sports teams!

Kevin   Mon, 27 Feb 2012, 3:38pm PST   Share   Quote   Permalink   Like   Dislike (1)     Comment 27

TPB says

And You're confusing Doctors with God.

William E Baughb

Wrong, doctors exist.

TPB   Mon, 27 Feb 2012, 10:57pm PST   Share   Quote   Permalink   Like (1)   Dislike     Comment 28

Har!

Mick Russom   Thu, 1 Mar 2012, 4:56pm PST   Share   Quote   Permalink   Like   Dislike     Comment 29

Yeah, they did this single atom transistor at ultra low temperature (below 1 kelvin).

But the idiot main stream media doesnt even know what a kelvin is.

I believe they got a 12 atom transistor working at normal temperatures, thats more newsworthy.

TPB   Thu, 1 Mar 2012, 10:25pm PST   Share   Quote   Permalink   Like   Dislike     Comment 30

The inventor of the Kelvinator.

Dan8267 is moderator of this thread.

Email

Username

Watch comments by email

home   top   questions or suggestions? write p@patrick.net