Monthly Archives: July 2012

Explaining Quantum Computing – Blogroll Memory Hole Rescue

So what is quantum computing?

This is the most dreaded question for anybody involved with this field if posed by a friend or relative without a physics background.  When I am too tired or busy to make an honest effort, I usually answer by quoting Feynman's quip on quantum mechanics (click here to listen to the man himself - the quote appears about 6:45 min into the lecture):

"A lot of people understand the theory of relativity in some way or other.  (..) On the other hand, I think I can safely say that nobody understands quantum mechanics."

The crux of the matter is that Quantum Computing derives its power from precisely the same attributes that make this realm of physics so alien to us.

It's small comfort that greater minds than mine have been mulling this conundrum. For instance, a while back Scott Aaronson described his struggle to write a column for the New York Times that describes Quantum Computing.

Then there is Michael Nielsen's take on it, a brilliant write-up illustrating why there is really no reason to expect a simple explanation.

But if this hasn't utterly discouraged you then I have this little treat from D-Wave's blog. You need to be willing to tolerate a little math, understanding that an expression like this

$ \displaystyle ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ \sum_{i} x_{i} $

means you are summing over a bunch of variables x indexed by i

$ \displaystyle ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ \sum_{i} x_{i} = x_{0}+x_{1}+x_{2}+ ... $

Other than that it just requires you to contemplate Schroedinger's light switches.  Just as his cat can be thought of as dead and alive at the same time, his light switches are in a superposition of On and Off.  Strictly speaking, D-Wave's description is specific to their particular adiabatic quantum chip design, but nevertheless, if you get your head around this, you will have a pretty good idea why a Quantum Computer's abilities go beyond the means of a classical Turing machine.

More Lost Papers

Lost Papers Dropping from the Page of Time

As promised, the translation of the paper that contains Einstein's last important contribution to modern physics has been completed.   Paul Terlunen graciously provided the initial translation, and my wife Sara helped immensely with the final editing.

Starting point for this effort was the news that Einsteins hand-written manuscript had been re-discovered.  This has been reported in various physics LinkedIn groups, and subsequently there was some interest to look at this paper in an English translation, but there was none readily available.

This is the second part of Einstein's publications on this topic and I will now start on a translation for the first paper as well.

Nevertheless, if you already have some familiarity with quantum mechanics, you can read this last paper without having worked through the previous one.  It is intriguing to follow the author's line of thought, and to share in his intrigue with the puzzling nature of the quantum statistics that were uncovered for the first time. To the modern reader, it is also interesting to see in hindsight where Einstein erred when speculating on the nature of the electron gas; At the time, he did not know that his statistics only applied to bosons, and that electrons would turn out to be fermions.



Vested in IT Security and Losing Sleep Over Quantum Computing?

If so, then this special edition of the IT security magazine hakin9 might be just what the doctor ordered.

Your humble blogger was asked to provide an article and so I contributed the "Who’s Afraid of the Big Bad Quantum Computer?" piece.

It is very smart of the editor staff of hakin9 to approach bloggers, as we already write material for free and an ample sample portfolio of articles can be readily perused. As with this blog I do not benefit in any way financially from this. For me this magazine is just another good venue to get the word out about this exciting technology that is now becoming an IT reality.

Feynman would have approved

Astute followers of this blog know that quantum computing was the brain child of Richard Feynman whose contribution to the quantum field theory of electrodynamics earned him a Nobel prize. Feynman was the first to remark on the fact that classical computers cannot efficiently simulate quantum systems. Since then the field has come a long way and it has been shown theoretically and experimentally that quantum computers can efficiently simulate quantum mechanical multi-body systems.  And recent experimental setups like NIST's 300 qbit quantum simulator are destined to surpass anything that could be modeled on a classical computer.

Yet, for the longest time it was not clear if quantum computers could also efficiently simulate quantum field theories.

Fields are a bit more tricky.  Just recall the classic experiment to illustrate a magnetic field as shown in the picture.

Every point in space is imbued with a field value, so that even the tiniest volume of an element will contain an infinite amount of these field values.

The typical way to get around this problem is to perform the calculations on a grid.  And the algorithm introduced by Jordan et. al. in this, just three pages long, paper does that as well.

Unfortunately, Feynman is no longer around to appreciate the work that now made it official: Quantum Field theories can be efficiently simulated i.e. with polynomial time scaling.

It is quite clever how they spread their simulation over the qbits, represent scattering particles and manage to derive an error estimate.  The fact that they actually do this within the Schrödinger picture makes this paper especially accessible.

If you don't know the first thing about quantum mechanics, this paper will still give you a good sense that the constriction of quantum algorithms does not look anything like conventional coding - even, as is the case here, one using the gate based quantum computing model.

This goes to the heart of the challenge to bring quantum computing to the masses.  Steven Job's quip about the iPhone is just as true for any quantum computer:  “What would it be without the software? It would make a nice paperweight!” (h/t R. Tucci) Only difference is that a quantum computer will make a really big paperweight, but otherwise it'll be just as dead.

This somewhat resembles the days of yore when computer programs had to be hand compiled for a specific machine architecture.  Hence, the race is one to find a suitable abstraction layer on top of this underlying quantum weirdness, in order to make this power accessible to non-physicists.

Just in case you wondered:  It is still not clear if String theories can be efficiently simulated on a quantum computer.  But it has been suggested that those that cannot should be considered unphysical.