Quantum Bayesian Networks

January 9, 2014

The NSA’s top secret Kane Quantum Computer

Filed under: Uncategorized — rrtucci @ 4:40 pm

Check out the following article which has generated much buzz on the Internet

“NSA seeks to build quantum computer that could crack most types of encryption”, by Steven Rich and Barton Gellman (Washington Post, Jan 2, 2014)


“The geographic scope has narrowed from a global effort to a discrete focus on the European Union and Switzerland,” one NSA document states.

By the end of September [2013], the NSA expected to be able to have some building blocks, which it described in a document as “dynamical decoupling and complete quantum ­control on two semiconductor qubits.”

I had heard that Booz Allen Hamilton (Edward Snowden’s former employer) had posted some advertisements offering jobs in quantum computing. I had also heard that the head of the NSA, General Keith Alexander, has, for the past 5 years, been bragging at public symposia about his agency’s growing prowess in quantum computing and predicting that once QC’s arrive, it will be a “game changer” (his words) for his agency. I knew that the U. of Maryland was doing some kind of classified work in quantum computing for the NSA. But I have to admit that the extent and depth of NSA’s classified work is bigger than I thought. I had thought that NSA’s secret work was just meant to follow the front-runners from behind. But now it appears that they are trying to be one of the front-runners themselves, at least for some types of quantum computer design. It even appears that the NSA has decided to classify TOP SECRET most of its work on the very promising Kane quantum computer. If that is true, it really sucks.

Let me give some background. Enter Bruce Kane (not to be confused with Bruce Wayne, Batman’s alter ego). I have no inside information about the guy. All I know is what I’ve read on the Internet: that he started off brilliantly in quantum computing, by postulating the Kane quantum computer. After working in Australia for a while, he accepted a position at the U. of Maryland, where he still works. I always wondered how come, after such a brilliant start, we haven’t heard much about his work at the U. of Maryland. (His arXiv record as “B.E.Kane” shows 20 papers in the last 17 years, nothing about “dynamical decoupling and complete quantum ­control on two semiconductor qubits.” ) And now I learn from the Snowden revelations that the NSA has a TOP SECRET project at the U. of Maryland that is building a QC that sounds very similar to the Kane computer.

It is widely known that the NSA has been throwing money at quantum computing ever since Peter Shor came up in 1995 with his algorithm for breaking most public encryption codes. The NSA even funded at the U. of Maryland (College Park campus), two institutes (one for unclassified work, one for classified work), which I’m sure has cost them a bundle, probably well in excess of 100 million dollars by now. Their institute for unclassified work is called the Joint Quantum Institute (JQI, pronounced Yucky?)
Their institute for classified work is called the Laboratory for Physical Sciences (LPS, pronounced Lupus?)

The Lupus website makes it fairly clear that Bruce Kane is one of the top honchos there, which explains why they are partial towards the Kane QC design. According to the Lupus website, and I quote,

LPS has housed an internal research program in quantum computing (QC) for several years, and it currently includes seven research groups. Four of the research groups experimentally investigate various solid-state systems at low temperatures, and are connected to either semiconductor-based or superconductor-based quantum computing. These groups are led by Bruce Kane, Kevin Osborn, Ben Palmer, and Michael Dreyer.
The remaining three research groups theoretically investigate a broad range of topics which include: solid-state quantum-computing systems, ground-state quantum computation, and quantum control. These groups are led by Frank Gaitan, Ari Mizel, and Charles Tahan. Further details on QC research at LPS can be found by visiting the links shown above.

Further thoughts about the Washington Post article:

It’s interesting that the Lupus people think their main rivals are the “European Union and Switzerland”. Ouch! No mention is made of Waterloo, Canada, or the Aussies. Dismissing Waterloo, Canada can be explained since those guys are still hard at work on NMR quantum computers and quantum crypto BS. As for dismissing the Aussies, who have made significant advances in building the Kane quantum computer, that might be explained if the Lupus people view the Aussies as just their faithful minions. That might mean that anything the Aussies have done so far vis-a-vis the Kane computer, the Lupus people have already done twice as well. After all, I doubt the US would passively take second seat to their wimpy Aussie allies/lackeys 🙂

It’s lamentable that due to all the secrecy, we can’t tell for sure how advanced the Kane QC model is. I’ve written some previous blog posts lamenting the fact that Intel seems to be totally uninterested in quantum computers, despite the fact that Moore’s Law is coming to an end (2014: 14 nanometer semiconductor nodes, 2019: 5 nm). The Kane QC, based on semiconductor qubits, would be a perfect fit for Intel, but due to all the infantile secrecy, Intel is blissfully ignorant of Lupus’s advances…or are they? It’s possible that Intel is secretly doing joint work with Lupus, although I have no evidence for that wild conjecture.



  1. […] Ever since the Edward Snowden-provided news broke that the NSA spent in excess of $100M on quantum computing I meant to address this in a blog post.  But Robert R. Tucci beat me to it and has some very interesting speculations to add. […]

    Pingback by Quantum Computing NSA Round-Up | Wavewatching — January 12, 2014 @ 12:04 am

  2. University of Maryland highlighted a new building to house the institutes in a recent Odyssey newsletter. http://www.cmns.umd.edu/sites/default/files/uploads/docs/odyssey_dec2013_web.pdf

    Comment by mikemcglincy — January 13, 2014 @ 4:32 pm

  3. OK, I got my physics PhD in the “Lupus” quantum computing group, and Bruce Kane was one of my bosses. He’s a smart and hard-working guy, but like many people who came up with one idea that made the cover story in Nature and launched a substantial research enterprise, he hasn’t yet come up with another one.

    The Kane computer has not been made by the NSA or anyone else because it is beyond the state of the art to make such a thing. A fair amount of progress at the single-qubit level has been made by Michele Simmons’ group in Australia, but they are still far from being able to make even a small, toy computer. If the atomic-level fabrication technology existed today, it would have many more uses than making a quantum computer, most of them much more important, and it is so far beyond what is being done in the best labs that we can be pretty sure the NSA doesn’t have one in a basement somewhere. Yes, the NSA is spending in units of $100M on QC-related research, but nearly all of it is unclassified and that includes everything of interest (probably some classified overview reports and trials of the D-Wave machine which are showing that it is a piece of junk).

    Quantum computing is a boondoggle project that physicists love because it’s fun and definitely physics, not just 2nd rate engineering. Quantum information is a profound and fascinating topic. But I never believed a practical quantum computer would be built and used to decrypt via Shor’s algorithm, or that QC was going to prove an important technology for any application. If it worked, sure, but after 20 years of effort we are still trying to make an adequate single-qubit register, let alone an entire computer, and there are good reasons to doubt that this effort will succeed before the government and corporations decide to pull the funding.

    Comment by Mark Gubrud — March 15, 2016 @ 12:19 pm

  4. Thanks Mark. Very wise and honest assessment.

    I agree with most of it, except for one major thing. I believe gate model quantum computers will be built in 10 years. Perhaps I’m a naive gambler, but maybe not. I think the reason that we didn’t make much progress the first 20 years but now it’s different is that before it was purely an academic endeavor but now industry heavyweights like Google, Intel and MS are in the driver seat. These companies are far from perfect but have a much better track record at getting practical things done than academia. Academics have little incentive to change their modus operandi, move fast, meet deadlines, or do practical things. They aren’t hungry enough. Look at Moore’s law. All it is is the military discipline of Intel. Now look at what some academics want to spend their entire careers working on: NMR qcs, quantum crypto and boson sampling, all dead end streets that industry would not touch with a ten foot pole.

    I think Shor’s algorithm is a nice theoretical result but I’ve never programmed a simulation of it. I’ve never believed it to be commercially viable. It is the hardest QC application to implement and we’ll be using post quantum crypto long before we can do Shor’s.

    But I believe QCs will shine in simulating quantum systems from chemistry and condensed matter physics.

    I think qcs will also excel in promoting the connection between classical and quantum Bayesian Networks, which are, from my point of view, the foundation of classical and quantum information theory. If classical Bayesian Networks are so useful, and believe me, they are, can you imagine how much more useful they can be if we extend them to include both classical and quantum physics?

    Comment by rrtucci — March 15, 2016 @ 2:33 pm

RSS feed for comments on this post. TrackBack URI

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

Blog at WordPress.com.

%d bloggers like this: