Hi everyone, best wishes as we all wait patiently through the COVID situation.  What a challenging time for our leaders at home and abroad as they try to balance the trade-offs of personal health / safety versus damaging economic impact.  I hope that everyone seeks to understand the data and subtleties on BOTH sides of this issue.  However, what an appropriate time for quantum computing advancement that might bring new tools for R&D towards vaccination and related medical research.

Next Steps with Algorithms – Part 1:

As knowledge on basic principles of quantum computing and gate-level mechanics improves, the experience begs the questions “So what?” and “What next?”.  With a little extra free time at home, I took some first steps in trying to understand basic principles of quantum algorithms, picking a copy of O’Reilly’s Programming Quantum Computers:  Essential Algorithms and Code Samples.

To get started, it’s useful to remind yourself of a few of the basic principles of quantum computing we’ve encountered so far:

• Superposition:  The property of a qubit to exist in a probabilistic state of 0 and 1.   Aka “zero and one at the same time.”  The best analogy I’ve heard is that of spinning a coin on a table.  It’s neither heads nor tails until it stops.  While spinning it’s both heads and tails at the same time.
• Entanglement:  The concept that the value of two (or more) qubits are directly related to each other, despite not being physically connected.  This principle is particularly hard to believe as it defies logic in many ways, but has been proven to be true.   The primary example is the simple controlled not (C-NOT) where the value of qubit A “controls” the value of the other.
• Reading a Qubit’s Actual Value Destroys the Superposition State:  Both a useful property, such as in Quantum Key Distribution (QKD), and a limitation in that you can’t just “read” a qubit without losing information about the superposition state.  That same spinning coin is either heads or tails once it stops, but not both.

I’m only about half-way through the book, but it’s a good time to relay some initial impressions.  Initially, the book is written in a way that emphasizes learning of the algorithms vs the underlying math and physics.  The approach has been helping in taking a next step in quantum learning, particularly as someone without a quantum physics background.

The immediate first take-away is perspective on using superposition with traditional binary logic or operations.  This was an initial taste of what “zero and one at the same time” can do:  multiple state solutions to a “problem” / circuit exist concurrently with a probabilistic rate of final occurrence.  For example, suppose you have two 2-bit registers initialized to zero, and the second order bit in the second register is in superposition: ’00’ + ‘[0|1]0′.   You have a certain chance of getting ’00’ (|0>) as the answer, and a certain chance of getting ’10’ (|2>).  

Related, it became apparent qubit gates can be configured to achieve traditional binary techniques in the context of superposition (as described above).  For example, bit shifting with Qubits:  shifting all the bits to the left acts as a multiplication by 2.  Brings back fond college homework memories.

Next, and the second major take-away from the book for me, the concept of qubit phase is introduced as a second method (behind 0 and 1) to encode information.  To get an idea of this we need to revisit the Bloch Sphere and the concept of phase shifts in our good ole friend the sine wave…

When we looked previously at the Block Sphere we considered the Hadamard gate placed the qubit in a superposition of |0> and |1>, which we specifically described (|0> + |1>) / √2.  It became apparent that the ‘+’ part of that refers very specifically to the state that results from applying the Hadamard to a qubit initially in the |0> state.  A Hadamard applied to a qubit in state |1>, on the other hand, results in a superposition state at the |-> position on the opposite side of the Bloch Sphere and implies a 180 degree phase shift in the qubit state.  From there, many of the expected concepts of familiar signal processing processing / analog analysis begin to become apparent.

At this point it’s relevant to note that O’Reilly introduced a new visualization method (new to me at least) that incorporates the logical value of the qubit(s) (0 or 1) as a circle, the superposition state (probability of the value) as the amount of the fill of the circle, AND the phase as indicated by the line on the circle.   Up for |+> state, down for |->.  Here’s a quick snapshot.

Source: Programming Quantum Computers by Eric R. Johnston, Nic Harrigan, Mercedes Gimeno-Segovia

In order to begin thinking about algorithms, a visualization such as this is required to go beyond simple binary state.  So far, it’s been tremendously helpful for understanding the importance of basic phase concepts.  While I don’t want to re-create the book, I do want to mention a couple key quantum primitives as take-aways thus far:

Amplitude Amplification:  This technique takes a quantum register, where qubits are in superposition, with certain qubits in differing phase states, and converts the register to a corresponding state value that indicates those qubits that were in a unique phase.  In other words, a phase shift finding algorithm.  This method, which uses a technique called mirroring, is apparently a building block for future quantum algorithms.

Quantum Fourier Transform (QFT):  Just like Discrete Fourier Transform (DFT) from signal processing, frequencies in qubit register state values can be calculated directly.  This is a useful primitive in future period finding algorithms.  One that quickly comes to mind is Shor’s Factoring Algorithm, which is referred to frequently (pun intended) in early quantum reading as one that has quantum advantage,  that leverages period as a sub-task to finding prime number factors.  A quick look-around in the chapter titles confirms this.  I’m looking forward to reaching those.

Again, you can buy the O’Reilly book here:  Programming Quantum Computers.

Industry Updates:  Honeywell, D-Wave

Honeywell – New provider for IBM Qiskit:  In my previous post, I spent a decent amount of attention on the buzz coming out of Honeywell that they had (or at least will) be releasing the fastest QC to date.   And as part of that post, Microsoft will be offering Honeywell as a back-end to their quantum service.  Overall I was left with the impression Honeywell, with their tech being based on Trapped Ion qubits, would be a distinct competitor to IBM; however, I read a surprisingly “quiet” announcement that a Honeywell provider for qiskit had been released.   Sure enough, here it is.  Another step towards major providers (IBM in this case) offering multiple back-ends to their service.

D-Wave – Leap 2 Launched (in February, sorry I’m just getting to it):  One announcement I breezed by was D-Wave releasing their second generation development environment – Leap 2.    It includes a refreshed, python-based IDE, which uniquely includes a hybrid solver services, that “automatically runs problems on a collection of quantum and classical cloud resources, using D-Wave’s advanced algorithms to decide the best way to solve a problem.”  Very cool.  It’s worth checking out their 30min overview; however, the example portion in the second half will be more meaningful if you have quantum programming experience – more than I have, I’ll admit I wasn’t able to follow the examples very well.

Quantum & COVID-19:

I’ll end on a couple quick COVID-19 notes.  Hopefully the day is around the corner that quantum computing could potentially assist pharmaceutical research in quickly developing vaccines.  In the current situation, quantum companies have been opening their doors to medical personnel as needed with easy / free access to quantum platforms as well as data & quantum techniques.  There hasn’t been to be a lot of specifics currently on directly contributing success; however, the collaboration to help those closer to the front lines is inspiring.  Here’s a few articles for reference.

Quantum Computing inc. Supports State & Local Governments with COVID-19 Data

National Science Foundation Taps Clouds for Quantum Computing Research (including Amazon, IBM, and Microsoft)

D-Wave gives anyone working on responses to the COVID-19 free cloud access to its quantum computers

Inside the Global Race to Fight COVID-19 Using the World’s Fastest Supercomputers
** This was by far my favorite article.  Written by Dario Gil, Director of IBM Research, it ends with:

Humanity has more tools at its disposal in this pandemic than ever before. With data, supercomputers and artificial intelligence, and in the future, quantum computing, we will create an era of accelerated discovery.  The consortium is an example of a unique partnership approach, and it shows that the bigger the challenge, the more we need each other.

Stay safe out there!

Hi, all. This week brought an updated preview of Honeywell’s upcoming quantum system release. Personally, I’m still getting over that feeling “Wait….Honeywell???”, but this seems to be real. It’s a good time to update an industry perspective. What’s important here with Honeywell’s announcement:

• The underlying technology is Trapped Ions. This is a departure from the Superconducting approach that IBM & Google have been pursuing. Note: Honeywell is not alone. IonQ is another notable in the Trapped Ion space.
• Microsoft Azure is going to be an available “cloud” front-end to Honeywell quantum services. This is a move seemingly out of the recent Amazon playbook (part of the usual flurry of re:Invent launches), where recently Amazon announced the Braket service. Braket will front-end quantum systems from D-Wave, IonQ, and Rigetti. (Btw, at this point we should all know Bra-“Ket” is a play on quantum Ket notation, such as state |0> or |1> ).
• Honeywell expects to make an exponential, leap-frog step to quantum volume 64, from IBM’s latest release at 32. Aside from the raw performance statement, this is a nudge towards quantum volume as a measurement standard vs qubit count.

From an industry perspective, IBM has continued to feel like the industry leader, at least in terms of apparent commercial system availability, marketing, and awareness. IBM Qiskit/QASM SDK tooling has become a standard point of comparison, although first half of 2019 Microsoft made moves to open source their quantum SDK, Q# (of course, before there was C#). We have yet to see a pure Microsoft quantum system, so this Azure news paired with Honeywell is hinting at their strategy. Like Amazon, it appears they could be taking a front-end approach, vs the high-capital system development path where Google, IBM, Honey, and smaller players such as IonQ, D-Wave, and Rigetti are battling for IP.

Recapping tech incumbent position:

• Google: High R&D investment in superconducting systems. (Due for a splash!)
• IBM: High R&D investment in superconducting systems AND large developer push with Qiskit SDK and QASM programming language
• Microsoft: Large developer push with Q#, apparent Azure strategy (Updated 3/9) with Honeywell, IonQ, and QCI.
• Amazon: (Expected) Cloud-first strategy with D-Wave (invested in by Bezos ventures), IonQ (headed by a former AWS exec), and Rigetti (just took another investment round of $71M).

Before wrapping this up, given my previous blog with a qubit count view, I also need to create a performance view using quantum volume. The Honeywell reads have been leverage quantum volume metric notation, but it should be noted that IBM originated the quantum volume concept in 2019.

Updated Q-bit Count View

Quantum Volume view per IBM Blog FYI:

And, here’s a fresh chart for quantum volume based on the latest article. Yah, it’s a little sparse. More to come on a cross-company view as this evolves….

Finally, I’d like to point out the headline in the Rigetti investment article:

…..Because Quantum Computing is Hard.

I’m glad others agree. 🙂 However, it’s exciting to see the rapid advancement, and investment, in the space.

Hi, all!  Happy holidays and happy New Year!

In the Blog thus far, the discussion has been focused on qubit basics and technology/coursework, admittedly, oriented to the IBM approach: gate-model quantum computing, which seeks to leverage qubit states (0, 1, or a superposition of those values) conceptually in a similar way to classical logic/bit gates that implement a computationally intensive formula or algorithm to achieve a numerical result.  Such use case examples include Shor’s algorithm, for prime number factorization, or Grover’s algorithm.

Gate-model quantum computing; however, is only one approach.  We have yet to discuss quantum annealing, which uses an approach of measuring energy states to solve optimization use cases such as a molecular/protein modeling and the traveling salesman.  Quantum annealing has been in use longer, most prominently by D-wave Systems who has had a number of system announcements dating back to a 1,000 qubit system circa 2015.  Thus far with IBM Q and comparable gate model systems we’ve been talking about computation on the order of 50 qubits and quantum advantage occurring theoretically in the 50-100 qubit range (we’re not there yet), but clearly a different order of magnitude with the annealing approach.

Let’s take a second look at our qubit release chart with D-Wave applied.  Note trend lines are applied here for IBM Q (left y-axis) and D-Wave (right y-axis).  D-Wave has been on an exponential pace for adding qubits.  With only a year or so of data for so-called noisy intermediate scale quantum (“NISC”) computers leveraging the gate-model approach, it’s not immediately obvious what the future development pace will be.

So what is quantum annealing?  While there are numerous explanations online, I’ll give my own explanation with the intent of practice in mind.  Earlier I described annealing as an approach of measuring energy-levels of qubits rather than a pure logical state measurement.  Such “energy-level” measurement, or more specifically energy minimization, seems to be the most common description; however, I like to think of the concept as a fancy way of describing good ‘ole maxima/minima determination from classic calculus, just applied to a qubit architecture.

“Wait…calculus?!  Ugh.”  Yah, sorry.  Here’s a reminder image from mathisfun.com (See!  Math is Fun!).  One use case of derivatives is to find maximum and minimum points in a function, where such points tend to represent a key transition point of the function.  A similar technique is used in so-called “Gradient Descent” in many of today’s AI/ML learning algorithms.

In the case of quantum annealing, the solution to an optimization problem is represented by the lowest point of energy, or ground state, in the system.  The energy function itself is known as the system “Hamiltonian” and is derived from the interconnected relationship of the qubits.  In the case of D-Wave systems, this could be upwards of 5,000 qubits in their latest 5th generation release.   Note:  there is a wealth of information on Hamiltonian concepts available online; however, I will treat as out of scope here.

Here‘s an example of one of many online visualizations of quantum annealing showing the value of the Hamiltonian as a function over time, including points of minimum energy (local and global Minima values).

In an annealing algorithm, all qubits are set to an initialized state, typically a neutral or x-axis oriented state comparable to the effect of a system of Hadamard gates setting all qubits to a superposition state.  Then, by modifying the energy state of an initial set of qubits in the system as a trigger, over time the relationships of the individual qubits drive the entirety of the system to converge towards a respective maximum (“peak”) or minimum (“valley”) states.  The system could also diverge to an excited state.  Along the way each qubit is measured for it’s state, whether 0 or 1 or superposition, and summed to achieve the total energy value.  Upon repeating this process multiple times, the lowest measured state (“global minima”) is said to be the ground state – the solution (or >a< solution) to the problem.

Here’s a visualization of the algorithm above from a topological point of view.  Assume a vertical spin upward is a value of ‘0’, a spin downward as a value of ‘1’, and a horizontal spin in the middle as a superposition of the two as an initial state (this is consistent with a qubit modeled as a Bloch sphere).   The lines simply represent relationships or couplings between qubits, as defined by the Hamiltonian.  In this case the sample represents a measurement of ‘2’; however, this is only a point in time measurement.  Over the current pass and subsequent test cases, the energies may be altered differently in Step 2 in which case a lower total energy state (“global minima”) might be found.

One could only imagine the interconnection relationship of a 5,000 D-Wave qubit system, likely being wildly complex, and illustrating the permutations that would make an optimization problem impossible by a classical computer.

One additional topic, a key concept in quantum computing, is quantum tunneling, which is illustrated in the (2x) previous diagram showing a quantum system Hamiltonian progressing over time.  In quantum tunneling, the qubit state “tunnels,” conceptually it jumps, through an energy barrier to reach a lower energy state.  The best visual I could think of is…a video game reference – ever seen those critters in Dig Dug transform into ghostly eyeballs and mystically traversing through the barriers of dirt??  (advance to :31)

Now, in contrast to tunneling, imagine trying to travel OVER a hill (a “peak”) – a certain amount of energy would be required to get over the hill, assuming that amount of energy is even possible.  Quantum tunneling; however, occurs without overcoming such a requisite energy threshold, going directly “through” the peak to the lower energy state.  I’m not confident in my knowledge of quantum physics to attempt to explain further.  Suffice to say, the occurrence of quantum tunneling does occur with certain probability.  This probability is relatively small and decreases with increasing mass, hence the benefits of tunneling are typically only seen in quantum processes.

Real-world applications of quantum computing are starting to show up in press.  Here is a recent one published by Volkswagen used to perform traffic optimization – note they used D-Wave.

Volkswagen AG has successfully demonstrated the world’s first live use of quantum computing to help optimize traffic routing. During the Web Summit conference in Lisbon, Portugal, nine public transit buses used a traffic management system developed by Volkswagen scientists in the United States and Germany, powered by a D-Wave quantum computer, to calculate the fastest travel routes individually and in near-real time.

D-Wave’s annealing systems are now also available through the recently announced ‘Braket’ service by Amazon  (a play on ‘ket’ notation in quantum computing).  In addition to D-Wave, Amazon Braket also front-ends IonQ and Rigetti quantum services.  In should be noted that Jeff Bezos is an early investor in D-Wave (2012), and Amazon is also an investor in IonQ.  IonQ uses a trapped ion modality (qubit physical architecture) approach to gate-model quantum computing, unlike superconducting qubits used by IBM Q and Google.  One can see the investment by Amazon in technologies competitive to other big players.  I’ve yet to research Rigetti and will add an addendum at some point.

Summary:  Quantum computing continues to present new levels of learning, starting with basic qubit behavior & concepts.  Beyond that, it is important to understand that there are more than one way to leverage qubits and quantum properties – gate-model and annealing being the leading methodologies.

Thanks for reading!  Here’s a list of resources used in the pre-reading for this post:

• D-Wave Intro to Quantum Annealing Video Series:  (Each of these is about 6 minutes – WATCH THEM!)
  • What is Quantum Annealing
  • How the Quantum Annealing Process Works
  • Physics of Quantum Annealing
  • Measuring Quantum Physics in a Quantum Annealar
• Quantum Tunneling Intro: (Each of these is short, <2min)
  • Quantum Tunneling
  • Quantum Tunneling Explained
• In-depth Tech Talk by Google:  (~50min)
  • Quantum Annealing (2014)

 

 

 

The race to “quantum advantage” heated up recently with what felt like a series of one-ups from IBM and Google.  IBM was first to break news with the announcement of an enhanced version of the IBM Q system from earlier this year – now with 53 qubits compared to the earlier and first system launch at 20.  At more than double the first IBM Q launch and exponential quantum gain by qubits, this is an enormous step in potential performance.  IBM noted this 53-qubit would be available online initially.

Shortly after the IBM announcement, there was an announcement that Google achieved “quantum advantage” in lab experiments.   (Quantum advantage is the concept that a quantum computer can solve a problem that a classical computer cannot, i.e. in a viable amount of time.).  The paper, published in Nature magazine, was title ‘Quantum Supremacy Using a Programmable Superconducting Processor’ can be found here.

As stated by Google Researchers, “To our knowledge, this experiment marks the first computation that can only be performed on a quantum processor.”

My first reaction was “Already?!”.   In the coursework and reading I’ve done thus far, most have projected quantum supremacy to occur at on the order 100 qubit systems, so how was this possible?  What problem had they solved?  According to the article, the use case involved “comparing [the] quantum processor against state-of-the-art classical computers in the task of sampling the output of pseudo-random quantum circuit.”  It appears the goal was to create and sample randomness across the set of output bits, which is cool but not immediately clear how it would solve a real world problem.   The quantum processor used was a 54-qubit system and was called “Sycamore.”

Interestingly, it didn’t take long for IBM to dispute Google’s results, and publicly at that.   Per IBM Research blog:

Google’s experiment is an excellent demonstration of the progress in superconducting-based quantum computing, showing state-of-the-art gate fidelities on a 53-qubit device, but it should not be viewed as proof that quantum computers are “supreme” over classical computers.”

Ouch.  However, IBM’s analysis seems rather credible.  IBM points out the baseline algorithm referenced by Google was synthetic in nature, and given the extensive classical computing tools and resources available today, it is possible to beat Google’s quantum performance (or “quantum volume” as IBM may argue) using enhanced memory and storage techniques available to classic computers.

The important point here, which IBM did acknowledge, is that Google’s work does demonstrate a step forward in quantum computing.   I can’t help but wonder if it a coincidence that the quantum computer used by Google was based on 54 qubits, exactly 1 qubit more than announced by IBM just earlier?  IBM was forced to make the point in their article, acknowledging as such and ultimately giving credit for Google at least demonstrating progress in theoretical quantum computing technology.  As competitive as the quantum race has become, one wonders if “no press is bad press” in this situation, at a 1 qubit more, just enough for Google to smirk a little at IBM in appropriate forums.  

As a recap, here’s a quick scatter plot of the system announcements by IBM & Google – scatter plot for now until a better trend-line visual makes sense. I’ll work on keeping this up to date as new announcements occur.

 

Thanks again for reading!

Hi everyone. In my previous post, I started by apologizing for the delay in posting; I will begin today with both a similar apology and also a similar excuse. I’ve been CONTINUING to work on online classwork, completing the 3rd and 4th (final two) courses in the MITx PRO Quantum Computing Fundamentals curriculum.  Interestingly, since I last posted the 3rd and 4th courses don’t seem to be part of the training any more.  I’m not sure why that is, but I can say these courses on “Quantum Computing Realities” are a step more detailed, focusing on the physical limitations of current quantum technology, namely noise, and the workarounds.   Both the math and use cases examples are less about practical application and more for those interested in understanding the second order physics.  On to the update…

In my previous post, I revisited the concept of superposition, in which quantum bits (qubits) can supposedly be both 0 and 1 at the same time.  I will argue this is a bit theatrical.  In reality, they can be “between” 0 and 1 states, bring forth the concept of “maybe” or partiality to quantum computing that does not exist with binary logic.

We reintroduced the Block Sphere, where the state of a given qubit is logically expressed as it’s position on the sphere and better conceptualized as a vector than a simple 0 or 1 state.

The state at the top pole of the Block sphere is referred to as |0>, the state at the bottom pole as |1>, the state at the middle (not going into the math here) as (|0> + |1>) / √2.  Voila!… 0 and 1 at the same time.  Again, it’s not so much that the qubit is in both states at the same time, it’s that it’s somewhere in between, or indeterminate.  NOTE:  After thinking about the last I want to emphasize that the Bloch Sphere representation is a logical representation.  They are many physical ways to represent this behavior, which would make a good future post.

The power of a quantum computer comes from the ability to perform tasks on this “in-between” state, achieving a degree of parallelism that does not exist with classical computers.  One place where quantum computing becomes interesting is, should we measure the state of the qubit, we will measure a 0 or 1, but we can’t predict which get with the exception we tend to get either state with equal probability.

I offered to demonstrate superposition, here goes.  I will be using the IBM Q Experience online simulation tools.

Without going too far into the details, a basic first-step to modeling a quantum circuit is to initialize the bits to a superposition.  This is done, typically, but setting all the bits to a known |0> state, then applying what’s called a Hadamard gate, represented by the letter H, which sets the qubit midway between |0> and |1> (drastically oversimplified).   To further illustrate, I will introduce what’s called a C-NOT gate which logically behaves like a classic XOR gate.   Here’s the truth table for a reminder:

 In a C-NOT gate, the first bit (X) “controls” the second bit (Y), inverting the second bit (Y) when the first bit is set to 1.

Here’s what that circuit look like in the simulator.

Again, the Hadamard gate sets the [0] and [1] bits to the “in-between” state.  When we measure a bit after a Hadamard gate, we’ll get a 0 or a 1, but we won’t known which.  This applies directly to the [0] bit as we are measuring the output of the H gate directly.  The [1] bit is the controlled bit and we would expect a truth table like the one show previously – i.e. when bit [0] is 1, we will expect the [1] bit to become inverted.

Here are the simulated results:

As you can see, we get the expected results where bit [0] is 0 half of the time and 1 the other half.  And within each of those, the [1] is appropriately inverted (or not) depending on the value of [0].  Each row in the truth table accordingly is shown at about 25% probability of occurring.

More sophisticated quantum algorithms leverage this behavior with more meaningful circuits.  Finding a simple enough example algorithm to show here remains part of the journey.

Hi, everyone.  It’s been a few months since my last post; my reason is that I’ve been working in the background to get some real, working knowledge behind this blog by completing some online training – the Quantum Computing Fundamentals course offered online through MITx PRO.  My short review is: while it came at a pricey cost, it was well worth the investment in terms of the level of depth, hands on programming experience, and breadth of concepts taught.  This takes us, somewhat indirectly, to today’s topic: what do all the hype articles mean when they same quantum bits (“qubits”) are unique from classical bits because they are capable of being “zero and one at the same time”.

Related to this, you’ll typically hear this phrase as the short-hand explanation of the term superposition.  Here’s a few examples:

• Quantum Computing for the Qubit Curious “Quantum computing relies on quantum bits, or “qubits”, which can also represent a 0 or a 1. The crazy thing is, qubits can also achieve a mixed state, called a “superposition” where they are both 1 and 0 at the same time.”

• Quantum What? The Future of Computing and Electronics Is All About Qubits “You can think of a qubit as the equivalent to the classic bits in modern computing, with a twist. Like bits, qubits are also measured using our binary system of 1s and 0s. But unlike a classical bit, qubits can be both a 1 and a 0 at the same time. It gets even stranger.”

• How Quantum Computers Work: Defining the Quantum Computer “This means that the symbols on the tape can be either 0 or 1 or a superposition of 0 and 1; in other words the symbols are both 0 and 1 (and all points in between) at the same time.”

By the way, each of these articles is a solid introductory article – you should read each of them.  In my experience, this notion of “zero and one at the same time” repeats itself in online coverage similar to these, seemingly as a way to build exciting around quantum computing.  Of course, however, it should – it’s true…for the most part.  After taking the online class, I found “0 and 1 at the same time” to be an overloaded term, so let’s go through.

Pause for the history:  classical bits (built from semiconductor-based transistors, vacuum tubes, what have you) are the basis of modern electronics.  They can hold a state of 0 or 1, on or off.  That’s it, and it turns out you can do amazing things with 0 and 1.  Now imagine something could be 0 or 1 at the same time, it’s not hard to imagine you could do way MORE amazing things.  But how the heck is that possible?

A qubit does have the notion of 0 and 1, but it also has the hard-to-grip concept of in-between.  That’s what superposition is all about.  To visualize that, we have to go back to an earlier article that introduced a theoretical visualization of a qubit, based on a sphere.  This is formally called the Block Sphere.  The state of a given qubit is expressed as it’s position on the sphere and better conceptualized as a vector than a simple 0 or 1 state.

The state at the top pole of the Block sphere is referred to as |0>, the state at the bottom pole as |1>, the state at the middle (not going into the math here) as (|0> + |1>) / √2.  Voila!… 0 and 1 at the same time.  It’s not so much that the qubit is in both states at the same time, it’s that it’s somewhere in between, or indeterminate.

Getting into the next phase of “interesting”, the fascinating potential of quantum computing is in it’s theoretical ability to leverage such indeterminate states to produce faster computations than classical computers (“quantum advantage”).  Now that we have the means to create qubits (not too long ago we did not), the challenge of modern quantum computing is to find applications that leverage these properties.   I haven’t yet identified a straight-forward enough example that I can illustrate in a blog post, but I’m working on it.  As a near term alternative, I’ll at least demonstrate the effect of superposition, utilizing an online simulator.

There’s been a flurry of announcements following IBM’s commercial release of the IBM Q System One, so I read a couple additional articles in follow-up to yesterday’s post.

First up:  Goppion for the future: the display case for IBM first commercial quantum computer

IBM also announced plans to open its first IBM Q Quantum Computation Center for commercial clients in Poughkeepsie, New York in 2019.

This is a good time to take an aside and do a little “about the author.”  My first job coming out of college was in the IBM System z mainframe development team located in Poughkeepsie, NY.  I met some of the most intelligent, hard-working people I’ve ever come across in my professional career, and they were nothing short of a “family” dedicated to building amazing technology.   So, it’s hear-warming to hear this will also be the home of the Quantum Computation Center.   It should also be noted Poughkeepsie boasts a state-of-the-art Executive Briefing Center (EBC) – that helps too.

Much as classical computers combine multiple components into an integrated architecture optimized to work together, IBM is applying the same approach to quantum computing with the first fully integrated universal quantum computing system.

I think that just means putting it all in a box that you can actually transport.  Think PC, but then think mainframe, which IBM has a history of putting extensive design work into much as they’ve apparently done with the Q System One.

IBM Q System One is comprised of a number of custom components that work together to serve as the most advanced cloud-based quantum computing program available, including:

• Quantum hardware designed to be stable and auto-calibrated to give repeatable and predictable high-quality qubits;
• Cryogenic engineering that delivers a continuous cold and isolated quantum environment;
• High precision electronics in compact form factors to tightly control large numbers of qubits;
• Quantum firmware to manage the system health and enable system upgrades without downtime for users; and
• Classical computation to provide secure cloud access and hybrid execution of quantum algorithms.

Let’s break this down.  ‘Stable and auto-calibrated’ – makes, we read previously that quantum systems are highly sensitive to vibrations.  ‘continuous cold’ – check, need that.  ‘high precision electronics in compact form’ – I can only imagine getting the components down to a portable size has been an immense challenge.  Think of early cell phones and how mobile phone industry has been able to consolidate technology.  We’re at this point of history with quantum computers:

‘Quantum firmware’ to manage the health and enable system upgrades.  That’s basic block and tackling for building a computer, but >quantum< firmware does have a nice marketing ring.  ‘classical computation to provide secure cloud access and hybrid execution’ – This is a more subtle statement, but raises a common theme around quantum computers:  the notion that quantum computers will operate in a hybrid mode with classical computers. 

Extending the point in the previous post, the decision to use Goppion (self marketed as the “The Art of Case Design”) is impressive.  IBM clearly “spared no expense” (Jurassic Park homage) on this prototype system.  Congrats to IBM for combining all the necessary mechanical and environmental requirements to create qubit stability AND make it visually attractive, if not beautiful.  It will be interesting to follow adoption of the system.

Here it is one more time…pretty nice!

Also, check out the video:  https://youtu.be/LAA0-vjTaNY

 

 

 

Not wasting anytime since the last post, where I collected a few baseline terms for quantum computing, let’s jump into an announcement from IBM this week – the commercial release of the IBM Q System One quantum computer.  How does our first pass at quantum terms help read an article like this?  Before getting to that I have to say:  Man, does that box look sweet!

EDIT 9/22/19:  The img link from the original post is broken, here is a reference article:
https://www.engadget.com/2019/01/08/ibm-q-system-one-quantum-computer/

I appreciate that the author ( Meriame Berboucha,  Contributor Science, Laser Physicist), expressed sentiment of sci-fi movies when describing quantum computing.  Very much the way I reacted and described in my first blog, so we agree on that.

The author describes the notion of bits being in both states simultaneously – that’s superpositioning.   Her follow-on statement:

As a result, when more than one quantum bit (qubit) act in a similar manner they can interfere with each other meaning that more than one process can occur at a given time

I don’t quite follow how interference translates to more than one process occurring at a given time.  But the next comment…

Therefore, quantum computers are much faster than classical computers since they can carry out multiple processes in one go. This is normally known as parallelism and allows the computer to carry out a million computations at a given time.

…certainly introduces a new, however, familiar one – parallelism.  From classic computing, the “goodness” makes immediate sense.  The more processes I run in parallel, the faster my workload is completed (yah, yah, I know it depends on the workload.  Assume ones optimized for high parallelism.).  Assuming this behavior to be true, the relation to optimization problems is clear.

Next:

A qubit is typically either a photon, atoms, ions or electrons and computer scientists control them with control devices. Some of these include ion or optical traps or superconducting circuits.

Ah, this is interesting.  There are different types of transistors – different substrate/gate technologies.  Seems to be the same for qubits; however, this range of atomic & subatomic structure seems a level more complicated.

One of the main problems faced when building a supercomputer is how fragile qubits are. Qubits have to be made in specific environments that need to be isolated from the outside world. If a quantum system is to interfere with the outside world it collapses back into a classical state and so works like a classical computer, which is not ideal.

Ok, now we’re getting into some practical realities.  We haven’t figured out to make qubits stable yet.  “Isolated from the outside world” sure sounds like a relationship to ultra-cold operating environments we read about previously.

The Q System One looks very sleek in its half-inch thick glass

That’s for sure!

IBM worked with numerous scientists and engineers and even a Milan-based manufacturer that designed display cases for the Crown Jewels at the Tower of London.

The designer appears to be Goppion.  Don’t underestimate the importance of design appeal, ala Apple products.

More on specs:

It sits in a series of independent aluminum and steel framed which help eliminate vibrational interference. IBM’s Q System One has 20 qubits.

Quantum computers are apparently sensitive to vibrational effects in addition to cold – got it.  20 qubits is an interesting point.  We’ve read previously that beta quantum systems can go to the hundreds of qubits, so it’s interesting that IBM found 20 to be commercially viable.  Could be a first-mover advantage play.

But for it to achieve performance greater than classical computers the number of qubits needs to increase to at least 50 qubits, a milestone known as quantum supremacy.

As expected, there is a threshold where quantum computers outpace classic computers.  Apparently that’s around 50 qubits, but there is more to learn on why this number.

In the next post, we’ll run through this follow-up article on Goppion, the Milan-based designer of the IBM Q.

Terms:

1. Parallelism:  (A refresher)  Performing multiple processes at the same time.
2. Quantum Supremacy:  The point at which quantum computers outpace classical computers

Questions:

1. How are the different types of qubits constructed?  (photon, atoms, ions or electrons)
2. What kinds of problems can be solved with 20 qubits and why is IBM targeting this market?
3. Why is 50 qubits the specific threshold for performance outpacing a classical computer?

Fblo

Hi, welcome back.  Here we go.  In today’s blog we’re going to start the journey by walking through the article I referenced in the previous post – “WTF is a Quantum Computer?“.   Why this one???  Mainly because I find the title hilariously appropriate.  The purpose of the walkthrough is to create a simple list of common terms and ideas in quantum computing – a baseline, if you will.  Once again, most of us are not PhD’s in quantum physics, so it’s important 1.) to create an inventory of basic terms/concepts and, probably more important, 2.) to create a list of new questions to answer.

Here we go, in real-time thought….

“Explain…in 1,000 words”

Wow…that will be impressive.

I doubt that you’re afraid to ask, no one knows what a quantum computer is.  However, it’s great you made this meme.  I like the article already.

two important states of matter known as superposition and entanglement

Got my first two – superposition and entanglement.

Quantum computers do not use transistors (or classical bits), instead they use Qubits.

Next one – Qubit.  Note it’s ‘Qubit’, not ‘Qbit’.  People keep asking me which it is.

Qubits are the basic unit of information in a quantum computer.

That’s an excellent picture.  Yep, that’s how I envision a classical bit.  I didn’t know how to envision a Qubit, but that looks like the bit could be anywhere on the sphere (orbiting), or maybe just on the red line….0,1 or in between.  Wait, those are 0’s and 1’s, those have some funky | and > and square root symbols…what do those mean?

Qubits can be either a -1 or a 1, or have properties of both of these values, which is called superposition.

Wait, what happened to 0 or a 1, that’s not what the picture shows.  “Both” sounds confusing, but that’s ok.  We’re talking about quantum computing, so there’s going to be possible things that aren’t supposed to be possible.  Superposition…check.

Right away there’s a whole lot more possibilities for performing computations.

I don’t understand why yet, but that’s what everyone seems to think.

the Qubit can leverage a state known as quantum entanglement, whereby pairs or groups of quantum particles are linked so that each particle cannot be described independently of the others, even when the particles are separated by a large distance; opposite ends of the universe for example.  Einstein called this “spooky action at a distance” and it’s the theoretical basis for quantum teleportation.

And…woah.  Lemme read that again.  Ok, still a mouthfull.  Can I get a picture of that??

At this point you may be wondering, what’s really in that pipe, Albert?

Yes, that’s exactly what I’m thinking.  Excellent…pictures coming…

What matters (to those of us who aren’t quantum physicists) is that thanks to Qubits and the phenomena of superposition and entanglement, a quantum computer can process an immense amount of computations simultaneously, and much faster than a classical computer.

That’s funny, but not the picture I was thinking of.  Now I’m getting that feeling that I still don’t understand how a quantum computer works.

2. What are the practical applications of this stuff? 

Wait, go back, tell me more about quantum entanglement!

a thought experiment. Imagine a phone book, and then imagine you have a specific number to look up in that phone book. A classical computer that uses transistors will search each line of the phone book, until it finds and returns the match. A quantum computer, because it has Qubits, can search the entire phone book instantaneously, by assessing each line simultaneously and returning the result much faster than a classical computer.

How?

These massive variable problems are often called optimization problems.  For example, optimizing every airline route, airport schedule, weather data, fuel costs, and passenger information,

New term…bookmark it.

Classical computers would take thousands of years to compute the optimum solution to that problem. Quantum computers, theoretically, can do it in a few hours, or less as the number of Qubits per quantum computer goes up, which is already happening …

How?

That’s a great chart.  D-wave seems like the only vendor quantum computer vendor shown, did they write this article??

Steve Jurveston, managing director of the investment firm Draper Fisher Jurvetson, and an early investor in D-Wave Systems, the company widely regarded as a quantum computing pioneer and standard bearer, dubbed the phenomenon of the increasing capacity of quantum computers as “Rose’s Law.” (Geordie Rose, is the CTO of D-Wave, so it’s named after him.)

I get it, like “Moore’s Law” for transistor-based CPU’s.  Ok, I should probably note D-Wave Systems and “Rose’s Law.”   That’s pretty bold of D-wave to take the name, we’ll see if that holds.

Rose’s Law for quantum computing parallels Moore’s Law for semiconductor processor development. Basically, quantum computers are already getting really, really fast.

Yep, I get the analogy.

D-Wave sells and leases quantum computers to clients such as Google. The machines are rumored to cost between $10M and $15M, so start saving.”

No idea what Google is using it for, but hardly shocked to see their name or ability to spend that kinda cash.

Oh, and the latest generation D-Wave 2X system has an operating temperature of about 15 millikelvin, which is approximately 180 times colder than interstellar space.

That’s cold, got it.

If a D-Wave machine isn’t in the cards, IBM is already offering “the world’s first quantum computing platform delivered via the IBM Cloud,” meant to unleash quantum processing power to the masses

IBM also in the picture.

Modern cryptography (secret codes) relies on a mathematical function called prime number factorization. Basically, large numbers are broken down into prime numbers that can then be multiplied together to get the large number. Classical computers are not good at this and take a long time to crack cryptographic codes based on prime number factors. But, you guessed it, quantum computers are really, really good at it.

That’s interesting.  How does a quantum computer do prime number factorization?

Governments all over the world are racing to build quantum computers that can render all modern forms of cryptography obsolete.

Cybersecurity is a huge market right now, apart from quantum computers, but a government buyer makes A LOT of sense.

In an effort to develop hack-proof communications, the Chinese government recently launched into orbit what is said to be the world’s first quantum satellite. That satellite’s name is Micius.

International technology race?  Yah, I think so.  Jotting down Micius.

Quantum encryption is the idea of sending entangled particles of light (entangled photons) over long distances in what is known as Quantum Key Distribution (QKD) for the purpose of securing sensitive communications.

Woah.  “Entangled” Let’s test my understanding.  From earlier, that means the particles are linked somehow.  New terms ‘quantum enryption’ and ‘QKD’.

In QKD, both the sender and recipient measure the polarization of entangled photons they receive, by assigning each photon a 0 or 1. This creates a quantum key, that can be used to decipher an encrypted message.

I kinda get it.  The sender and recipient should be using the same key, that’s the definition of cryptographic key exchange.  Something still missing in my understanding of how the key exchange works.  But QKD is a new term, probably should record that.

The most important point is that if the quantum entangled photons are intercepted by anyone, the system will show immediate signs of disruption and reveal that the correspondence is not secure.

That helps a little bit.  If there’s disturbance, the keys must break somehow.

In short: Quantum computers rely on the fundamentals of quantum mechanics to speed up the process of solving complex computations. Often those computations incorporate a seemingly unfathomable number of variables, and the applications span industries from advanced genomics to finance. Also, quantum computers are already reinventing aspects of cybersecurity through their ability to break codes based on prime number factorization, as well as their ability to offer advanced forms of encryption for protecting sensitive communications.

End of the article.   The pictures were entertaining and I appreciate that.  It didn’t leave me with a good sense of how a quantum computer actually does calculations (“How?”), but there seems to be a large amount of use cases and I learned some terms.   Oh, how many words was that?? I got 1283, that’s close enough for me to count it.  I’d like to thank the author – I know more now than I did before.

Let’s summarize:

Terms:

1. Qubits:  The basic unit of information in a quantum computer.
2. Superposition:  Qubits can be either a -1 or a 1, or have properties of both of these values
3. Quantum entanglement:  Pairs or groups of quantum particles are linked so that each particle cannot be described independently of the others, even when the particles are separated by a large distance; opposite ends of the universe for example.  Einstein called this “spooky action at a distance”
4. Optimization problems:  Massive variable problems.  For example, optimizing every airline route, airport schedule, weather data, fuel costs, and passenger information,
5. “Rose’s Law“:  The phenomenon of the increasing capacity of quantum computers, similar to “Moore’s Law“.  Named after Geordie Rose, is the CTO of D-Wave
6. Micius:  “World’s first quantum satellite” launched by the Chinese government” to develop hack-proof communications
7. Quantum encryption:  The idea of sending entangled particles of light (entangled photons) over long distancesns the particles are linked somehow.
   • Quantum Key Distribution (QKD), both the sender and recipient measure the polarization of entangled photons they receive, by assigning each photon a 0 or 1. This creates a quantum key, that can be used to decipher an encrypted message.

Questions:

1. In a superposition sense, what do the | and > symbols mean?
2. How can a quantum computer search the entire phone book instantaneously.  (How does a quantum computer execute on an optimization problem?)
3. D-wave & IBM are building quantum computers.  Who else is currently in the business?  How many Qubits?
4. How does a quantum computer threaten current cryptography techniques?  How does it do prime number factorization?

Plenty to research, looking forward to it!

If you’re like me, your first reaction to the topic “quantum computing” doesn’t have anything to do with AI, cryptography, molecular analysis, or any other of the VERY interesting applications quantum computing aspires to solve according to headlines.  If you’re like me, your first reaction takes you back to an earlier period of life when computers weren’t as big of a deal, and dreams were inspired by shows like Star Trek™ The Next Generation™ (TNG) and The Last Starfighter™.  Inertia could be dampened, light speed was possible, things like that.

Eventually college came around and computers WERE a big deal, as was learning how to build them…with transistors of course.  It was all about 0’s and 1’s.   Other than a token class in modern physics (where I still remember being stunned by Einstein’s proof of E = mc2), subatomic particle theory was the future, maybe even fantasy,   Does time really dilate?  Could we someday travel at warp speed like the Enterprise?   That was all fantasy too.  However, these were fascinating questions, and they required imagination of all that was possible or could be true in the universe.  But who had the time or justification to spend on that?  Oh, and by the way, it required the intelligence of an Einstein to even participate in a well-informed conversation.  If you were like me or most other IT professionals over the last 10-20 years, you went on to study the 0’s and 1’s and probably have had a great career in the digital age, fueled by advancements such as multi-core processors, virtualization, and cloud computing.

Enter quantum computing: Read an article on quantum computing from the past year and you’ve probably run into terms like “quantum entanglement”, “superposition”, and “qubits”, and you’ve probably run back into Einstein as well (his theories, anyway).   Those concepts that once seemed futuristic sound like they may be becoming a reality, and THAT IS EXCITING.

Problem 1:  We’ve spent the last 20/30/more years learning how to build computers with transistors, not qubits.  With logic, not superpositioning.  The concept of 0’s and 1’s are relatively straightforward.  Binary if you will:

• 0+0 = 0
• 0+1=1.
• 1+1 = 10.  (You get that joke if you’ve read this far.)

We’ve learned that given enough 0’s & 1’s, enough transistors, you can build a processor.  And with enough of those, you can build some cool computers or wireless gadgets or robots and solve some really hard problems.

Which leads to the first question we ask ourselves – what does a quantum computer do that is so different?  Or perhaps you prefer “WTF is a Quantum Computer?”  This is where the journey begins.  To get to that answer today requires a PhD in theoretical physics, and I don’t have one, nor do many of my peers who have been in the IT industry for decades now.

Problem 2:  Many of the people trying to explain quantum computing today to the general population ARE PhD’s in theoretical physics….and explain like it.  There have many articles that try to explain quantum computing in a simple way.   Here’s an example:  Entanglement Made Simple.  However, I still end with a feeling of “Cool, but I still don’t get it.”  How exactly can I, say, calculate prime factorization using a qubit?  Or for that matter, how do I add 0+1.  Should I?  Why or why not?  There’s frustration simply with having the word ‘bit’ in the term ‘qubit’ but not still understanding how quantum computing relates to a classical computing ‘bit’.

This is where the journey begins.  In this series of blog posts, I’m going to document my learning about quantum computing and do so in a way that is understandable by the average person, or at least the average person with a basic knowledge of current day computing.  I hope you’ll join.   I hope we’ll learn a practical understanding of quantum computing, open our mind to future problems that could be solved, and maybe even get a step closer to the future reality we watched on TV when we were kids.  Please feel free to post questions or challenge any conclusions made – I’m not an expert…yet.

In the next post, I’ll walk through a sample article and begin documenting terms, context, and the questions that remain.   Thanks to wordpress.com for the following picture and quote which seemed rather appropriate.

Good company in a journey makes the way seem shorter. — Izaak Walton