Quantum Computing

I want to build a Quantum Computer at home! — July 18, 2015

I want to build a Quantum Computer at home!

Qbitrex Dual GPU Water Cooled PC
Qbitrex Dual GPU Water Cooled PC with under the house radiator

For anyone who knows anything about the theory of Quantum computing that statement is the equivalent of a hobbyist saying

“I want to build a Fusion Atomic Reactor in my Garage!”

Why should I be taken seriously in this highly esoteric field when Governments and Corporations are throwing enormous amounts of money at the extraordinarily difficult technical problems associated with this new field of technology.

For those readers who have no clue about QC you need to watch some youtube and bring yourself up to speed to follow my argument.

A simple 5 minute cartoon style video can be seen here: Quantum Computers Animated

A more advanced short video as basis for get your head around a quantum register: How Does a Quantum Computer Work?

I have no degree, had a go at a few in Chemistry, Computer science and Arts. I’m a 60 year old small business computer network consultant semi-retiree grandfather who has more time than money on his hands but I am a technical reader from age 13 with a photographic memory inherited from my University Degree-ed mother.

Not being an academic I am free of the constraints of having to deal with the politics of being accepted by the elder high priests of the arcane mathematics directing research in the QC field.

I am a string theory devotee thanks to Professor Brian Greens “The Elegant Universe” books and videos, but that’s a whole new blog. Watch the vid at your leisure it’s not necessary to follow the logic of this blog.

The take-away from my String Theory research is forget about using anything more exotic than electrons for quantum computing. The electric field is the foundation of the physics of the Universe .

Here’s a free Nobel Prize giveaway to any Physicist wanting to do the work to claim it.

The four Maxwell equations line up with the Weak, the Strong, the Electro-magnetic and Gravitational forces. Everything exhibits polarity. The missing piece of the puzzle in the Grand Unified Theory of the Universe is Anti-Matter exhibits anti-gravity so it can never clump together to form planets, stars and galaxies. It is formed on the up swing of the entropy arc of the Planck Frequency Pendulum so it appears to exist in the opposite direction of time becoming less en-tropic over time.

This cloaks it from appearing in any statistical quantity on the Newton/Einstein scale of the Universe during this cycle of the big bang, it may be the opposite in a future or past serial universe finally making the old 18th century electric convention that that current is formed from moving positive charges true.

The US Government has reportedly quietly blown a billion $USD to date funding research on QC and prototypes are now in existence. The big carrot is QC’s theoretical ability to crack cryptographic code you wonder if its abilities where more in alignment with the Arts that they would be as interested.

DWave have built a multi bit quantum device from our best understanding of what QC hardware should need. You sort of read between the lines of the promo videos it’s producing output that customers are struggling to apply to real world problems.

Never-less i would love an opportunity to program it in Python in order to design a better iteration based on some of my ideas in this blog.

My approach at Qbitrex.com would be to use magnetism as a de-coherence filter to force the majority of the electron cloud in my quantum gate to spin in one direction thereby shutting off a potential leak of data due to perturbation of the probability field due to variations between entangled electron clouds.

We need a new math to make sense of what a  quantum computers will output.

I’m defining a system which has never been explicitly named but is the wood within the forest of quantum computing discussions. According to Holevo’s theorem a Qubit can only be measured as a binary bit of information either 1 or 0.

While it remains in the Quantum State it can do magical things but the moment we measure it to output a result the magic stops. OK lets consider a new ordinal system I’m calling it the Unary Number (Greek Symbol lower case Psi) its the ordinal system before the binary.

This intuitionally obtuse number system has at its unit a Qubit. Unaries are mentioned in the literature of Quantum mechanics and pure math but to my knowledge has never been dignified as a number system in its own right. I might stand corrected if some mathematician recalls an obscure paper that did define it as such, so I’m not claiming it as my invention I’m just re-phrasing something that has been passed over as self evident without the need for a detailed understanding until the new field of QC was imagined.

Once we dignify it as a number system in its own right we have at our disposal a new set of tools to consider QC calculations as they appear in the Newton/Einstein scales of the Universe.

Like a real number in higher ordinal systems it also has an axis at right angles to it that defines ts imaginary component.

To build a quantum register we need two entangled qubits. If we consider the basic NMOS FET transistor configured as an NOT gate or inverter as one qubit. It has the state of either being on or off. Entangle two of those transistors in a RS latch and viol-ah we have a quantum register albeit a very noisy one. A little bit more work has to be done to add some de-coherence filtering by placing them as close together as possible in mirror image and enclosing the whole package in a magnetic field circuit formed from a simple split torus choke with a permanent magnetic supplying the magnetic field to force all the electrons within the cloud to the same spin direction to prevent the aforementioned data leakage and we have a working device we can do some real world calculations at a fraction of the cost of the more exotic cryogenic-ally cooled devices. Because I am using a cloud of electrons singing like a soccer crowd we don’t need the confinement necessary to hear what an individual electron is singing.

After all if Quantum computation is the algorithm the Universe runs on it must also happen at room temperature. My thought processes within my neurons are quantum entanglement at the foundation Planck level and I don’t have to wear a cryogenic helmet to think. I’m not writing off cryogenics yet I don’t have the maths chops to prove it either way but my spider instinct tells me a QC should be energy neutral and what’s the point of spending all that cash on a containment system when the computational result is statistically the same as a room temperature device.

Make no mistake we are going to have to cool this sucker, quantum stuff happens at the Planck frequency of 1.85E×10+43 hertz our fastest super computer clock cycle is in the order of 33.86 petaflop/s or 3.386Ex10+16 hertz so we can’t build a digital device fast enough to slurp up the data from a QC running at full fizz. The question is what clock frequency to use, the fastest your digital interface can handle or will it be like one of those spooky Planck Units a specific value or harmonic of that value.

Relatively speaking the QC operating at the Planck frequency has been calculating for 5.46Ex10+26 or 546,000,000,000,000,000,000,000,000 years between our fastest computers clock cycle! Longer than the age of the Universe estimated at 1.38Ex10+10 or 13.8 billion years. Another hint as to the potential and power of QC. The question is does cryogenic cooling get us a significantly better statistic result than less expensive water cooling on a cheaper 16 nano meter IC form factor exploiting the quantum effect of Meta-stability in electronic circuits that I’m proposing in the Qbitrex concept.

See Figure 1 below a single Qubit Register serially clocked using 18 FETS exploiting the metastabilty of an RS Latch when the R & S pins are placed in a disallowed truth table configuration.

FET Quantum Register
Single Qubit Quantum Register using 18 NMOS FETS

This design is based on a true random number generator altered to account for the NOT Q result and with the ability to be initialized with a known state by Michael Epstein, Laszlo Hars, Raymond Krasinski, Martin Rosner, and Hao Zheng in the white paper “Design and Implementation of a True Random Number Generator Based on Digital Circuit Artifacts

I’m experimenting with individual plastic encapsulated ones but the ultimate solution is if they were fabricated on the same silicon wafer.

The output from the Quantum RS latch is Q and Not Q representing a real and imaginary number or a vector with magnitude and direction. The trouble with the existing prototypes they through away half the result by letting the not Q result flow back to Earth leaking precious energy/information!

But remember the two entangled Unaries are completely described in this state by Four Numbers or indices representing the 2×2 states.

Hang on a minute where have we met a number that has an real and imaginary parts that is described by four indices! Enter stage right to thunderous applause “The Quaternion“. I can see the maths students roll their eyes not those nasty beasties again I remember nodding off to sleep during that lecture.

Getting your head around quaternions is as frustrating as understanding quantum mechanics but they make computations of rotations in 4D space without gimbal lock and are returning with vengeance from their 18th century discovery!

Quoting Wikipedia:

However, quaternions have had a revival since the late 20th century, primarily due to their utility in describing spatial rotations. The representations of rotations by quaternions are more compact and quicker to compute than the representations by matrices. In addition, unlike Euler angles they are not susceptible to gimbal lock. For this reason, quaternions are used in computer graphics, computer vision, robotics, control theory, signal processing, attitude control, physics, bioinformatics, molecular dynamics, computer simulations, and orbital mechanics. For example, it is common for the attitude-control systems of spacecraft to be commanded in terms of quaternions. Quaternions have received another boost from number theory because of their relationships with the quadratic forms.

So the language spoken by quantum computations is Quaternion therefore we must phrase our questions to the quantum oracle in Quaternion Format and interpret the result as a Quaternion answer! Actually I believe its Fractal Quaternion in a multi-register system.

Don’t laugh but lets now consider my Quantum register modeled as an egg carton to see how this might work.

Here is a picture of my 12 Unary Number Quantum register made from a modified dozen egg carton giving us 6 quantum entangled registers.

6 Qubit Model Quantum Register Closed
6 Cubit Quantum Register model in the Quantum State
6 Cubit Model Quantum Register in Null State
6 Cubit Quantum Register Model in Null State prior to power up
6 Cubit Model Quantum Register Iniatilised
6 Cubit Model Quantum Register initialised with a real number.
6 Cubit Model Quantum Register with NOT Q Set
6 Cubit Model Quantum Register with NOT Q Set ready for close up and shaking
6 Cubit Model Quantum Register after shaking
6 Cubit Model Quantum Register after shaking opened up to show the Quantum Calculation

I painted it black to simulate the classical black box of the quantum state entanglement state and it contains 6 fuzzy blue balls representing electron clouds when energized otherwise its is in a null or empty set state before we turn the electricity on. It has barriers to restrict the balls from moving out of their individual dual register.

The egg carton can output a real and imaginary number number in the range 0 to 63 or 64 bits. It is initialized with a real number 43 binary 110101 in this example. The NOT Q result of this number is 20 binary 001010 the imaginary number. After entering the quantum state (close the lid and shake) when you halt the quantum calculation and open it up the new real and imaginary quantum number result is displayed.

The Q and Not Q outputs must always sum to 63 this is an error condition check that the Quantum Register is correct and ready for close up and shaking to place it in the Quantum State. Similarly when opened the Sum of Q and Not Q must equal to 63 or 2^n-1 for a n sized Quantum register. This can model a 64 Real x 64 1 bit Imaginary Complex Number matrix probability landscape. The initial complex number and its resultant quantum calculation a final complex number can be thought of as two vectors in 3D space the quotient of which is by definition a quaternion or If the initial real number is a magnitude that has to be conserved under the laws of energy conservation the quantum calculation provides the j imaginary k imaginary for the quaternion (Q Initial Real, Not Q i,Q j,Not Q k). It can be reset to its original state and run again in multiple trials which you would expect to follow the normal distribution curve for a real and imaginary numbers in the 2^n range of an n qubit sized register.

Another approach similar to the egg carton idea where the real world problem is more like a lotto game of 45 numbers of which only 6 is the winning combination is have a single line of 45 egg cups without side barriers between egg cups and only place in it 6 fluffy blue balls and shake it about. It will give a real number binary within the factorial 45 probability space of the problem the NOT Q imaginary result can then be calculated after the quantum calculation by NORing the quantum register with zero. To do that with entangled transistors you would need to cross connect the whole 45 transistors qubits representing the probability set ring.

If the QC was implemented as a FPGA then it can be quickly re-configured between runs to suite the real world problem.

The next step to create artificial intelligence is teach two conventional computers with attached QC’s to interpret the circuit designs of its connected QC implemented in FPGA and allow the one to observe the circuit design of the other and respond to the other by re-configuring its FPGA in a dialectic discussion back and forth between the two entangled machines.

I need to do some more research how the compressed QC data output can be applied to meaningful real world dialogue with a Quantum Oracle but suddenly a complex multi dimensional number (a quaternion) is available from a previously single real. That should make the mathematicians sit up and take notice and open up some fresh avenues of research.

As we add more registers  also entangled on the same chip in a real QC the registers I believe are in Fractal arrangement with each register a power of two higher than it’s brother before it. Each register is one fractal scale higher in the solution. The plot of which would look like a crack through a square matrix.

I believe we can build Universal Quantum computers but we will need to have scaleable registers that can be built up or down depending on the probability area of the quantum landscape problem being considered. For example to calculate a quantum result from say a Cancer drug combination requiring 100’s of thousands of permutations compared to a Chemistry materials problem with 100’s of millions of permutations requires a larger array to make sense of the result. That’s why a FPGA device (I love Xilinx) that can be quickly re-configured between jobs is a more elegant solution,then trying to scale a fixed array device requiring a non quantum calculation.

We also need to standardize on a computing language rather than let the customer use their own legacy language as used in the Dwave model, so there is a common language for a open source type community to flourish.

I totally understand why you would initially let the customer decide how he uses a QC nobody this side of the black projects fully understands how a QC which is really an analogue computer can ever be made to produce a real world result. The safe bet is build a device based on the best understanding of how to achieve a machine with the properties of a theoretical QC and let the customer develop the software suitable to their field of expertise but I believe that strategy is not conducive to the evolution of QC as the the customers are only interested in their market niche and will not share their results with the world.

We need to solve the problems of building a real world QC device as a community from both sides of the divide from the theoretical highly expensive nano tech end and the IT world of the traditional internet object orientated software end like an arch bridge over a river eventually we’ll meet in the middle and a new quantum leap will literally take place in human history.

My dream if I ever attract some serious money to kick off this startup is to hire a Pure Math Director to fully develop the Unary Number Math, a Python Software Development Director and a FPGA Hardware Director to head up their own hand picked teams. Open source will be utilised whenever possible but some core IP may need to be kept trade secret to set aside as an asset to build the Corporation.

Qbitrex needs enthusiastic positive types who enjoy a challenge and will be given the freedom to follow their dream using whatever methods the can argue will get a result. Anyone interested in my ideas feel free to contact me.

Where to from here, well that’s a fair slab of my IP floating in the public domain breeze for anyone to pick over like like the desert bar at Sizzlers. I do have some aces up my sleeve but.

I’m waiting on my order for 25 N enhancement mode NMOS FET transistors to continue building my one bit register in my electronics kit. I blew up my IC flip flop and one Nand array IC by not tying the unused pins either high or low, yeah I know a rooky mistake but the silver lining in this cloud is I realised they build a lot of extra gates in these IC’s to minimise the Metastability condition I’m trying to exploit that are another potential source of complexity and potential data leakage so I’m building with basic transistors. MOS FET type are foundations building blocks of modern IC’s, BIPOLAR are not suitable as they require additional resistors that are another potential noise source. They also rely on current and therefore leak energy continuously compared to voltage controlled FETS which are voltage controlled resistors which are set and forget holding a miniscule charge like a tiny capacitor to turn on. The N-Type are also naturally off until a gate voltage above threshold turns them on.

My Lab
My Electronics Lab showing an RS latch built with Bipolar Transistors. It gives a random display of LED’s when turned off then on again

My plan is to keep going on my one and build the 18 transistor 1 Qubit Quantum register and clock it from the kits 4 bit assembler language computer. The QC circuit like the Dwave model has to be isolated from the the management microcomputer as the metastability I’m exploiting drifts across multiple clock domains if they are on the same chipset, causing bugs and glitches in the management software. Then scale up to a multi-register to have something to demonstrate to interested parties. Whether it shows some promise or just ends up a cryptographic grade true random number fire hose I’ll soon know!

Thanks for reading my blog.