Quantum computer

Who wants to build a quantum computer with me?

Other urls found in this thread:

discord.gg/zkPsa3
minecraft.gamepedia.com/Piston
discord.gg/Jbnm2Wf
twitter.com/NSFWRedditGif

el scientismo

How do you propose to build it?

I want to build at least a 2 qubit superconducting quantum processor in the next 5 years. I currently have a physics and EE major with me. We are undergrads, but we are going to grad school after we graduate.

I am looking for more people to join me.

With respect to how we are going to build it; We are planning on doing it the standard way. The way IBM, Rigetti and Google are doing it. We are going to fabricate our qubits on aluminum or niobium, set up the microwave controls that will drive the state of our qubits, and put our processor in a dilution fridge. Of course there is more to it than this, but I am willing to chat in depth about it if you are interested.

How are you going to get access to the fridge?

I'm a CS major, but how can I learn more about this user?

Every house has a fridge lol

That is one barrier to entry. The cheapest fridge costs $600k.

We are planning on going to graduate school for quantum computing research. There are some Physics departments that have dilution fridges.

Worst case scenario is finding a group that will let us rent time on their fridge.

Superconducting quantum processors need

just wondering but let's say you build one, what kind of experiments do you run in it?

There is a lot of resources online. It really depends on what sub-field of quantum computing you want to focus on (complexity theory, algorithms, hardware implementations e.t.c).

With respect to hardware, there is a lot of physical implementations for quantum computers. The leading ones are superconducting quantum computing and ion trap quantum computing. Superconducting quantum computers are currently winning because you can fabricate qubits using standard nanofabrication techniques.

if you cant build it in minecraft its not worth my time

The above post is for you, I accidentally deleted the last number on your id.

One of the simplest quantum algorithms you can run on a universal quantum computer is the bell state experiment.

Quantum computers are not meant to replace classical computers. They are meant to serve as co-processors to classical computers. Like how the GPU is a co-processor to the CPU.

using pistons

What?

>CS major
>build something tangible

>> 1 mm
That's a big computer
The way Britain is planning to do it is with a football field(european) full of hard vacuum, ultra cold bits that are actively aligned to micrometer tolerances. It is much more than just microwaves and fridges
No motherfucker, quantum computers are for simulating quantum physics. That's worth a lot of $$$$$$$$$ unlike vidyagaems. Finding new drugs, materials, and ways to make solar cells cheaper than what your mom charges for blowjobs is worth a lot of money.

where the fuck are we going to get money that we would need??

Getting money at first will be hard. We are planning on building one qubit and detecting it's rabbi oscillations. That will probably be enough to get us published in a low-tier journal.

We can use the credibility of being published for making a qubit to raise money from venture capitalists.

>No motherfucker, quantum computers are for simulating quantum physics. That's worth a lot of $$$$$$$$$ unlike vidyagaems. Finding new drugs, materials, and ways to make solar cells cheaper than what your mom charges for blowjobs is worth a lot of money.

I never said that quantum computers are GPUs. I said that they will be used like GPUs; in the sense that they will be co-processors to classical computers. And yes; one application for quantum computers will be for simulating quantum physics.

>The way Britain is planning to do it is with a football field(european) full of hard vacuum, ultra cold bits that are actively aligned to micrometer tolerances. It is much more than just microwaves and fridges

Yes. It is more than microwaves and fridges. It is why I said that another physical implementation for quantum computers is ion traps here:

and cracking cryptocurrency

>(((Rabbi oscillations)))

I meant rabi oscillations.

Yeah. Quantum computers can be used to break RSA; but not anytime soon. You need large scale fault tolerant quantum computers for that, which will take a couple of decades at least.

wrong, quantum computers are for mining bitcoin

I wanna build an iron man suit.

>that's a big computer

For you

For the people who are interested, how do I talk to you outside of this board?

from our parents

start a discord. I'm a soon-to-be ee who's interested

Here is the discord link for the people who want to join me: discord.gg/zkPsa3

I want a single dimensional mutli (10+) qubit computer. Dream big or go home

I've taken a course in quantum computing this semester with one of the inventor of quantum cryptography (i.e. Gilles Brassard). It was dope.

We have to start small because we don't have a lot of resources. It will be easier to get our project funded once we build our credibility by fabricating one qubit. We are dreaming big; but we also need to be realistic and have a plan to meet our goals. If you are dreaming big too, I invite you to join us.

i'm another user but he was probably talking about minecraft.gamepedia.com/Piston

Nice. What did he teach you any quantum algorithms?

I'm completing degrees in mechanical and nuclear engineering , if i can be of any assistance

It's different from binary right? Like there isn't just on and off there's several other states dependent upon the atoms rotation or something if I remember correctly. And It can be in multiple states at once too I believe. That about right? I'm really curious and fascinated!

As a material scientist I can tell you right now that the components you are going to need are either unobtainable or extremely high cost. Some of which you have to make yourself since you are going to need materials with extreme purity.
And how are you going to get the proper cooling to reach superconductivity? You gonna need liquid helium for that.

That post wasn't me (OP). I answered how we might get some resources here:

what uni are you at? i have some projects that will require a quantum computer

>We are planning on building one qubit and detecting it's rabbi oscillations. That will probably be enough to get us published in a low-tier journal.

I'm no experimentalist but isn't one qubit completely useless even as a proof of concept? You can't even demonstrate an entangling operation with it.

It's not completely useless. The point is to gain some credibility so that it will be easier to get our project funded. We are strapped for cash. Fabricating and entangling 2 qubits will cost more.

On the technical side, we will gain a lot of information from that qubit; such as decoherence times, and how to fabricate qubits (there is only one person in our group that knows how to do it).

Observing rabi oscillations on a qubit is the minimum of the minimum that we need to get started.

We would love to have you. Since this project is going to take years, we are looking for people who will be dedicated.

What projects are those?

Got Production Engineering BSc, getting Applied Physics degree now. I am down for it. How to contact you?

You won't publish Rabi oscillations in any journal unless the qubit is pretty fucking novel. Since you've basically said you want to make transmons, you're definitely gonna get laughed out of even the lowest tier reputable journals unless you somehow manage to make a huge improvement in coherence or energy lifetimes. Better route is to join a QC group, actually learn the the fuck it is people are doing, and then maybe consider trying to get venture captial if you've established yourself as a successful scientist. That's my plan. Even if dots don't pan out I'm sure I could get some kind of job at IBM, Google or Microsoft. If I'm desperate I'll go the startup route.

You are right. We are planning on joining a QC group. Another option is going the startup route; but we are not sure if that path will lead to a greater chance of success than going to graduate school.

I created a team on discord for the people who are serious about doing this: discord.gg/Jbnm2Wf

I don't think the startup route is viable unless you have some novel idea for qubits that will btfo all the big companies pouring in money into research.

We don't need to have a novel idea for a qubit to get VC money. "Uber for Dogs" and other similar products get millions of dollars of funding every year. All we need to do is convince a couple of angel investors or capital firms that we are capable of building and scaling a quantum computer.

This is not a fair comparison because the founder of this company did research at Yale and worked on quantum computers at IBM before he started his company; but Chad Rigetti was able to start a quantum computing company from 2.5 million dollars of VC money. They released a 19 qubit quantum computer to the public yesterday. My point is that startups have a shot at competing with the likes of Google and IBM.

Well, I won't pretend I know anything about business. I'm very skeptical some undergrads can tackle a problem like this feasibly though, since decoherence and the related issues aren't anything to sneeze at.

We are aware of how difficult this task is. It is not something we woke up a couple of days ago and decided to do. We have been working on this for the past year or so.

We also have someone who does undergrad research on superconducting quantum computers at Cornell. He taught us a lot in the early months of this project. We are now confident that we know the field pretty well.

That's not really an algorithm. Still kind of interesting though.

Hey, do you need someone with the knowledge of RSFQ circuit design?

You are right. I shouldn't have called it an algorithm. You can implement the Deutsch-Jozsa algorithm using 2 qubits.

Why do you want to build your own when you can use IBM's?

Yes we do. We want to build a superconducting quantum computer. Your knowledge of RSFQ will be very valuable.

Sure. Do you have the hundreds of millions bucks required to start it off?

I've made a couple of posts in this thread about how we might fund our project. Specifically this one:

Get a physics degree. In a nutshell, the quantum computer is just like a Turing Machine, except the value on the tape is actually a superposition of values (qbit) instead of one single discrete value(bit), every time you move the head to a cell on the tape, the value on such tape assumes a single value, meanwhile a lot of other stuff is being computed by the non-observed qbits on the other cells.
Now, if you think this in terms of digital systems, you just replace logical gates with quantum gates, instead of operation using bits, we use matrixes, quaternions and so on and so on.

Do you think that superconducting qubits are currently the way to go? Or would it be feasible to switch to another implementation if it proves to be more scalable?

Yes. Superconducting qubits have made tremendous progress in the past decade; and that progress hasn't stopped yet. Switching to another implementation requires an insane amount of resources. The knowledge that you need to have for one implementation is a lot. This makes it difficult to switch to a new one.

We want to have something to show within the next couple of years. Switching to a different implementation won't help with that.

>Superconducting qubits have made tremendous progress in the past decade; and that progress hasn't stopped yet.
Afaik, current superconduction qubits are just too fat for chip designs to get into the number of qubit regimes where you can actually perform useful error correction and computations with them. Also other engineering related problems are predicted to impact the progress sooner rather than later. The big players are increasingly running into these problems now. I am not sure what the exact aim of your startup is, but from my point of view one should be careful with extrapolating the current growth of superconducting qubits. Of course, when planning for the next few years you should not be getting close to having these problems in this time.

I agree. Trying to predict future progress is dangerous because the predictions are usually wrong.

>Afaik, current superconduction qubits are just too fat for chip designs to get into the number of qubit regimes where you can actually perform useful error correction and computations with them.

Superconducting qubits have recently reached a point where you can perform error correcting codes on them; like the surface code. This review article goes into depth about it: "Building logical qubits in a superconducting quantum computing system" (I can't post the link to it because this site thinks it's spam.)

It depends on what you mean by "useful computations". Superconducting qubits can now perform some standard quantum algorithms like grover and quantum simulation. They are not "useful" because you can still simulate these computers on classical computers. You probably know that there is currently no quantum computer that can perform a task that cannot be performed on a classical computer. I'm assuming that is what you meant by "useful computations". That is true; but in the next couple of months, Google is planning on doing a quantum supremacy experiment.

>Of course, when planning for the next few years you should not be getting close to having these problems in this time.

True. All we want right now is a very basic "working system". The research on that for superconducting qubits is pretty much done on that end as there are currently ~20 qubit systems out there. What our group is working on is an engineering problem, not a research one.

Because it's a challenging problem. It is a combination of the fields that I love; physics and computation. We think that quantum computers will be useful tools for solving some problems that are intractable for classical computers in the near-future. We want to be a part of this computing revolution.

We do use IBM's. I learned how to program quantum computers using their SDK. We are not ignoring quantum software.

Those machines cannot create entangled states, making a large number of quantum algorithms impossible. Other methods are being researched because the superconducting rings are seen a s a dead end. Consider going to the SQuInT conference in NM on Feb 21 to learn where the industry is at.

If your group did figure out how to create the computer, enjoy your nobel prizes.

Let's build a quantum dildo, mah nikka

>Consider going to the SQuInT conference in NM on Feb 21 to learn where the industry is at.

I am going to SQuInT next year. Are you going to be there?

>Those machines cannot create entangled states

Can you go into depth about what you mean by this? Superconducting quantum computers can perform Grover's algorithm; which needs two-qubit gates.

>If your group did figure out how to create the computer, enjoy your nobel prizes.

They don't give out Nobel prizes for things that have already been discovered. Universal superconducting quantum computers have existed for at least the past 10 years or so. In fact, Jerry Chow's PHD thesis from 2009 is about a 2 qubit universal quantum computer that can perform the Deutsch-Jozsa and Grover's search algorithm.

>Those machines cannot create entangled states, making a large number of quantum algorithms impossible.
Pretty much all quantum algorithms require entanglement, otherwise it would basically just be independently controlling a bunch of qubits.

What exactly do you mean?

sorry user, would love to help, but I gotta build a parallel computing light-wave processor to prove p=np. gotta flash