Pyquil Easy, Hybrid Quantum Programming | Scipy 2018 | Guenevere Prawiroatmodjo

So helium is indeed a finite resource just go all the way back so here is so we actually don’t use up the helium we actually recycle it we recover it so I don’t know if you can.

Picture but there are these cans here in the back these gray cans are cylinders and all these tubes coming out here they’re actually a closed cycle so it’s a.

Closed cycle dilation refrigerator so what we do is we have a.

Mixture of helium 3 and helium 4 two different isotopes of helium helium four is pretty abundant it’s what you use in party balloons.

And so on helium 3 however is pretty rare and it’s a byproduct of oil drilling actually and that is our pressure dry so that’s that’s a material recycle so we pump it.

Through our system continuously and it does happen sometimes that we like lose some helium-3 and yeah currently it’s just very expensive and in terms of sustainability there have been.

Some theories that there is helium-3 on the moon that could be mined so who knows I might go I have a helium mine on the moon hey I.

Just a very interesting I don’t know anything about this I was just curious is there any sense of how the energy.

Demand of the computer would increases you have more and more things entangled does that modeled yes that’s a great question you mean energy to men in terms of the usage of the quantum computer or you mean in useful and this is curious if it.

To does it take more energy to have more entangled things or is that not it does yeah so indeed having more qubits takes more cooling power so we would need a larger fridge so I should see this this brass plate over here it’s about this size and that can cool down a chip that’s about this size so if we wouldn’t would increase the number of qubits we would need to increase the cooling power one.

Other problem with quantum states is that they decohere so naturally nature wants to be wants to default to the ground state so this means that qubits when they are any excited state they naturally want to relax into the ground state and this is correct rise.

By the decoherence time or the relaxation time which is currently on the scale of about 50 microseconds to 100 microseconds and then there’s another there’s a D phasing time which is the time in which the cubed loses its phase information which is even shorter which is.

About 20 microseconds so if you want to entangle a lot of qubits you have to do that within the time that the quantum bits are going to retain their quantum information and that is currently the bottleneck so it’s not necessarily energy but it’s more the quality of the qubits that limits how many qubits you can entangle but the assumption is.

That will be you’ll be able to decrease that just continuously sorry like I mean so that but that’s that bound is not it’s.

Something you you expect to be pushing down more and.

More yeah definitely that’s something something we’re pushing on right now we’re trying to extend these coherence times as as.

Much as possible yeah thank you for the presentation this is really interesting I’ve recently just graduated from undergrad computer science I’m going on to get my masters and I’m curious as to how someone in higher education can continue to learn about things such as quantum computing cuz I’m not sure whether or not these.

Types of like I don’t think there’s a quantum computing course yet or anything like that.

So where could I go to find resources on how to learn about this that’s a really great question currently actually a lot of people at Righetti we’re not all quantum physicists so there’s actually a lot of computer scientists and people with traditional software engineering backgrounds working at Righetti one resource I would recommend is our doc spike well don’t read the docs I oh there’s a.

Pretty nice comprehensive overview of introduction to quantum computing you can start playing around the examples you can actually.

Start quantum programming yourself so there’s this binder that I already mentioned go to Righetti dot-com such quantum dice then there’s also go to Riga datacom / Pikeville.

Github account see that would be my main advice is to just get your hands dirty and start coding your own quantum programs started includes also some some really physics II examples such as this cubed relaxation that I mentioned where you can actually see.

The cubed relax into the ground state there’s also some really good books one.

Book quantum computation by qatayef I can really recommend or Mike and Ike as it’s called by Mike and Ike Isaac Shuang and Michael Nielsen quantum computation and quantum information Abby I’m happy to follow up with you after.

A talk and give you some more reading material so Ken or even how can they or no computers be used commercially for like the general public and for businesses yes that’s right yes so currently our platform is usable for free so you can just get your API key and you can use our simulator for free if you want to use our qpu you fill out a form on our website to request access and yeah you can be a business it’s for academics currently sorry currently most most of.

Our users are academics but we also work with companies we recently renovated our website so I’m not really sure where everything is are there any more questions hi thank you first hug as you mentioned there are physics limits to the qubits like they have declarants time right so.

Are there any constraints when I write a program for these content computers anything I need to pay attention to yeah that’s a great question actually one thing that a lot of people run into when they first start writing quantum programs is the is the topology of her chips so quantum information theory.

Sort of assumes that all qubits are connected.

To each other so also if you’re going to run programs on our qvm or quantum virtual machine or simulator you’ll see that you can just connect any qubit with any qubit so a lot of for example.

Shor’s algorithm or Grover’s search algorithm they all assume all – all connectivity while in practice that’s actually not the case so if you look at our qubit topology here which is in the docs you see that it’s a ring so so there’s qubit zero it’s connected qubit one q1 is connected to Cuba two etc so to simulate this all-to-all connectivity you have to do something called a swap operation where you swap the quantum states and do sort of like a jigsaw puzzle of quantum states and this comes with a cost.

These swaps which takes time which in which the qubits can decohere so yeah that’s definitely yeah so if you’re if you’re strong in graph theory this is very.

Much an unsolved issue so so thank you for your nice talk and thank the entire set of speakers who give excellent talks we have to break now for lunch so if you have any.

Further questions I encourage you to meet the speakers directly thank you you.