Dan: Hey there and welcome
back to the quantum divide.

Uh, this week we're talking to Dr.

Mirella..

Koleva.

And, uh, she's from contact
to con she's the CEO.

They developed software for
simulating quantum photonic devices.

And specializing quantum
modeling, quantum systems.

From first principles.

So we're going to dive into
really the physics of it here.

And I hope you enjoy it.

It was a great conversation.

Let's go for it.

Mirella, thank you very much for
joining us on the Quantum Divide.

Why don't you start with a bit of a
background on your story, your path into

quantum, and um, we'll go from there.

Mirella: Yes, of course.

So I have a background in physics.

I studied for my master's degree
at Imperial College London.

And after I completed that degree,
I became interested in exploring the

crossover between the physical and life
sciences to learn how physics can be

applied to other scientific domains
like biology and medicine to improve

our fundamental understanding of the
biochemical processes that occur within

living organisms with a view to opening
the door to new pharmaceuticals and.

better treatments.

So I have a bit, I had a very different
focus back then, and I was really

fascinated by that field at the time,
and I decided to do a PhD on it.

This was followed by two postdocs in
the same field, one at the University of

Oxford and one at King's College London.

And while postdoc, my mother,
who's a theoretical physicist,

well known in her field, shared her
latest research with me and stated

that she had developed a unique.

Theoretical model capable of faithfully
simulating quantum optical devices.

And this was back in 2013, just after
the then Chancellor of the Exchequer

had announced that the British
government had was to invest 270 million

pounds into quantum technologies.

So I was very intrigued by what she had
accomplished and quickly realized that

her method could soon become extremely
relevant and play a major support role in

the building of this whole new industry.

and that we could turn
this into a business.

However, there was one snag her
implementation of the method into

code at that point was very rigid.

It was meant to simulate only one
quantum system of a specific kind.

So in my spare time on evenings and
weekends, I set about to generalize her

algorithm to arbitrary quantum systems.

It took a while, but by summer 2016, the
entire computer program was finished.

And we founded Quantopticon
shortly after, in early 2017.

At the beginning we used it essentially
as a side hustle for a couple of years

to productize a theoretical method
into a commercial piece of software.

And by the time I was nearing the
end of my contract at the latter

postdoc, I decided that academia
was not a good career choice.

And I dedicated myself fully
to starting up Quantopticon.

Dan: Yeah, it's a really
fascinating um, story you've got.

Especially the connection
with your mother.

I mean, does, Does physics go back
more generations than that, or

perhaps broader than that with her
siblings or other family, or is it,

is it just in your little family

Mirella: Yeah well, my
father is also a physicist.

He has an undergraduate degree in
physics and a PhD in biophysics.

And my uncle is a physicist
too at the undergraduate level.

And he also helped with the
development of the software.

So it was very much a family business.

Dan: Yeah.

Real family affair.

Brilliant.

Well, I think that's very
unique in the industry for sure.

I haven't heard of any type of
similar, like siblings working

together or anything that I can
think of off the top of my head.

So tell me about Quantopticon and
Quantillion, I've also seen them.

Is that the name of your product
as opposed to the company?

Give me an overview of the team and the
vision and everything that would be good.

Because I guess that's the end
result of all of that hard work that

you've just described and the slow
building of the finalized product.

Mirella: so at QuantOpticon,
we are making the world's first

software platform for the design and
optimization of quantum optical devices.

And by quantum optical devices,
I mean devices that are driven

and manipulated with light.

These are the bread and butter components
of quantum computers, quantum networks,

interconnects, quantum sensors, such
as single and entangled photon pair

sources, qubit registers, quantum
memories, as well as all the remaining

passive components that you would find
in a integrated circuit like waveguide

splitters, which is couplers and so on.

At the core of the software is the
now patented in the UK algorithm that

I mentioned that my mother conceived.

It describes the quantum physics of light
matter interactions unfolding within

the device with very high precision
and granularity in both space and time.

And that is what makes it possible
for us to predict how the device

is going to behave on the basis
of the particular materials it's

made of, its size, geometry.

The light pulse, the pulses that
we're feeding it, and so on.

And using parallelized simulations
where we can vary each of these

parameters, we are then able to find
the particular set that yields the

most desirable operation of the device.

For example, if we're talking about single
photon sources, we would want the single

photon source to have the highest possible
purity of emission of single photons.

That is to minimize, minimize
bunched emission of photons.

And a little bit about the team there
are six of us at present Gabby and

I are at the helm of the company.

I'm the CEO and she's the
chief scientific officer.

We have a part time IT manager who
takes care of hardware and software

installations, security and updates on
the numerous work computer workstations

that we use to run our simulations.

We have a PhD student at University
of Graz called Felix Hitzelhammer.

who is helping us to quantize the
way we treat light in our model.

Currently we're treating matter
quantumly and light classically.

So our model is actually
semi classical at this point.

And the goal is to transition to a
fully quantum approach in order to

make sure we're taking into account
all quantum phenomena that may arise

and hence to improve, improve our
predictive capabilities even further.

A month ago we were joined by
a master student in engineering

at the University of Bristol.

Her name is Kriti Goel, and she's working
on the problem of implementing a micro

ring geometry into our model in order to
meet the modeling requirements of one of

our de facto customers that we're working
with through through an Innovate UK grant.

And uh, just this week, on Monday,
we acquired Andrew Lingenfelter,

a very talented PhD student at the
University of Chicago who has won

several prizes and scholarships.

And we're extremely excited to have him.

He will be working closely with Gabby
in developing a model for spontaneous

forward mixing for probabilistic
entangled photon pair generation

within the framework of our theory
which again is one of the objectives

of our current Innovate UK grant.

Dan: Thank you.

Yeah.

So It's good to see the growth that
you've got happening and great to

hear that news about the different
students you've got joining you.

I guess I'm just thinking back to the
rest of my podcast episodes, I've had a

couple of people who are involved in nano
fabrication and some who were involved in

quantum dot type technology, for example,
but there that's the very physical world.

Really your kind of unique position in
the industry is around world, right?

Virtualizing all of that and all of
the mathematics and, using mathematics

to model the behavior of how all of
the different atomic and photonic

particles interact with each other.

It's it's fascinating.

You spoke there about working
with Innovate UK and the number of

people you've got working for you.

It like you're kind of
startup scale up type.

Um, I don't know whether you've
had a seed or whether you've had

a series A yet or whether that's
the journey you're going on.

I'd love to know about the, you
know, your through the different

accelerators and funding rounds.

And you mentioned a couple
of grant applications there.

I just, I heard a new word earlier on.

I was reading an article that
said some VCs don't like startups

that become grantrepreneurs.

So I I guess that means somebody that's
just applying for grants and trying to

survive that way, but it doesn't sound
like that's the case with you, but

yeah, keen, keen to know what you can
share about that part of your journey,

Mirella: Yeah, I think the case is
It's very much still the case that

a lot of the quantum startups are
relying heavily on grant funding.

And yeah the field is still emergent.

There are a lot of engineering
challenges to be overcome and that

requires money to get to the point
where there is commercial interest.

So we have been through
a lot of accelerators.

First we started with the Kings.

King's 20 which is based at King's
College, London, where I was a postdoc.

We then went on we went
through the duality accelerator

at University of Chicago.

And this is why I'm actually
based in Chicago today.

It's all down to the the duality
accelerator and the fantastic

conditions that they provided to us.

We have access to free office space
state of the art office space.

We have networking connections with
local partners, with stakeholders in the

government in big industry corporations.

So they have been very key in
our survival and help, they have,

they've helped us really to thrive.

We have raised a pre
seed round of 100, 000.

dollars.

That was a couple of years ago,
and that was from the Luminate

accelerator, which we also went through.

Luminate is based at Rochester, New York.

And we have been recipients of many
different types of grants and awards

and uh, yeah from all sorts of sources,
including the European Union, from the

European Space Agency, from Innovate
UK some private donors in the US

some public money through Illuminate
Accelerator as well from the US.

Really a variety of different sources.

Dan: yeah, just goes to show the
um, all the different breadth

and uh, wide different sources
that you need at that level,

I guess.

so yeah, in terms of your company being
a software orientated company, I wouldn't

say that would shield you from any
risk, but it does put you in a different

place to a lot of quantum technology
companies that are building hardware.

Obviously, It's much more expensive
and difficult to build a whole

quantum computer, but there are lots
of small component manufacturers.

What are your thoughts on I guess the
differences between being a hardware

and a software company in the quantum
industry at this point in time?

Mirella: We are very unique because we
are the only ones doing what we're doing.

And it's because we have a,
developed something so advanced

and complex and complicated that.

It's difficult for others
to enter this field.

There's a big barrier to
entry as it were, and we have

already patented our algorithm.

In that sense, we are,
yeah, we're very different.

Yeah, they're not really that
many software companies that are

yeah, doing modeling as we are.

Most of the other software
companies are doing algorithms.

Developing algorithms for applications,
really, of quantum computers to to

business problems, whereas we're dealing
with simulations of physical systems

and trying to design and optimize them.

So it's a different kind
of simulation and software

development work that we're doing.

And, yeah, so we can't really
compare ourselves directly

with those software companies.

. Dan: Very interesting.

I really want to know a bit
more about your product.

I read and obviously you described their
modeling light atom interactions And

to be brutally honest, you know, as an
person that's not a trained physicist.

sometimes I don't quite follow,
but I'm going to call those out as

we go through and you can put me

Mirella: Sure.

Dan: But yeah, so it's, it's niche kind
of prototype modeling and, you know, what

are the benefits that, The customers get
and what kind of customers do you have?

Does it come down to anybody built
this building hardware first of all?

And does it go beyond that?

Mirella: Yeah.

So what we are offering really is like
performing virtual prototyping in silico.

So we scrapped the need for
quantum optical scientists and

engineers to conduct trial and
error experiments in the lab to

establish what works and what doesn't.

And we replaced those experiments with
accurate parallelized simulations.

This naturally, this leads to
enormous cost and time savings on

the order of 10 times for both.

Meaning that you could be done designing
and optimizing a device within a two week

period instead of six, a six month period.

This allows you to make better use
of your time and in the process

save hundreds of thousands of pounds
that you would otherwise spend

on equipment, time, consumables,
labor, and other running costs.

And all you need to run the simulations
on our software is an average spec tower

workstation, preferably multi core.

so we can model we can produce
models down to the Angstrom

femtosecond resolution.

In terms of the limits to our simulations,
yes, we cannot really we cannot model very

large structures at at the moment, but
yeah, very large complicated structures

will take quite a long time to simulate.

As we're employing uh, the FTDD
method in our calculations,

which is computation intensive.

But having said that, in future
we envisage employing something

called an adaptive step algorithm
to speed the simulations along

as well as harnessing GPUs.

So at this as a software stands
at the moment, as a rule of thumb,

it is able to model structures
of up to a few microns in length.

In terms of where we stand
in the design stack, we are

addressing the physical layer.

So we are working on the device,
not system level at present.

Eventually as we continue to build
out our software and make it faster

and more powerful, we will get to a
point where we'll be able to model

entire quantum photonic integrated
circuits which is a big aim of ours.

And they will have, they will be complete
with all the necessary elements, You'd

find in, for example, a repeater node,
and in fact we have already received

commercial inquiries about this.

And that leads us to our
customers and collaborators.

They can be categorized
into three different kinds.

First, we have the academic researchers.

So we team up with PIs and the research
groups from various universities around

the world in order to verify and validate
our model as we develop new aspects of it.

Of course need to verify that the model is
reflecting reality and that we're not just

making up some theory about how the world
works without any justification for it.

So, We team up with these university
professors and their groups to

carry out the experiments that that
demonstrate that our model is correct.

These are experts in their own niche,
for example Professor Kai Muller.

At Technical University of Munich is
a longstanding collaborator of ours

and he's famous for, he's famous in
scientific circles as being an expert

at etching micro pillar structures
with very high quality factor.

The second kind of collaborators
we have and customers,

Dan: Can I

Mirella: of course,

Dan: what's a micro pillar?

Mirella: A micro pillar is a very
tiny optical structure on the, on the

scale of, Nanometers to micrometers
that is composed of reflective layers

with different refractive indices.

And it's uh, it is essentially
a, an optical cavity that

confines light inside of it.

And it um, it helps to control the
emission from a quantum dot or some kind

of quantum system embedded in it and
to obtain better optical properties.

From that quantum dot for whatever
purpose that you're using it.

Dan: Do they work together with cavities?

Or is it like one or the other depending
on the architecture you've got?

Mirella: Yeah, so normally you would uh,
pair, you would embed a quantum system

within a cavity, within an optical cavity,
in order to control its properties.

You could also have a quantum system on
its own, but it's you would probably have

a very low collection efficiency and yeah,
and the properties won't be as desirable.

as if it were actually embedded
in an optical cavity, which is why

people do this sort of embedding.

They like to embed, embed uh, in cavities.

And cavities can be of
different geometries.

They can be made of different
materials and they can be, they

can serve different purposes as
well for different applications.

Dan: But the micro pillar, is that
something that goes with the cavity?

Mirella: Oh, the micropillar
is itself a cavity.

Dan: Got it.

Thank you.

Mirella: a micropillar

Dan: I'm very good at
asking stupid questions.

Mirella: Yeah, no problem.

There's a lot of jargon in this field.

No problem at all.

Dan: yeah, I'm getting there.

I'm getting there.

Okay.

Sorry.

I

Mirella: Yeah, no problem.

So, The second type of
collaborators and customers that

we have is quantum startups.

So we engage with companies making
components for quantum networking

like, Ki3 Photonics, who are
based in Montreal in Canada.

And this is all happening through
grants at the moment as a market

pool is strong at this early stage.

And the third kind is, There's
multinational corporations coming

from the aviation, telecom, and
datacom industry sectors mainly.

Although there's, we also are relevant
to the security and defense industries.

So these uh, corporations usually
have R& D divisions that lead

active efforts mainly on the
quantum networking side of things.

Trying to build robust,
efficient components.

They're at the stage where they're
exploring different physical platforms,

modalities, protocols, and materials.

And so it's quite an early stage and
our software comes in very handy for

this as they can try out many different
combinations in a very short period of

time and home in on the best solution.

Dan: Cool.

So you mentioned.

A limit to the scale that you
can model probably because of the

computational complexity that's involved.

The larger you get, the more
molecules, the more photons or the

more interactions you've got going on.

And you mentioned networking.

type use cases.

This is super interesting for me.

If you wanted to model a quantum
memory and a transducer and the

link off to a telecoms fiber, for
example, that would be too much to

model at this point in time, I guess.

Mirella: yes, that's right.

I've been told by Gabby that
we cannot model fibers yet.

And that is something that, that is a
capability that we'll develop in time.

But we are doing our utmost to speed
up our code as much as possible so

that we can model bigger things.

Dan: Yeah, that's great.

So at this point in time, then I
guess it would be the, if there's a

comb or something in the transducer
that's doing the taking the energy or

the state from photonic to microwave,
or if it's something else, then you

could model maybe that interaction.

And that would inform the manufacturing
and the optimization of that individual

Mirella: exactly.

So we can uh, identify how we can
make this transduction efficiently

from one, convert from one type of
photon into another type of photon of

a different wavelength, for example.

And minimize photon losses as well,
like insertion losses and, in general,

end to end losses because every
time you put in an interconnect or a

transducer in the network, it will lead
to some loss of photons, and we talk,

when you're talking about only a few
photons, every photon counts, so you

want to preserve of them as many of
them as possible, and we can do that.

Dan: Yeah.

And they're all of the
systems you model photonic

Mirella: yes,

Dan: or any, any just atomic
type structures as well.

Mirella: we can we can model
both matter and photonic qubits.

On the matter side, it's any
kind of solid state qubit.

It could be a, a defect center
embedded in our host matrix like

uh, diamond or silicon carbide.

It could be a quantum dot.

It could be spins silicon spins atoms.

And principle, we can model ions as
well, but we're focusing on quantum

dots at the moment because we think that
they are the most promising physical

platform because they are deterministic
sources and they are very bright.

Uh, And we have also figured out a
way to make them all emit homogeneous

photons, which has been a problem in
this industry for quite a long time.

Dan: Okay, so I'm going
to ask a question now.

It's probably going to I'm not
sure if it's going to make sense

or not, but I'll give it a go.

You're modelling a quantum dot, as
far as I understand, depending on

the type of dot, it has different
structures around it to either lock

it in place or control it in some way.

There are Bragg structures
sometimes, which are the layers

of material above and below.

Do you model all of that in order to
model the quantum dot, or is it purely

the mathematical kind of behavior of
having those structures there, which is

Mirella: Yes, exactly.

So we model all the material around
it, which forms the photonic cavity.

So those bragg reflection
layers that I was talking about.

But we also model the discreteness of
the energy levels of the quantum dot,

because we, because it's a quantum
system, so it's got discretized energy

levels, and we fully take that into
account which is very important to be

able to really understand how it behaves.

Dan: Yeah, and different qubit systems
have different behaviors, don't they,

in terms of their excitation levels.

And sometimes there are
Intermediate levels and

Mirella: Yes, that's right.

Dan: so every single qubit has to
be modeled into your, your software,

Mirella: yes, so we have to, you have
to know which levels you're exciting

and what energies they have and we can
as far as I'm aware, we can work out

the best excitation scheme for you.

If you have a selection of
different levels that are present

in the quantum system that you're
considering using, we can tell you

which one is the most efficient.

But yeah, there's a certain number
of input parameters that you have

to put into the software in order
for it to make the calculations.

And you can find these from either from
papers in the literature or you can

conduct experiments to measure them.

And yeah, there's a handful of these, but
uh, the rest we can calculate ourselves.

Dan: Yeah.

That's really interesting.

I always thought it would be the ground
state and up would always be the best.

Place to go, but ultimately, I
guess that depends on the system.

And if you want the ground state, you've
got to really put the atomic structure

and uh, Very high degree of control
to get it down to that point, right?

Yeah, I mean, could you just
elaborate on that a bit with

different states of excitation?

How high would you go?

Is it am I thinking about it the
right way when I think about height?

think of it like a ladder
of levels inside the the well,

Mirella: there's some levels that are.

more beneficial than others.

And also it depends on what kind of laser
wavelengths you have available, of course.

So you're constrained by that.

And uh, yeah it depends on whether you're
exciting resonantly or off resonantly.

And uh, I really am not the most
appropriate person to speak about this

because I'm not so deeply knowledgeable.

And Gabby is really the right person
to talk about this in more detail.

Yeah.

Dan: No, that's useful.

Thanks.

Mirella: I hope it helps somewhat.

Dan: Yeah, it does.

Of course, you're reminding me that
it depends on the laser, and that's

because the laser, the energy of
the laser has to be tuned to the

system that you're exciting, right?

And if you don't have the hardware to
go, maybe there's thresholds or kind of

guardrails on where you can go based on
the hardware you have is what I'm hearing.

Mirella: Yeah, and it could be
single photon excitation or two

photon excitation, and that will
have a consequence on the type

of signal that you get out of it.

For example, four wave mixing is,
you know, you need two pump photons

to get your signal and idler.

It's things like that
you have to consider.

Dan: Can I just ask about modeling the
interactions between light and atoms?

So, I'm imagining interference
patterns, and of the light that is,

I was going to say the wave functions of
the photons, but it sounds like you're

modeling them classically at the moment,
so maybe you don't look at them that

way but then there's pulses of light
as they go into the and things like

Mirella: Mm hmm.

Dan: Does that sound about right?

Mirella: Yeah, Yeah.

Yeah, that's right.

So we model the photons using
Maxwell's electromagnetic equations.

They are continuous wave equations.

Currently we do not really consider
the quantum properties of light in

our simulations, but we're working
very hard on that, and we hope

that we'll be able to breakthrough
very soon that we can publish.

And hopefully our PhD student from
UChicago will be a key, will be

instrumental in getting us there.

Dan: Now it's exciting to hear because
then you're moving purely into the

quantum domain then, aren't you?

And of course with photonic computers
and the way that different modalities

are stimulated using photons and
lasers to either hold them in place

or encode information and perform

Mirella: Yeah, you have
to start from somewhere.

We know that our model is not perfect,
but we are continuously improving it.

We are making it more and more quantum.

Currently we have we have
completed the 1D model.

We can model in a one spatial dimension,
but we have an interface for the one

spatial dimension part of the software,
but we also have developed a two

dimensional model, and we're working
on a 3D model, which obviously will

be the most desirable because we can
fully take into account asymmetries

that inevitably occur in real life.

So yeah, with a 2D model
we can model a 3D model.

rotationally symmetric structures.

We're confined to that at the moment.

But yeah, we have a roadmap where
we are implementing improvements on

the model to make it more realistic.

Dan: you got to have that, yeah, as a
small company and also as a software

company, you've got to have a short,
medium, long term vision and plan

of where your software is going.

So that's great to hear.

So I think we covered the customers
and collaborators primarily.

So yeah, I had this question,
which really goes to the more

fundamental physics of photonics.

Yeah.

It's a bit of a tangent, isn't
it, around combining, this is a

network engineer speaking, right?

Combining QKD and network
traffic on the same fibre pair.

Let me explain the context a little bit.

So all of the QKD systems out there at
the moment need to have a dark fiber

or a dedicated wavelength or something.

And there's interesting scaling issues
and cost issues in building QKD networks.

And at the, down to the world that you
operate in, the very small nano type

structure and the physics involved there's
In order for QKD and traditional network

traffic to go on the same fiber, there
has to be some kind of level of control

around emitting photons, either individual
ones to be used for quantum measurements,

and the classical type of photonic use
for network traffic, which is, I don't

know, should we call it a bit messier?

It's, you know, where We're too
bothered about what's happening

at the individual photonic level.

We're just encoding information
in the pulses or something.

Mirella: Yes.

Dan: any thoughts on that?

Mirella: Yeah.

Um, I think there are a number of
challenges that need to be overcome in

order to succeed in propagating classical
and quantum signals down the same channel.

And yeah, obviously when you're
talking about individual or a handful

of photons, optical loss becomes a
very major factor of consideration.

As far as I'm aware, fused silica fiber,
a telecom wavelength has been very

popular in classical telecommunications.

That is, it is an excellent
conduit for this purpose.

But this also means that due to its
success, there hasn't been much effort in

developing low loss fiber with with some
exceptions, like ultra low loss fiber,

which has an attenuation rate of about 0.

142 dBs per kilometer.

And I think ZB lan fluoride ultra low loss
fibers for repeater free communications.

But as I mentioned earlier,
photons can also be lost at

interconnects and optical couplers.

So insertion loss will have to be
lowered by a factor of of about 10 which

would also extend, this would permit
to extend the quantum repeater spacing

requirements and make it possible to
actually house them at the same sites

as inline, inline amplifier sites, along
with the classical amplifiers, which is an

important practical consideration for the
commercial realization of these networks.

It makes sense from a a cost perspective
and it's important to network carriers.

Um, And then there's also the question
of uh, the matter of channel degradation

factors effects like polarization,
modal dispersion, polarization dependent

loss, fluctuations in the state of the
polarization yeah, Riemann scattering

and nonlinear optical effects and.

Companies like Qunnect are
really prominent in this area.

They're making big strides in this
direction, having developed networking

components that compensate for the
optical fibers polarization drift.

So I think steps are being
made in that direction.

It's just, it's a complicated
engineering challenge.

Dan: Oh yeah from so
many different angles.

Actually I had an interview
with Noel the other day.

I haven't met her, but I had an interview
with her and she was talking about her

compensation scheme that they have acting
a bit like noise cancelling headphones,

which I quite like the analogy for that.

But yeah, it all comes
down to noise, doesn't it?

If you want to send quantum state over
a fibre, of whether it, on its own

is one thing alone, but if you start
mixing it with other sources of light

and higher likelihood of loss, then
it's just going to, the performance

is going to get worse and worse.

So, And I think another thing about
that is it that the adoption of

technologies like QKD is really
important because if the adoption

isn't there significantly then why
develop the technology to share fibers?

Okay, cool.

I said, I wanted to ask you what
papers you've you've worked on

that you'd like to highlight.

I'm quite keen to hear about that because
you've got an academic background.

You must have a, a massive profile
on Google scholar, I'm sure with

a plethora of papers, but is there
any, any you'd like to call out?

Mirella: You have to remember
that I come from a different

background, from a different field.

Having worked with biophotonics
for many years and then

suddenly jumped into quantum.

So I don't really have the uh, the track
record of publications, but I'm still

interested in contributing where I can.

And as I mentioned, I developed the
a portion of the software that, that

generalized the application of the
software, made it much more widely

applicable to different types of problems.

I'm very much interested in furthering
that and helping to make it more relevant

to problems that people are experiencing.

As a CEO, I don't really get
the time anymore to be properly

involved in scholarly work.

So I only have to jump in
when I'm absolutely required.

And uh, the last paper that I
was involved on was actually a

paper that Gabby presented at the.

Photonics for qu for the, at the
Photonics for Quantum Conference in

Rochester, New York a couple of years ago.

It was entitled single Photon
Generation through Cavity STIRUP in

a Neutral Quantum Dot, embedded in
a micro LAR cavity FTDD model study.

So that's quite a mouthful.

But this is a really exciting paper
in my opinion because we actually

showcased a novel methodology
which we dub programmable photons.

that can help to elicit photons
with the same properties

from dissimilar quantum dots.

And essentially what this means
is that we can control the shape

of the output pulse from a quantum
dot and make it whatever we want.

And this is an important breakthrough
because quantum dots are bright,

efficient, deterministic source,
sources of photons, which is

exactly the kind that the industry
wants to eventually move towards.

but their popularity has thus far been
impeded by issues related to scalability.

So it's really difficult to make
quantum dots emit in the same

photons with the same properties.

And with this paper, we show that
this problem is surmountable and

that quantum dots really are a great
commercially viable candidate for

future quantum networking componentry.

So yeah, this was the last paper
I was involved in two years ago.

Dan: One term that jumps out to me
there is the shape of the pulses.

What does that mean?

I always thought a pulse was just
a flat thing or a, it's a period of

time at which photons are emitted.

Mirella: Yeah, we can control.

Dan: in physics terms, the shape is the,
is it a mathematical representation of the

pulse or is it a, is something physical?

Mirella: yeah, it's definitely physical.

I mean, We can control the pulse area,
we can control the polarization state

the central wavelength all of these
intrinsic physical properties yeah

the, the frequency content and so on.

We have pretty much
complete control over that.

We have worked out a method that we can
exert almost complete control over that.

Dan: And what are the
techniques used to control it?

Mirella: It's something called
cavity STIRUP, which is a term

that's featured in the title.

And it's an adiabatic approach which
involves exciting, exciting the

quantum system in a kind of a non
resonant way through something called

the lambda system, which is a system
of three different energy levels.

And you almost have Almost excited
to the upper level and then

bringing down to a lower level.

And this is an adiabatic approach
that it doesn't lose energy as it

were, and it produces predictable
results unlike other approaches.

So you can program it yeah you
can use it to produce programmable

photons or photons that look
exactly like you want them to be.

Dan: programmable photons?

I like that.

And what are the controlling
levers you've got?

Is it the, is it kind of voltage?

Is it a rf?

Is it how do you stimulate the,
what's the physical properties

that are used to stimulate this

Mirella: Uh, So it is the excitation
scheme, the driving pulses used the

properties of the driving pulses and

Dan: Pulses to make pulses.

Mirella: yes, driving pulses, input pulses
make output pulses as they propagate in a

complicated way through the structure and
interact with the quantum system inside.

So it's a, yeah, you have to
consider everything in the

system and how it influences.

the driving pulses and how, yeah, how
they get shaped by those interactions.

Dan: Thank you.

Yeah, so I had a couple of questions
I said that I wanted to throw in.

They're a bit different to
my normal kind of format.

One is about, have you got a
favorite paper or an influential

piece of work in the quantum domain
that really sticks out for you?

Mirella: Yeah, actually it's funny
that you mentioned Noel being on a

podcast earlier this week and the fact
that she mentioned noise cancelling

headphones because I was just
thinking about the same kind of thing.

So really there, there are several
impressive stories out there published

out there all the time, but I'll
pick the one on spectator qubits.

Yeah, if I had to pick one, that
would be about the spectator qubits.

Spin decoherence, or degradation of
the properties of your qubit, is a

consequence of unwanted interactions
with the environment, and it's a big

problem in all the pillars of quantum
technology, in quantum networking,

quantum computing, quantum sensing.

A means to cancel out that noise would
be very beneficial, and spectator

qubits are exactly that kind of, They
are a set of qubits embedded in a

quantum computer whose sole purpose
is to measure the outside noise.

They're not actually used for
storing or processing data.

The information from the environment
is gathered by the spectator qubits

and then used to cancel out the noise
in the actual data processing qubits.

And this is an important advance because
it improves the quality of the data

qubits and reduces the need to run
quantum error correction algorithms to

compensate for complex to the environment.

Yeah, it's a, it's another type of
noise cancelling mechanism accepting

quantum computers rather than
quantum networks, as in Noel's case.

Dan: And.

Yeah, it's a good analogy, isn't it?

So this is like a physical error
correction approach, where you're taking

the noise received by one, making some
assumption that's also being, an

influence on the computational qubit, and
then somehow changing the control of that

other qubit to counteract it, like a,

Mirella: Yes.

Dan: Exactly like noise
cancelling headphones.

Mirella: Just providing a
signal with the opposite.

amplitude, so it cancels out

Dan: Exactly.

Fascinating.

So when was this paper or who is it?

Who is it by?

Mirella: I think that was in maybe a
couple of years ago and it was I think

it came out of Hannes Bernin's lab.

He's a professor at University of Chicago.

He works on quantum computers
made up of Rydberg atoms.

I think he uses rubidium and cesium atoms.

And I see him from time to time.

at conferences here, so he's a local.

Dan: Yeah.

Does he walk around wearing
noise cancelling headphones?

Mirella: I haven't seen him yet, but

Dan: of course.

Yeah, that's great.

Do you think that technique is
underused and you think it's, do

you think it's going to be used more
and more in different modalities?

Mirella: yes, I think so because
you can do error correction to a

certain point after which it's just
been, the noise is too great and

you cannot really correct for it.

So if you have, if you can figure out
a way to shield your quantum system

from extraneous noise or to cancel
it out in a reliable, robust way.

Then this is the way forward, but
it's definitely a big problem.

You don't want your quantum
systems to decohere.

You want them to stay coherent for
pretty much as long as possible, really.

For at least long enough so that they can
be useful for performing computations.

Dan: Exactly.

The more I get into this industry and
the deeper I get into the physics, more

I realise it's all about fighting noise.

Mirella: Yes.

Dan: Really.

Mirella: Yes, quantum systems

Dan: in all aspects

Mirella: systems are so incredibly
delicate that, Anything can perturb them.

So they have to be very carefully
looked after and protected.

Dan: Absolutely.

Okay.

And do you have a vision for the
kind of future of quantum optics and

maybe the modeling of them as well?

Is this kind of a pitch you'd
like to describe for us?

Mirella: I mean, For Quantopticon, I think
the vision is to become the primary tool

for modeling quantum optical systems for
sure, and on the integrated circuit level.

On a system level, I guess.

yeah, there's some time, there's some
developments for us to get to that point.

But this is really our ultimate goal.

I think, apart from all the devices
that will emerge, I'm really actually

very excited about developments in this
field because I think there is some

very exciting physics to be discovered.

I'm particularly excited about the fact
that entanglement distribution over

very large distances may provide insight
into the workings of quantum gravity.

And possibly, if we find surprises
along the way, we might need, this might

necessitate corrections to the standard
description of quantum mechanics,

which would be very interesting.

And I think also advancements in
quantum sensors in ultra sensitive

quantum sensors could also help in the
search for dark matter in the universe

which is again, a bit of fundamental
science that I'm very excited about.

Dan: Oh yeah.

You're, You're going big
with your answer on that.

I was going to say your, your comment
about quantum gravity, I was going to say

it's the dark matter of quantum mechanics,
but actually they're all intertwined.

It's the dark matter of the standard
model in a way that we know these,

or the theory is that these particles
exist and there are many of them.

We just don't know how to detect them yet.

Mirella: Yeah, yeah, exactly, this
advancement in quantum sensors in a

particular frequency band will enable
us to detect them for the first time.

It's another engineering
challenge that will lead to some

fundamental physics insights.

Dan: yeah.

Can I explore that a bit more?

So, Is it the interaction of the
quantum states of atoms or systems with

dark matter that will help us forward?

I don't think it is, is it?

Because we don't see any particular, we
see a relationship with mass and so on,

but we don't see interactions with it.

Mirella: Yeah, I think you're
reaching the limit of my knowledge,

but I can definitely look it
up and tell you more about it.

Dan: No, I like to test people.

I'm trying to, I'm trying to get my
head around these concepts as well.

Mirella: Yeah, so I, I think I found
a bit in that literature I was reading

about dark matter and it's apparently
it's going to be superconducting

cavity qubit systems with 40 percent
photon detection efficiency and 15.

7 dB advantage over standard quantum limit
which hold promise for axion detection.

I don't even know what an axion is, but
it, I think it's a type of dark matter.

Dan: So this is a forecasted,
or a kind of calculated,

Mirella: yeah.

So this is the metrics needed
of the uh, the sensor in order

to be able to detect them.

Yeah.

And.

I think uh, they're gonna,

Dan: there were a few
assumptions made in that.

Mirella: yeah, I think they're gonna be
using squeeze states for 100 kilohertz

to 100 megahertz signals to, to be able
to detect dark matter and maxin, maxins.

But yeah that's all I know.

That's literally everything.

Dan: Mind blown.

Mirella: Yeah,

Dan: cool.

Just to wind down the let me ask
you my question about winding down.

What do you uh, what do you like to
get up to to disconnect from science?

Is there anything you'd like to share with
me and the audience on what you get up to

Mirella: Yeah.

Dan: when you're not thinking
about science, if there is a moment

in your life at all where that

Mirella: Yeah, it's rare, it happens.

So over the past.

year and a half, I've really gotten
into playing chess and I've become

quite addicted to playing chess.

I play that daily.

I'm also a keen musician.

Music has been a underlying
thread throughout my entire life.

And I've I'm mostly playing the
acoustic and electric guitar these days.

And one of my favorite pastimes is
to learn epic lead guitar solos.

So yeah, I'm a very big fan
of some electric guitar sound

there, mostly from the seventies.

Dan: Oh, cool.

Mirella: and

uh,

Dan: Well, I mean,

First of all, chess is a real rabbit hole.

You can go down and lose your entire too.

I know.

Do you study the, the history and
the masters, or is it more about just

Your score or

Mirella: yeah, the at the beginning
it was just trying to find out how,

whether I was really rubbish at it
or I was okay because I've had some

past trauma in my childhood where I
was told that I wasn't any good at it.

But then as I became an adult,
I thought maybe, actually,

maybe maybe that was not true.

Yeah, the first year was really to see
whether I could play averagely at all.

And now I've, I think I'm at the
point where I'm an intermediate and

I'm getting more and more into it
and getting more and more interested

in the um, some of the famous.

figures throughout history and what kind
of moves they've played and why they

were brilliant and what styles they had.

And I also like to take lessons.

I'm subscribed to one of the platforms.

Dan: it seriously.

Mirella: Yeah not from actual masters or

Dan: Oh yeah.

Mirella: but just I'm subscribed to a
platform online and there are lessons

that you can take that take you through,
that improve your game, essentially.

Dan: Yeah, that's a
good, it's a great game.

I got to the point where you
had to really start memorizing

all of the different openings.

Mirella: Oh, okay.

Dan: And particular gambits
and things and I just

Mirella: Yeah.

Dan: I just thought okay.

I'm gonna have to fill my head with these
things and there's not gonna be room for

Mirella: Yeah, I find that it's like,
there's a point maybe where all your

internalized knowledge becomes intuition.

So when you look at a particular
configuration of a board, you just

have this intuition about what
is a good move and what isn't.

So you don't even have
to memorize anything.

You'll just do whatever feels right.

And most of the time it will be correct.

But of course, we all have bad days
when you think something's a good move

and it actually turns out not to be.

Dan: Go with your gut as they say,

Mirella: Yeah.

Dan: And yeah, I mean quite a few
Scientists and quantum physicists

I've spoken to are also musicians.

And that must be a common thread, I think.

Mirella: Yes.

Dan: yeah.

So what's your favorite solo?

Which tune?

Do you know any Pantera?

Mirella: Pantera, I'm not
really into, yeah, Pantera is

like heavy metal or very hard

Dan: Yeah, but the Dimebag was a
soloist and he did some, he's, did

some classic, world changing solos,
which just blew people's minds.

Mirella: see.

I think I'm yet to check it
out, but the latest one of,

Dan: You should.

Mirella: yeah, thank you for the tip.

I'll definitely add it to my list.

But at the moment I'm into
solos by Eric Stewart.

I think he's Yeah, he's a virtuoso
guitarist, so giving me a very hard

time to try and learn his solos.

But there's a song
called Feel the Benefit?

I think it's Feel the Benefit by 10cc.

And I'm learning the solo at
the end, which is very long.

There is no, I've searched the
internet and I couldn't find

any transcriptions of the solos.

I'll have to work it all out by ear.

And the arpeggios are very
tricky because it's so fast.

Dan: Yeah, that's difficult.

Yeah, that's a talent, isn't it?

You've either got it or
you haven't, I think.

Mirella: It was,

Dan: I hadn't heard of Eric Stewart.

I've just looked him up.

He's got a nice hat.

Mirella: oh, okay.

Yeah, he used to be with 10cc.

I think he played with
Wings as well for a while.

He was involved in
several different bands.

I think yeah, 10cc are still going.

I'm aware.

But yeah.

Dan: I'm gonna have to
go and have a listen.

Mirella: Yes.

I think at one point I was thinking
about a way to contact him just to,

to tell me how he played that solo
because it's so tricky to transcribe.

Dan: Just find him on Facebook
and send him a message, surely.

Mirella: Yeah he's probably in his 80s
now, so I don't know if he's on Facebook.

It'll be

Dan: Why aren't there tools that
automatically transcribe music?

Mirella: Yeah, really.

It

Dan: Um, Because ultimately it's just
about identifying the, is it that once you

record it into a single track, you lose
some of the fidelities of the particular

notes and therefore you can't pick them

Mirella: yeah.

Thank you.

Dan: I'd like to take this moment to
thank you for listening to the podcast.

Quantum networking is such a broad domain
especially considering the breadth of

quantum physics and quantum computing all
as an undercurrent easily to get sucked

into So much is still in the research
realm which can make it really tough for

a curious IT guy to know where to start.

So hit subscribe or follow me on your
podcast platform and I'll do my best

to bring you more prevalent topics
in the world of quantum networking.

Spread the word.

It would really help us out