Page 59 - EE Times Europe Magazine – June 2024

EE|Times EUROPE 59

Delft Quantum Ecosystem Addresses Connectivity Challenges

I/O TAKES THREE FORMS
EE Times Europe wondered how Delft Circuits
keeps its products in step with innovations
in a constantly evolving industry. “When we
talk about I/O, we consider it three things,”
Kuitenbrouwer said. “It’s signal transmission,
so bringing a signal from A to B; it’s signal
conditioning, filtering the signals and adjust-
ing them; and then it’s interfacing, which is
getting the signal onto the cables and from
the cables to something else. When we look
at the three approaches that require the most
cables, which are superconducting, spin and
photonics, what they have in common is the
need for signal transmission from a hot place
to a cold place.”

As the quantum industry

advances, there are
growing signs of grassroots
collaboration forming value High-density
chains and establishing modular vacuum
feedthrough
ecosystems. Delft Circuits has for microwave
contributed to this step. drivelines using
Delft Circuits’
Cri/oFlex flexible
cryogenic I/O
SIGNAL TRANSMISSION CHALLENGES
Kuitenbrouwer told EE Times Europe about
signal transmission challenges between
places of significantly different temperatures.
“At the cold place, you have a limited cooling applications send a signal via a connector cite physical resonant frequency reasons for
power, so it’s hard to subtract the heat. And to another component and then another operating at those frequencies, with no bene-
since cables are metal, they bring heat into connector and component arrangement. How- fit in performance achieved by attempting to
the system. You need to bring in the signal ever, Delft has a more integrated approach: operate at higher frequencies.
but not the heat. That’s one of the major “We use the structure and special geometries When asked about the innovation chal-
tricks. You also don’t want to lose your signal to make our low-pass filters, or we make lenges that Delft Circuits faced creating its
integrity.” resistant networks to make our attenuators, product lineup, Kuitenbrouwer noted, “Our
To compound the connectivity challenges, or we adjust the dielectric to make IR filters. superconducting flex is one of our key inno-
Kuitenbrouwer said, “If you need to wire up a Our in-flex integrated components are one vations. A superconductor is needed to reduce
quantum computer with two qubits, you need of the major innovations that achieve two the amount of heat you bring in and its resis-
six cables. That’s not an issue. But if you start things. First, it allows you to scale up because tance. Our normal cables consist of a stack-up
to go through high numbers of qubits, say you need much less space for all the compo- of different materials, the most important
100 qubits, you need 300 cables. That’s nents. The second thing is that it reduces the being silver and the dielectric Kapton. We use
already a lot more, and you begin to get more number of cable problems you get, because titanium and Kapton for the superconducting
issues of space—where does it fit, how does it at every interruption, at every interface, you cables. Getting the stacking of all the mate-
fit in the system? Then there is the thermal have an interruption that might break. And rials and sizes precisely right, it’s a materials
question: How can we avoid bringing heat in as we integrate everything in one monolithic science question.”
and maintain signal integrity?” flex, you don’t have these interruptions, so Kuitenbrouwer noted that microwave
you’re much less prone to failure.” testing at such low temperatures compounds
STRIP-LINE INTEGRATED the challenges. Performing short, open and
FLEX COMPONENTS UNDERSTANDING MATERIALS SCIENCES load VNA testing requires cycling the fridge
EE Times Europe asked Kuitenbrouwer how IS CRUCIAL between each test, which takes a lot of time
Delft Circuits overcame the challenges of Kuitenbrouwer said Delft’s products are and is not precise. Commenting on the need
quantum computing connectivity. “Our designed for operation to at least 10 GHz to build systems, research material properties
technology uses a strip-line structure, and the and up to 15 GHz. He sees some demand and develop test mechanisms and protocols,
parts we add are integrated into flex compo- for operation up to 20 GHz but noted that it he said, “None of it is straightforward; we had
nents,” he said. “These are low-pass filters, varies from customer to customer. Most of to develop everything so that we could make
attenuators and IR filters.” the primary superconducting companies have sure that we could reliably make
He observed that, typically, many quantum qubits operating at about 6 to 8 GHz. They something.” ■

www.eetimes.eu | JUNE 2024