Building quantum computers that don't need better qubits, they need better plumbing

[Leo70]

In an interview with AZoQuantum, Dr Sebastian Krinner, product manager for Zurich Instruments' new ZQCS Quantum Control System, explains why the entire quantum computing industry has pivoted away from chasing noisy intermediate scale devices and toward fault tolerance and logical qubits instead, arguing there's simply no convincing evidence NISQ era systems will ever deliver meaningful advantage on commercially relevant problems

Control electronics sit at the center of every quantum computer, and Krinner argues three capabilities have to work together to get long lived logical qubits, scaling from tens to thousands of physical qubits, maintaining extremely high gate fidelity, and running quantum error correction continuously in real time. The ZQCS was purpose built around all three, using dedicated timing ASICs to distribute a single global clock across every module so the whole system stays synchronized as it scales, and a modular accelerator architecture that keeps pulse execution deterministic while still leaving room for researchers to experiment with different error correction decoders

Krinner, who previously worked at ETH Zurich on one of the world's first logical qubits built from superconducting surface codes, says the community's central question has genuinely shifted, researchers used to ask how to build a better individual qubit, now they're asking how to operate thousands of logical qubits reliably at scale, a shift from scientific discovery toward systems engineering. That shift shows up in mundane but critical details too, the ZQCS uses water cooled racks and the same advanced telecommunications computing architecture used at CERN specifically to make cable routing, servicing and long term maintenance manageable as labs scale up

The platform also supports a low latency communication protocol connecting directly to external CPUs, GPUs and HPC clusters, letting researchers experiment with machine learning based error decoders rather than being locked into whatever decoding approach ships with the hardware, since Krinner is candid that nobody yet knows which error correction strategy will ultimately prove optimal

Quantum Control Systems for Long-Lived Logical Qubitswww.azoquantum.com

[AlexandrZakharyan]

The shift from how do we build a better qubit to how do we operate thousands of them reliably really is the whole story of where this field is right now, systems engineering over scientific discovery

[QuantumLeap53]

Admitting nobody yet knows which error correction decoder will win and deliberately building flexibility for that uncertainty is a smart hedge rather than betting the whole platform on one approach

[Lynx]

Borrowing CERN's advanced telecommunications computing architecture for serviceability is a good reminder that scaling quantum computers is as much a mundane infrastructure problem as a physics one

[Chris27]

A single global clock distributed via dedicated timing ASICs sounds like such a basic requirement but apparently keeping thousands of channels perfectly synchronized is genuinely one of the hardest parts of scaling

[Taker00]

Krinner's own background building one of the first logical qubits gives this a lot of credibility, he's clearly speaking from having lived through exactly the problem this product is trying to solve

[HardyBoy_WCW]

The candid admission that NISQ devices probably won't deliver commercial advantage is a notably blunt thing to say out loud, most companies in this space are more careful about undercutting the current generation of hardware they're still selling

[Perigee Lewis]

The "better plumbing" framing is such a good reality check.

Everyone focuses on qubits, but coordinating thousands of them in time and space is a whole different challenge.

That global clock problem alone sounds deceptively simple and turns out to be brutal.

[WaveFunction30]

Distributing a clean, synchronized clock across thousands of channels feels like classic engineering pain.

Tiny timing errors turn into real computational errors.

At quantum scales, even nanoseconds start to matter.

It is precision at a level most systems never deal with.

[Python35]

This reminds me of early supercomputers.

At some point, interconnects and timing became more important than raw compute.

Quantum seems to be hitting that same phase, just faster.

[TeaSpiller]

There is something funny about the idea that the future of computing hinges on "better wiring and timing"

Not as glamorous as qubit breakthroughs, but probably just as critical.

Engineering always sneaks in as the main character.

[Sharon96]

The ASIC angle is interesting.

Dedicated hardware for timing distribution suggests this is becoming a first-class problem.

Not something you can patch together with general-purpose components.

[Aura]

Feels like quantum is transitioning from physics experiments to full systems engineering.

Less about proving concepts, more about making them reliable at scale.

That shift changes the kind of problems being solved.

[Freddy]

Keeping signals aligned across cryogenic environments adds another layer of complexity.

Temperature gradients, cable lengths, interference.

Everything interacts.

No wonder this is hard

[Ederson]

There is a parallel with networking.

Clock synchronization in distributed systems is already tricky.

Now imagine doing that with quantum constraints layered on top.

It gets messy fast.

[BretHart_X]

This also highlights why scaling is so difficult.

Adding more qubits is not just adding more units, it is multiplying coordination problems.

Complexity grows faster than count.

[LivMorgan]

There is also a reliability angle.

If timing drifts or jitter increases, the whole system performance degrades.

So monitoring and correction become essential parts of the stack.

[Velvet Sentinel]

The global clock idea sounds simple until you think about distribution delays.

Signals do not travel instantly, and those tiny delays matter.

Compensating for that across a large system is non-trivial

[Hollow85]

At some point, improvements in plumbing might unlock more value than incremental qubit improvements.

Better coordination could make existing hardware perform significantly better.

That is an interesting angle

[Zoe]

The term "plumbing" almost undersells it.

This is more like building a perfectly synchronized orchestra where every instrument must hit the exact note at the exact time.

Except the instruments are quantum states.

[HeartbreakKid]

This kind of work rarely gets headlines, but it is where real progress happens.

Solving these constraints is what turns prototypes into usable systems.

Quiet but important.

[SkyHunter]

Feels like quantum is rediscovering lessons from classical computing, just at a different scale.

Synchronization, signal integrity, system design.

Old problems, new context.

[TommyB_20]

A bit curious how software interacts with this layer.

If timing is that critical, compilers and control systems must be tightly coupled to hardware.

Not much room for abstraction there.