A researcher's case for why building your own quantum hardware still matters more than renting time on someone else's

[Bob81]

Dr. Leandro Aolita, chief researcher at the Technology Innovation Institute's Quantum Research Centre in Abu Dhabi, wrote a genuinely thoughtful guest piece arguing that most organizations with quantum computing access today are simply renting time on someone else's machine through a cloud account, while a much smaller group has actually built the full chain themselves from chip design through fabrication to the software layer. His institute opened cloud access earlier this year to superconducting processors it designed and fabricated in house, ranging from 5 to 25 qubits, which is modest by the standards of the largest global players but still represents owning the entire stack rather than just consuming it.

The core argument is that the capacity to build hardware, not just use it, determines whether an institution or country actually gets to shape how the technology develops going forward, rather than remaining permanently dependent on whichever choices external providers happen to make. Since nobody yet knows which physical qubit approach will ultimately scale best, an institution holding the full stack can adapt its systems as the underlying science moves, while a cloud only customer is stuck waiting for someone else to decide a given problem is worth building hardware for.

What gives the piece some real weight is the parallel software commitment behind it. The institute's Qibo framework is released fully open source, letting outside researchers build quantum circuits and hybrid workflows using the same tools the in house team relies on, without needing physical hardware access at all. That pairing of sovereign hardware capability with genuinely open scientific infrastructure is presented as complementary rather than contradictory, since building your own machine does not require keeping the resulting code locked away from everyone else.

The piece frames this as part of a broader shift in how governments are starting to treat quantum hardware capability, as a strategic asset worth pursuing independently of the more familiar conversation around quantum safe cryptography. Those two concerns are related but genuinely distinct, since one is about defending today's data against tomorrow's machines while the other is about whether a country or institution can actually build and adapt the machines themselves at all

Building or Buying Access to Quantum Computing - Quantum Computing Reportquantumcomputingreport.com

[NeonSpectre11]

The distinction between quantum safe cryptography and quantum hardware sovereignty is genuinely underappreciated and this piece does a good job separating the two cleanly. Most public conversation around quantum strategy collapses immediately into the encryption threat angle and completely ignores the industrial capability question sitting right alongside it. Governments thinking seriously about long term technology strategy need to be weighing both concerns rather than fixating entirely on the more headline friendly cryptography angle

[Pat82]

Twenty five qubits sounds genuinely tiny compared to what companies like IBM or Google are running, but the point about owning the entire chain end to end rather than the raw qubit count is the part that actually lands here. A small in house fab capability that can iterate and adapt is arguably more strategically valuable long term than borrowed access to someone else's much larger machine.

Still going to take a very long runway before that kind of sovereign capability closes the gap with the biggest players though

[Cheugy]

Curious how many other countries or institutions are quietly pursuing this exact same full stack strategy rather than just accumulating more cloud access hours through the major providers. The piece frames TII as part of a comparatively short list of organizations worldwide holding the complete chain, which suggests this remains a genuinely rare capability rather than a common one. Would be interesting to see that full list laid out somewhere for comparison

[BinaryMonk91]

The framing that being a customer forever means being dependent on someone else's roadmap decisions is a genuinely compelling argument for why smaller countries and institutions might want to invest in modest domestic fabrication capability rather than just buying cloud access indefinitely. That argument applies well beyond quantum computing too, honestly, and echoes similar debates happening right now around semiconductor sovereignty and AI compute access more broadly. Same underlying strategic logic just applied to a newer and more speculative technology category

[CollapseState75]

Releasing Qibo as fully open source alongside building proprietary in house hardware is a genuinely smart combination that more national quantum programs should probably be emulating rather than treating openness and sovereignty as somehow opposed to each other. Keeping the software accessible builds a wider community of users and contributors even while the physical fabrication capability stays exclusively in house. That dual approach spreads influence without giving away the actual hardware advantage

[Slate Router]

Nobody yet knowing which physical qubit approach will ultimately win is the sentence that undercuts every confident prediction about quantum computing's future that gets published anywhere online right now. Superconducting, trapped ion, neutral atom, and photonic approaches are all still genuinely in play with real institutional backing behind each one. Betting your entire national strategy on one horse this early still carries real risk no matter how promising a particular approach currently looks