Xanadu Makes Quantum Computer Borealis Available in the Cloud with Advantage Over Supercomputers
Canadian company claims its photonic system with 216 qubits completes a task in 36 microseconds, while a classical supercomputer would need 9,000 years. This is the first publicly accessible cloud quantum computer to achieve quantum supremacy.
Insight: Xanadu Borealis in the Cloud. 9,000 Years vs. 36 Microseconds — It's Not a Breakthrough, It's the Start of the 'Photonic Revolution'
When on June 2, 2026, Canadian Xanadu announced that its photonic quantum processor Borealis became publicly available through Xanadu Cloud and Amazon Braket, the industry stood still. The numbers they provided sound like science fiction: 36 microseconds versus 9,000 years of classical supercomputer work.
But as an analyst who has been following the race between photonic and superconducting architectures for the past two years, I see this: it's not just another 'breakthrough.' It's the moment when the photonic branch of quantum computing split from the superconducting one and went its own way. And this path has three key differences: a room instead of a cryostat; optics instead of wires; and complete independence from Nvidia. And the main insight here is that Xanadu did what IBM or Google didn't dare to do — gave the world a chance to verify quantum supremacy with their own hands.
[The Gist]: What's Really Happening
Officially: Xanadu deployed the 216-qubit photonic processor Borealis in the public cloud. The system is built on the principle of 'squeezed-state qubits,' where qubits are generated as light pulses entangled in three temporal dimensions within a continuous optical fiber.
Unofficially: Xanadu demonstrated that 'quantum supremacy' is not a theory, but a service.
Let's read the numbers correctly. The task that Borealis solves is Gaussian Boson Sampling (GBS). This is not 'any task.' It's a specially chosen task for which the quantum computer has an exponential advantage. But! This is the world's first programmable photonic system that does this in the cloud, accessible to everyone.
The essence of the deal: Xanadu is not selling 'a solution to all problems.' Xanadu is selling the opportunity to verify. Any developer, any researcher, any skeptic can log into Amazon Braket, run a task on Borealis, and see for themselves: 9,000 years vs. 36 microseconds — it's not marketing, it's a measured fact.
Why does this change the game? Because until now, all claims of 'quantum supremacy' (Google Sycamore in 2019, Chinese Jiuzhang in 2020) were laboratory artifacts. You couldn't touch them. Borealis — you can. This is the first public cloud quantum computer in history to achieve quantum supremacy. And it opens the gate for industry validation.
Timeline and Context
To understand why this happened now, we need to rewind four years.
June 2022: Xanadu publishes a paper in Nature about Borealis — a 216-qubit photonic processor demonstrates a 50 million times advantage over classical supercomputers. But it's a lab prototype.
May 2026: Xanadu shares (XNDU) fall 55% in a month, despite technological breakthroughs. The market doesn't believe. Investors want not 'papers in Nature,' but 'live hardware.'
June 2, 2026 (Amazon re:Invent or similar date): Xanadu announces: Borealis is now on Amazon Braket. Richard Moulds from Amazon directly says: 'Researchers and developers will be able to independently verify the claim of quantum advantage.'
The context everyone missed: Xanadu's War with Nvidia.
A month earlier, Nvidia released AI models for calibrating quantum processors. Xanadu CEO Christian Weedbrook publicly stated: 'Nvidia is not interesting to us. We have our own tools for 5-6 years. And in general, we cannot use their chips for error correction because our photonic approach is too fast.' Instead of Nvidia, Xanadu works with AMD.
Translation: Xanadu is building an ecosystem independent of Nvidia. If photonic quantum wins, Nvidia loses one of the fattest pieces of the future market — control over quantum-classical hybrids. Xanadu is AMD's 'Trojan horse' in the war against CUDA.
Who Wins and Who Loses
Winner #1: Amazon. Amazon Braket gets exclusive access to the world's first public quantum-supremacy computer. This is a differentiator against AWS and Google Cloud. Amazon can now say: 'Google has Sycamore (unavailable), we have Borealis (available to everyone).' This is a win in the cloud marketing war.
Winner #2: The Research Community. Finally, there is a platform for validating quantum algorithms on real hardware, not simulators. Universities and R&D centers can publish papers with results obtained on Borealis. This will accelerate the development of quantum software manifold.
Winner #3: PennyLane (Xanadu's ecosystem). PennyLane is an open-source library for quantum machine learning. With access to Borealis, its popularity will skyrocket. Developers will learn on PennyLane because only there you have access to real quantum-supremacy hardware. Xanadu is creating a 'quantum TensorFlow.'
Loser #1: IBM Quantum. IBM has not yet provided public access to its quantum-supremacy systems. Their 433-qubit Osprey is an impressive machine, but it cannot do what Borealis does (at least publicly). IBM is losing the 'accessibility' race.
Loser #2: Google Quantum AI. Sycamore was the first to claim quantum supremacy in 2019. But Google never opened public access to it. Now Xanadu has done it. Google looks like a company that 'hides its toys.' Reputational blow.
Unexpected Loser: Nvidia. This isn't written in press releases, but Xanadu consciously distances itself from Nvidia. They have their own error correction strategy, their own software, their own stack. If photonic quantum becomes the standard, Nvidia will be left without a seat at the table in this niche. AMD, on the other hand, gains a strategic partner.
What the Media Isn't Saying
First and foremost insight: The GBS task has no practical application (yet).
All headlines scream '36 microseconds vs. 9,000 years.' But quietly add: 'in the Gaussian Boson Sampling task.' This is not RSA decryption. This is not modeling a caffeine molecule. It's a specially crafted task to demonstrate supremacy. Xanadu understands this. They don't even try to sell Borealis as a 'business solution.' They sell it as a 'research tool.'
Second silence — about the uncertainty of the future.
Borealis is a programmable system, but its program is limited. You cannot run Shor's algorithm or arbitrary quantum circuits on it. It's a specialized processor for GBS and similar tasks. A universal photonic quantum computer that can be reprogrammed for any task is still years away.
Third insight: This is not about revenue, it's about credibility.
TipRanks directly writes: 'Borealis is unlikely to become a significant revenue generator in the near term. But it gives Xanadu a stronger story to tell cloud users, corporate partners, government agencies, and future investors.' Moreover, in the first quarter of 2026, Xanadu reported revenue of $2.8 million — 4 times more than a year earlier. But that's still peanuts. The company's market capitalization is $4.53 billion.
Fourth and most important insight: Xanadu's financial position is shaky.
According to Simply Wall Street, Xanadu has less than one year of operating cash runway. Yes, they have access to $300 million through an agreement with Yorkville and potential $390 million from the Canadian government. But these are 'lifebuoys.' Shares fell 55% in the month before the announcement. Borealis in the cloud is an attempt to show the market: 'We are alive, we are working, come and see.' And the stakes here are the company's survival.
Forecast: Next 30 Days and 90 Days
Next 30 days (until early July 2026).
A wave of 'validation reports' will begin from independent researchers who gain access to Borealis via Amazon Braket. If at least three respected laboratories confirm Xanadu's numbers (36 microseconds vs. 9,000 years), XNDU shares could recover some of the decline and rise 20-30% from current levels (currently around $14-15, peak was $42).
However, if it turns out that '9,000 years' is a calculation for a naive algorithm, and with optimizations a classical computer can do it in 100 years (or even 1 year), the reputational damage will be catastrophic. So watch arXiv publications.
Next 90 days (September 2026).
Xanadu will have to report Q2 2026 earnings. If revenue grows insignificantly (e.g., to $3.5-4 million), investors may decide that Borealis isn't helping sales. This could trigger a new wave of decline.
However, another scenario is more likely: Amazon will aggressively promote Borealis as part of its quantum portfolio. Case studies will emerge from researchers who found real applications for GBS (e.g., in optimizing photonic circuits for telecom). This will create a narrative of 'the first commercially significant application of quantum supremacy.' Xanadu shares could rise to $25-30.
Also, the Canadian government will likely confirm its support for Xanadu to the tune of about $390 million. This would remove bankruptcy risks and be a strong signal to the market.
Main takeaway on June 5, 2026: Xanadu did what no one had done before — opened quantum supremacy for public testing. Technically, this is a landmark event. Commercially, it's still a 'story for investors,' not a business. The company is on the edge: either this move convinces the market and government to invest, and Xanadu becomes the leader of the photonic revolution; or the hype fades, and $4.5 billion in market cap becomes a memory. But in any case, June 2, 2026, will go down in the history of quantum computing as the day when 'light' became accessible to everyone. And that's wonderful.
— Editorial Team
No comments yet.