March 26, 2026

Software revolutions begin in hardware, typically 5-10 years before they become visible for most end users.

Quantum computing promises a significant change in computational power. It could reshape industries, from chemistry to logistics. But today, that promise remains unrealized. Early demonstrations of “quantum advantage” exist, yet they are limited to useless or niche scientific problems.

The key milestone to watch is simple: the “AlexNet moment” of quantum computing.

Those who spot it early, across governments, investors, and corporate end-users, will capture most of the benefits.

Here is the pattern to follow:

Phase 1: HPC validation

Before AI exploded, GPUs were gradually adopted in high-performance computing (HPC). Originally developed for gaming, Nvidia GPUs were quickly used for scientific workloads such as physics and chemistry simulations. By 2010, 10% of top supercomputers already used GPUs.

Quantum processors (QPUs) are at this stage today: early deployments in HPC centers, validating performance on scientific computing workloads.

Phase 2: the “AlexNet moment”

In 2012, the AlexNet deep-learning model, trained on Nvidia GPUs, clearly outperformed CPUs in the ImageNet competition (40% less errors). It was a a clear demonstration of significant performance advantage on a problem with strong industrial relevance (autonomous vehicles, medical image analysis, industrial quality control, etc).

From that moment, the market expanded beyond HPC (~$20B) to the much larger enterprise automation opportunity (~$100B+), supported by the larger trillion-dollar IT market. AlexNet eventually led to ChatGPT-like moments, that revolutionize entire industries.

An investor who had bought Nvidia stock around that time would have achieved roughly a 600× return (vs. 50× for investors after the 2017 “Attention Is All You Need” paper from Google).

What to watch for in quantum

Possible “AlexNet moment” candidate applications for quantum computing include:

• Unprecedented accurate molecular and materials modeling which could cascade into transformative impacts across chemistry, materials, and drug discovery, sparking a huge demand for quantum computers from the industry.
• New algorithms or heuristics emerging as quantum hardware matures could unlock value in optimization or AI-related workloads, by performing beyond the capabilities of classical computers or by being more cost-efficient.

 April 16, 2026

Shor’s algorithm has been a powerful narrative to trigger public investments and, as such, a key driver of quantum technologies progress, but it may now begin to slow the field down.

Shor’s Algorithm as a Catalyst

At first, Shor’s algorithm acted as a catalyst. Discovered by Peter Shor in the 1990s, it showed that a quantum computer, if powerful enough, could break RSA and cryptocurrencies (ECC) encryption. Suddenly, quantum computing was no longer just a tool for exploring science (Feynman’s initial vision). It became a direct threat to the digital economy built on classical cryptography.

Governments took notice and started investing heavily to avoid lagging behind on this strategic infrastructure.

These programs funded hardware and algorithms development, academic labs and start-ups. They helped structure the field and gave it long-term direction, which is key for deep tech. In that sense, Shor was a booster drug.

The Same Force May Now Slow Things Down

Building a full quantum computing system is too complex for any single lab or company. Progress relies on open research and collaboration, as Austin Fowler recently emphasized. But as Shor’s implementation becomes more realistic, that openness may start to shrink.

Indeed, recent progress has cut the resources needed to break cryptosystems like RSA and ECC by around 100x. What once looked decades away, requiring millions of qubits, now appears in some 2030-2033 roadmaps. The threat is starting to feel real for governments.

And when a technology becomes a real security concern, openness tends to fade. As Scott Aaronson pointed out, the Manhattan Project followed that path: research on uranium fission was published, until it wasn’t.

We may already be seeing early signs with quantum computing. First, Google recently chose not to openly release its latest quantum algorithm for breaking cryptosystems. Critically, access to advanced fabrication could become limited under geopolitical pressure.

It’s a shame as quantum computing could benefit other areas, for the greater good. The cure may lie in post-quantum cryptography – more on this soon.