2024 was the year quantum computing stopped being mostly about bragging rights over qubit counts and started being about something more useful: making those qubits actually reliable. That shift matters more than it sounds, because a quantum computer with thousands of noisy, error-prone qubits is far less useful than one with a small number of qubits that behave predictably. This is a rundown of what actually happened in 2024, why researchers consider it a turning point, and what it means for where the technology goes from here.

Why 2024 Is Considered a Turning Point

For most of quantum computing’s history, progress got measured in raw physical qubit counts. More qubits sounded like more power, and headlines followed that logic for years. The problem is that qubits are extremely fragile. They lose their quantum state (a process called decoherence) from the tiniest interference: heat, vibration, even stray electromagnetic noise. Historically, adding more qubits to a chip just meant adding more sources of error, not necessarily more usable computing power.

2024 is when several major research teams demonstrated, convincingly, that this trade-off doesn’t have to hold forever. The breakthroughs below all circle back to the same underlying idea: grouping several unreliable physical qubits into a single “logical qubit” that can detect and correct its own errors in real time.

Google’s Willow Chip

In December 2024, Google Quantum AI unveiled Willow, a 105-qubit superconducting processor that many researchers point to as the year’s single biggest milestone. What made Willow significant wasn’t primarily its speed, though it did complete a benchmark computation in about five minutes that would take a classical supercomputer far longer than the current age of the universe to finish.

The more important result was that Willow achieved what’s known as “below-threshold” error correction. In plain terms: as the team added more physical qubits into an error-corrected group, the error rate went down instead of up. That’s the exact opposite of what had happened with most earlier quantum hardware, and researchers had been chasing this specific result for close to three decades. It’s the difference between a technology that gets messier as it scales and one that gets more dependable as it scales, which is the property any real, commercially useful quantum computer will eventually need.

Microsoft and Quantinuum’s Logical Qubits

In April 2024, a joint team from Microsoft and Quantinuum published results using Quantinuum’s trapped-ion hardware that the research community quickly recognized as a second major turning point for the year. The collaboration created four logical qubits with error rates roughly 800 times lower than the physical qubits that made them up, and ran 14,000 independent instances of a quantum circuit without a single error.

That error-free run at scale is the part worth paying attention to. A handful of clean results can happen by chance. Fourteen thousand consecutive error-free runs is a much stronger signal that the underlying error-correction approach is genuinely working, not just getting lucky.

QuEra’s Neutral-Atom Approach

Also in early 2024, QuEra Computing (whose systems are available through Amazon Braket) announced that it had built a logical qubit using only eight physical qubits, using a different error-correction method based on what are called transversal gates. QuEra laid out a multi-year roadmap: 10 logical qubits by the end of 2024, 30 by 2025 using a technique called magic state distillation, and 100 by 2026.

QuEra’s approach trades raw speed for error-correction quality, meaning its systems tend to run more slowly than some competitors, but with a stronger emphasis on getting the underlying reliability right first. It’s a reminder that “the biggest breakthrough” in quantum computing doesn’t always mean “the fastest chip.” Different hardware approaches (superconducting qubits like Google’s, trapped ions like Quantinuum’s, and neutral atoms like QuEra’s) are all racing toward the same finish line by different routes, and it’s still an open question which approach scales best in the long run.

IBM’s Heron Chip and the Shift Toward Circuit Depth

IBM’s 2024 updates leaned into a different but related theme: the industry moving past the idea that qubit count alone tells you how powerful a system is. IBM’s Heron processor, part of its broader roadmap, was built to run larger, more complex circuits (reportedly capable of executing around 5,000 gates) rather than simply packing in more qubits.

This connects to a concept called circuit depth, essentially how many sequential operations a quantum computer can run before errors pile up and corrupt the result. A chip with fewer qubits but a deeper reliable circuit can, in practice, be more useful than a chip with more qubits that can only run a handful of operations before falling apart. IBM’s 2024 messaging reflected an industry-wide move toward this more nuanced way of measuring progress.

Progress Beyond Hardware: Algorithms and Real-World Applications

Not every 2024 breakthrough involved new chips. On the software side, Microsoft’s Azure Quantum Elements platform was used to combine AI models with early quantum tools to simulate molecular behavior, running more than a million chemistry calculations to evaluate complex reaction networks in a single project. This matters because chemistry and materials science are widely expected to be among the first fields where quantum computing delivers practical value, long before it’s useful for more general-purpose computing tasks.

IonQ also reported reaching a new algorithmic performance benchmark (referred to as #AQ 35) in January 2024, a measure intended to reflect how complex a circuit a system can run with meaningful accuracy, rather than just how many qubits it has on paper.

What These Breakthroughs Actually Mean

It’s easy to see headlines like “quantum computer solves problem faster than a classical supercomputer” and assume quantum computers are now outperforming regular computers at everyday tasks. That’s not quite accurate, and it’s worth being precise about the difference.

The tasks quantum computers “won” in 2024 were specifically chosen because they’re hard for classical computers and well-suited to quantum hardware, not general-purpose tasks like running spreadsheets or browsing the internet. What 2024’s results actually demonstrate is that the fundamental error-correction problem, long considered the biggest roadblock to useful quantum computing, has a real, working solution, at least in principle and at a small scale.

Getting from “we proved this works for a handful of logical qubits” to “we have a machine that solves valuable, real-world problems” is still a long road. Most researchers and companies estimate fault-tolerant quantum computers capable of running commercially valuable algorithms are still somewhere between five and ten years away, with 2029 to 2033 being a commonly cited window. Those are targets based on current roadmaps, not guarantees, and quantum computing has a track record of ambitious timelines slipping.

Why This Matters Even If You’re Not a Physicist

You don’t need to understand quantum mechanics to care about this. A few practical reasons this affects people outside research labs:

• Cybersecurity. Some current encryption methods could theoretically be broken by a sufficiently powerful quantum computer. That’s part of why organizations and governments are already migrating toward post-quantum cryptography standards, years before such a machine exists, because data intercepted and stored today could potentially be decrypted once the technology matures.
• Drug discovery and materials science. Simulating molecules accurately is something classical computers struggle with as complexity increases. Quantum computing’s early, most promising commercial use case is likely to be in chemistry and materials research rather than general computing.
• Investment and national strategy. Quantum computing startups raised billions of dollars in 2024, and multiple governments expanded national quantum research programs. This isn’t purely academic interest; it reflects a broad bet that the technology will eventually matter economically and strategically.

The Bottom Line

2024’s real story wasn’t a single flashy chip. It was multiple independent teams, using different hardware approaches, all converging on evidence that quantum error correction, the field’s oldest and hardest problem, actually works at small scale. Google’s Willow chip, Microsoft and Quantinuum’s logical qubits, QuEra’s roadmap, and IBM’s shift toward circuit depth all point the same direction: the field moved from “can we build bigger quantum chips” to “can we build quantum chips that get more reliable as they scale.” That’s a genuinely different, and more promising, question than the one the industry was asking a few years earlier.

FAQ

What was the single biggest quantum computing breakthrough in 2024? Most researchers point to Google’s Willow chip, specifically its demonstration of below-threshold error correction, meaning error rates decreased as more qubits were added into protected groups. This solved a problem the field had worked on for nearly three decades.

Does this mean quantum computers can now do everything faster than regular computers? No. The 2024 results involved specific benchmark tasks chosen because they’re difficult for classical computers and well-suited to quantum hardware, not general-purpose computing tasks.

When will quantum computers be available for everyday, commercial use? Most current estimates from hardware companies and researchers place fault-tolerant, commercially useful quantum computers somewhere between 2029 and 2033, though these are roadmap targets rather than guaranteed timelines.

What is a “logical qubit” and why does it matter? A logical qubit is created by combining several physical qubits so that errors can be detected and corrected automatically. It’s the foundation of building quantum computers reliable enough to run real, useful computations rather than just short demonstrations.

Why is quantum computing relevant to cybersecurity right now, if usable machines are still years away? Because data encrypted today could be intercepted and stored, then decrypted later once a sufficiently powerful quantum computer exists. This is driving early adoption of post-quantum cryptography standards well ahead of the technology itself being ready.