Quantum computing made real progress in 2024, mostly by becoming more reliable. Researchers did not just add qubits. They found better ways to protect quantum information, catch errors, and control complex operations.
The latest breakthroughs in quantum computing in 2024 included Google’s Willow chip, new logical-qubit results from Harvard, QuEra, Microsoft, Quantinuum, and Atom Computing, and a stronger Heron processor from IBM. Google DeepMind also used AI to spot errors inside quantum processors. Eight results stood out, and each one tackled a different piece of the same problem.
That problem is easy to state. Today’s quantum computers are noisy and hard to scale, so a bigger chip is not always a more useful one. The real goal is a machine that can run long, reliable calculations.
What Were the Biggest Quantum Computing Breakthroughs in 2024?
The biggest 2024 breakthroughs made quantum computers more reliable, not just bigger. Google’s Willow showed that error correction can get better as a system grows, and several teams showed that logical qubits can do real work. The table below lists all eight results.
| Breakthrough | Main result | Why it matters |
|---|---|---|
| Harvard, MIT, and QuEra | Up to 48 logical qubits on a neutral-atom processor | Showed advanced logical operations with neutral atoms |
| Microsoft and Quantinuum | Logical error rates 800 times lower than physical error rates | Protected quantum information far better |
| IBM Heron R2 | 156 qubits, with circuits of up to 5,000 two-qubit gates | Deeper circuits on superconducting hardware |
| NIST | Three finalized post-quantum cryptography standards | Prepares data security for future quantum threats |
| Google DeepMind AlphaQubit | An AI decoder that made fewer errors than earlier decoders | More accurate error detection |
| Microsoft and Atom Computing | 24 entangled logical qubits | Logical qubits used for real operations |
| Hebrew University of Jerusalem | Cluster states worth more than nine qubits from two photons | A leaner way to build photonic systems |
| Google Willow | Below-threshold error correction | A key step toward fault-tolerant computing |
These are different kinds of progress. Some are processors, others are software, and one is a security standard. Together they show how many parts of quantum computing have to improve before large, useful machines exist.
1. Harvard, MIT, and QuEra Run a 48-Logical-Qubit Processor
Researchers at Harvard, MIT, and QuEra showed a neutral-atom processor that worked with up to 48 logical qubits. The study appeared in Nature. It was first posted online in December 2023 and reached the print edition in early 2024, which is why many roundups count it as a 2024 result.
How does neutral-atom quantum computing work?
A neutral-atom quantum computer stores quantum information in individual atoms. Lasers hold the atoms in place and can move them into new patterns. Extra laser pulses then control the atoms’ quantum states and make them interact. Because the atoms can be rearranged, researchers can link qubits in flexible ways.
The 48-logical-qubit result
The paper reported several results on a single processor:
- Up to 48 logical qubits
- 228 logical two-qubit gates
- 48 logical CCZ gates
The team did more than pack many atoms into one system. It showed that the processor could encode logical information and run operations on it.
Why do logical qubits matter?
A physical qubit is a real device that holds quantum information, such as an atom, an ion, or a superconducting circuit. It is sensitive to noise. A logical qubit spreads one piece of information across several physical qubits, so the group can catch and fix errors.
This work showed that neutral atoms can support advanced logical operations. That makes it a step toward larger, more reliable processors. It does not mean 48 fully fault-tolerant, commercial-grade qubits, because error rates, control, and scale all need more work.
2. Microsoft and Quantinuum Report 800 Times Lower Logical Error Rates
In April 2024, Microsoft and Quantinuum announced a major error-correction result. They paired Quantinuum’s trapped-ion hardware with Microsoft’s error-correction software. The team built four logical qubits from 30 physical qubits.
What made the result important?
The logical error rate was 800 times lower than the error rate of the physical qubits behind it. The team also ran 14,000 separate experiments without seeing a single error in the logical circuit. Errors that pile up can ruin a quantum calculation, so protecting the data while the computer works is the whole point of error correction.
The system used active syndrome extraction. This step checks for signs of errors without reading out, and destroying, the information being protected.
Why use trapped ions?
A trapped-ion computer holds information in charged atoms suspended by electric fields. Lasers control the ions and make them interact. These machines are known for precise control and strong gate quality. Their weak spots are speed and the difficulty of scaling up.
The result showed that logical qubits can be far more reliable than the physical qubits used to build them. Still, 14,000 clean runs do not mean the system never fails. They describe one specific set of experiments.
The collaboration made further progress in September 2024. Microsoft and Quantinuum created 12 entangled logical qubits on Quantinuum’s H2 processor, tripling their earlier logical-qubit count. They also combined logical quantum computing with an AI model and classical high-performance computing to estimate the ground-state energy of a catalytic intermediate. This was a hybrid research demonstration, not proof of practical quantum advantage in chemistry.
3. IBM Improves Its Heron R2 Quantum Processor
IBM kept pushing its superconducting hardware with the Heron R2 processor. This version has 156 qubits and is built to run deeper, more complex circuits. It uses a heavy-hex layout, which connects qubits in a pattern that helps limit unwanted interactions.
What changed in Heron R2?
IBM pointed to several hardware upgrades:
- Tunable couplers between qubits
- A heavy-hex qubit layout
- Methods that reduce noise from two-level systems
- Better performance on circuits with many two-qubit gates
At its 2024 Quantum Developer Conference, IBM reported that Heron R2, working with Qiskit, could accurately run certain classes of quantum circuits containing up to 5,000 two-qubit gate operations.
Why do two-qubit gates matter?
Gates are the basic operations of a quantum circuit. A single-qubit gate changes one qubit. A two-qubit gate links two qubits and creates entanglement. Useful algorithms need many of them, yet each gate adds a little error, and in a deep circuit those errors pile up and can ruin the answer.
IBM’s result was an advance in hardware and circuit performance. That is a different kind of progress from Google’s below-threshold error correction, even though both teams are working on related problems.
4. NIST Finalizes Three Post-Quantum Cryptography Standards
In August 2024, the U.S. National Institute of Standards and Technology (NIST) finalized its first three post-quantum cryptography standards. This was not a new quantum machine. It was a security milestone tied to the risks that powerful quantum computers could pose later.
| Standard | Algorithm | Main purpose |
|---|---|---|
| FIPS 203 | ML-KEM | Key establishment |
| FIPS 204 | ML-DSA | Digital signatures |
| FIPS 205 | SLH-DSA | Hash-based digital signatures |
These standards give organizations approved tools for protecting data against future quantum attacks. Read the official NIST announcement.
Why could quantum computers threaten encryption?
Much of today’s public-key security relies on math problems that classical computers find very hard. A large, fault-tolerant quantum computer could use Shor’s algorithm to solve some of them far faster. Current machines are nowhere near that scale, so today’s encryption is not under threat yet.
But sensitive data can stay valuable for many years. An attacker can steal encrypted files now and try to unlock them later, a tactic often called “harvest now, decrypt later.”
Is post-quantum cryptography the same as quantum encryption?
No. Post-quantum cryptography uses math designed to resist quantum attacks, and it runs on ordinary computers. Quantum cryptography is different because it uses the laws of quantum physics as part of the security itself. NIST’s standards give governments, banks, and tech companies a clear starting point for upgrading their systems.
5. Google DeepMind Introduces AlphaQubit, an AI Error Decoder
A quantum computer needs more than good hardware. It also needs fast, accurate ways to find its own errors. In November 2024, Google DeepMind and Google Quantum AI introduced AlphaQubit, an AI system that decodes errors in quantum processors. It uses a transformer neural network to read measurement data and work out which errors likely hit the protected information.
What did AlphaQubit achieve?
On data from Google’s Sycamore processor, AlphaQubit made fewer decoding errors than earlier methods. In the largest experiments, it made 6% fewer errors than a tensor-network decoder and 30% fewer than a correlated-matching decoder. Simulations also tested it at code distances up to 11.
How does AI help with quantum error correction?
A quantum processor’s measurements hold clues about what went wrong. A decoder reads those clues and picks the most likely fix. That is hard because errors can be linked in ways that simple rules miss. A machine-learning model can learn those patterns from data, which may keep decoding accurate as processors grow.
Why AlphaQubit is not ready for real-time use
Accuracy is only half the job, because a decoder also has to keep pace with the processor in real time. Google said AlphaQubit is not yet fast enough for that at the speeds large quantum computers will need. It is a promising research advance, not a finished answer.
6. Microsoft and Atom Computing Entangle 24 Logical Qubits
In November 2024, Microsoft and Atom Computing announced progress with neutral-atom hardware. Their system entangled 24 logical qubits. The team also ran a computation on 28 logical qubits using the Bernstein-Vazirani algorithm.
What is a GHZ state?
The 24 logical qubits were placed in a GHZ state, a special kind of entanglement in which many qubits share one linked state. Entanglement is a core ingredient of quantum computing. It lets qubits work together in ways ordinary bits cannot.
What did the 28-qubit run show?
Bernstein-Vazirani is a standard test algorithm for quantum computers. The team ran it on 28 logical qubits and used methods to detect and correct errors from atom loss and other faults. Together, the two results show logical qubits doing more than storing data, since they took part in operations and a full algorithm.
The 24-qubit entangled state and the 28-qubit computation were separate experiments. They should not be described as one run with 28 entangled qubits.
7. Photonic Quantum Computing Advances With High-Dimensional Cluster States
Photonic quantum computers carry quantum information in particles of light called photons. In September 2024, researchers at the Hebrew University of Jerusalem published a Nature Photonics paper on a way to build large entangled states with far fewer photons. They encoded several qubits in each photon using its spatial structure. The team generated cluster states containing more than nine qubits, encoded across just two photons, at a rate of about 100 hertz.
What are photonic cluster states?
A cluster state is a web of entangled quantum states used for measurement-based quantum computing. The computer first prepares the entangled resource. It then measures parts of it, one after another, to carry out a calculation. The team also showed that qubits stored in the same photon can be processed at once, which shortens the computation.
Why is this useful, and what is missing?
Photons travel well through optical fiber, which makes photonic designs attractive for quantum networks. Packing more qubits into each photon reduces the number of photons needed, so it saves resources.
Big, useful photonic machines are still far off. Photon loss, imperfect detectors, and the size of the states required remain hard problems. This result is a research method that could support future systems, not a general-purpose photonic computer.
8. Google Willow Demonstrates Below-Threshold Quantum Error Correction
Google’s Willow chip was the most talked-about quantum result of 2024. Announced in December, Willow is a superconducting processor with 105 physical qubits. Its main result was below-threshold quantum error correction.
What does below-threshold error correction mean?
Qubits are fragile, and heat, noise, and imperfect operations can all disturb them. Error correction spreads information across many physical qubits and keeps checking them for signs of trouble. Below threshold means that a larger error-correcting code lowers the logical error rate. Adding more physical qubits makes the logical qubit better instead of worse, which is a key step toward fault-tolerant quantum computing.
What did Willow show?
Google reported that increasing the surface-code distance from 5 to 7 reduced the logical error rate per correction cycle by a factor of about 2.14. The distance-7 logical qubit also preserved information more than twice as long as its best physical qubit. The Nature paper also reports stable operation over up to one million error-correction cycles. The real-time decoding result was demonstrated with the distance-5 code.
Google also reported that Willow finished a Random Circuit Sampling test in under five minutes. The company estimated that one of today’s fastest supercomputers would need about 10 septillion years for the same task. That test is a narrow benchmark. It does not show that Willow beats classical computers at everything, or that it can already help with drug design or finance.
What has Willow not solved?
Willow is far from a finished machine. Researchers still need more and better logical qubits, error correction that works at larger scale, and systems that can run long, useful calculations. Rare correlated errors and the need for better decoding are also open problems.
Willow’s real value is the evidence that error correction can improve as a system grows. Read the peer-reviewed Willow paper in Nature.
How Did Quantum Computing Improve in 2024?
Taken together, the eight results point to three clear shifts in the field.
Error correction became more practical
Google showed logical errors falling as codes grew. Microsoft and Quantinuum showed logical qubits that beat the physical qubits beneath them. Harvard, QuEra, and Atom Computing ran logical operations on neutral atoms. A useful quantum computer has to keep errors under control while it works, and these results moved that goal closer.
Quality started to matter more than qubit count
The number of physical qubits does not tell you how useful a processor is. What counts is the quality of its logical qubits, the operations they can perform, and how long they hold information. IBM’s gate-heavy circuits and the photonic work fit this theme from the hardware side.
AI joined the effort
AlphaQubit showed that machine learning can help quantum hardware by decoding its errors. It supports the hardware rather than replacing it, and it still has to get fast enough for real-time use.
How Do the Main Quantum Computing Architectures Compare?
No single hardware design has won. Teams are testing several approaches, and each one has its own trade-offs.
| Architecture | How it stores information | 2024 example | Main strength | Main challenge |
|---|---|---|---|---|
| Superconducting | Tiny electrical circuits cooled to extreme cold | Google Willow, IBM Heron R2 | Fast operations | Noise and cooling needs |
| Trapped ion | Charged atoms held by electric fields | Quantinuum | Precise control and strong gates | Speed and scaling |
| Neutral atom | Uncharged atoms held and moved by lasers | Harvard/QuEra, Atom Computing | Flexible, reconfigurable arrays | Atom loss and control at scale |
| Photonic | Particles of light | Hebrew University cluster states | Fits optical networks | Photon loss and detection |
| Silicon spin | The spin of an electron in silicon | UNSW Sydney and Diraq | Fits chip-making methods | Control and scaling |
Silicon spin qubits are the one design not covered above. In a March 2024 Nature paper, a team from UNSW Sydney and the company Diraq operated them above 1 kelvin. They reported single-qubit gate fidelity up to 99.85% and two-qubit gate fidelity of 98.92%.
Warmer operation matters because it eases the cooling demands of a large machine. Silicon is also the base of the chip industry, so this route could borrow existing manufacturing methods. The work is promising, but control and scale still need to improve.
What Are the Real-World Applications of Quantum Computing?
Quantum computers may help with some problems that are hard for classical computers. Most of these uses are still research goals. None of the 2024 results made quantum machines ready to replace ordinary computers in business or science.
Drug discovery and chemistry
Molecules follow the rules of quantum physics, so quantum computers are a natural fit for studying them. Someday they may help model chemical reactions and test new drug candidates. That will take better hardware and algorithms than exist today.
Materials and batteries
Quantum systems could help scientists model materials at the atomic level, which may support better batteries and catalysts. The aim is to simulate complex systems that classical methods struggle to handle.
Finance
Banks study problems like portfolio selection and risk analysis, and some of these can be framed as optimization tasks. Researchers are testing whether quantum methods can help with parts of that work. There is no broad proof yet that quantum computers beat classical systems in finance.
Cybersecurity
The clearest near-term effect is preparation. NIST’s standards let organizations update their systems before large quantum computers become a threat to current public-key methods.
What Are the Biggest Challenges Still Facing Quantum Computing?
The 2024 results were important, but several hard problems remain.
Noise and decoherence
Qubits lose their quantum state when they interact with their surroundings, an effect called decoherence. Noise also comes from imperfect control, materials, and unwanted interactions. Researchers must cut these errors much further before long calculations can be trusted.
The cost of logical qubits
Each logical qubit needs several physical qubits to protect it. The exact number depends on the error rates, the code, and the level of reliability needed. A chip with many physical qubits may therefore offer far fewer useful logical ones.
Real-time error correction
A quantum computer has to find and fix errors fast enough to keep a calculation going. Better decoders are more accurate, but they must also be quick enough to run alongside the hardware.
Hardware scale and cost
Building a small quantum processor is very different from building a large one. A bigger system needs stable control, reliable connections, and error management across many qubits. Many designs also depend on special cooling systems, lasers, or vacuum equipment, plus expert teams to run them.
No broad practical advantage yet
A quantum processor can beat a classical computer on one special test and still be useless for everyday work. Researchers need to show clear, repeatable benefits on problems that matter in science, industry, or business.
What Is the Future of Quantum Computing After 2024?
The field is moving from counting qubits toward making them reliable. Progress will likely depend on a few connected goals:
- More reliable logical qubits: They need to store information and run operations with very low error rates.
- Fault-tolerant processors: These machines keep working even when individual physical qubits make mistakes, which takes strong error correction and many well-controlled parts.
- Faster error decoding: Decoders like AlphaQubit will need to match the speed of large processors.
- Hybrid computing: Quantum processors will likely work alongside classical computers, handling only the parts of a problem they suit.
The goal is not to replace every computer with a quantum one. It is to find the problems where a reliable quantum processor offers a real benefit.
Frequently Asked Questions About Quantum Computing Breakthroughs in 2024
What was the biggest quantum computing breakthrough in 2024?
There is no single winner, because each result tackled a different problem. Google’s Willow made the clearest advance in error correction by showing that errors fall as codes grow. Harvard and QuEra ran up to 48 logical qubits, and Microsoft and Quantinuum cut logical error rates 800-fold compared with physical ones.
What did Google’s Willow chip achieve?
Willow is a 105-qubit superconducting processor that showed below-threshold error correction. Larger error-correcting codes gave lower logical error rates. Google also reported a fast Random Circuit Sampling result, but that test does not prove an advantage on real-world tasks.
How many logical qubits were demonstrated in 2024?
Harvard, MIT, and QuEra used up to 48. Microsoft and Atom Computing entangled 24 and ran a separate computation on 28. Microsoft and Quantinuum built four logical qubits from 30 physical ones. These came from different systems, so the numbers are not a direct ranking.
What is the difference between physical and logical qubits?
A physical qubit is a hardware unit, such as an atom, an ion, or a superconducting circuit. A logical qubit combines several physical qubits and uses error correction to protect the information they hold.
Can quantum computers break encryption?
Not today. Current machines are far too small and error-prone to attack widely used public-key encryption. A future large, fault-tolerant quantum computer could threaten some of it, which is why NIST finalized post-quantum standards in 2024.
Did quantum computing achieve practical quantum advantage in 2024?
No. There were strong benchmark results, such as Google’s Random Circuit Sampling test, but a benchmark is not a useful business or science problem. Clear, repeatable advantages on practical tasks are still an open challenge.
What is below-threshold quantum error correction?
It means that making an error-correcting code larger lowers the logical error rate. Adding more physical qubits then makes the logical qubit more reliable instead of less.
Which organizations made major quantum computing advances in 2024?
Google, IBM, Microsoft, Quantinuum, QuEra, and Atom Computing all reported important progress. Academic teams at Harvard, MIT, the Hebrew University of Jerusalem, and UNSW Sydney contributed as well. NIST added a separate milestone by finalizing post-quantum encryption standards.
Conclusion
The latest breakthroughs in quantum computing in 2024 brought real gains in error correction, logical qubits, processor design, and quantum security. Quantum computers are not ready to replace classical ones, and noise, scale, and practical performance remain open problems. Still, the field is shifting from counting qubits to making them reliable, and that is what useful machines will need.
You Might Also Like:
How Can You Protect Your Home Computer? A Practical Guide
Yellow Tint Display: Why Your Screen Looks Yellow and How to Fix It
I’m Emma Rose, the founder of Punstation. I cover technology, digital trends, business, and online tools, creating practical guides and thoughtful reviews to help readers understand their options and make informed decisions. For more on how I approach my content, see our Editorial Standards.