Quantum Error Correction Faces Unexpected Complexity from Memory Effects
New research suggests quantum computer errors are more correlated than previously assumed.
1Quantum Error Correction Hurdles
Quantum computers rely on error correction to function reliably, yet recent studies reveal that the landscape of errors is unexpectedly complex. Researchers have discovered that the fundamental assumptions guiding error-correction models, specifically the Markov assumption, are failing in practice. This realization poses significant challenges for achieving practical, fault-tolerant quantum computing systems.
2The Markov Assumption Flaw
The Markov assumption posits that qubit errors are memoryless and uncorrelated over time. However, experiments consistently show that qubits are sensitive to environmental interactions that create memory effects. These effects mean that a qubit's performance can change based on its history, effectively breaking the mathematical models used to design current error-correction protocols.
3Experimental Evidence of Defects
Ioan Pop at the Karlsruhe Institute of Technology developed a measurement method uncovering long-lived defects in superconducting qubits. These spurious quantum states, caused by material imperfections or stray molecules, act as memory sources. Similar violations of the Markov assumption were observed by Rami Barends and his colleagues, confirming these issues are widespread.
4Analyzing Google and IBM
In 2025, researchers including John Fidel Kam and Kavan Modi analyzed Google’s Willow quantum computer, finding errors consistent with non-Markovian models. Additionally, Gregory Quiroz at Johns Hopkins University found that models incorporating non-Markovian effects represented IBM quantum computers seven times better than standard noise models. These findings challenge the current industry-standard error-correction frameworks.
5New Types of Errors
Lorenza Viola at Dartmouth College identified that qubit environments can act as quantum systems that actively respond to the qubit itself. This introduces a novel category of errors that existing correction approaches do not account for. As quantum hardware scales, these previously ignored environmental effects may become critical obstacles to achieving long-term stability.
6Shifting Burden of Proof
Amit Hagar at Indiana University Bloomington notes that the burden of proof has shifted to firms building quantum computers. Since their promises were based on conditional theorems that may not hold, the viability of these machines is under scrutiny. Standard error-correction might only work for limited computations, undermining the goal of arbitrary, long-running programs.
7Evolving Correction Strategies
While the situation is serious, researchers are exploring new paths forward. Muhammad Usman highlights that error-correction codes are evolving, with QLDPC codes gaining traction. It remains unclear how these new codes handle correlated noise, but they represent a shift away from the 1990s-era standards that have dominated the field until very recently.
8Future Outlook and Mitigation
Experts suggest that non-Markovian errors are not necessarily unsolvable if taken seriously. New measurement techniques may allow teams to identify and avoid memory effects, while conventional control algorithms could be adjusted to mitigate correlated noise. The path to fault-tolerance requires adapting error-correction codes to account for these complex, real-world environmental interactions.