Building quantum error-correcting codes once meant hours of complex calculations. Now a Duke PhD Candidate has created a visual platform that cuts that time to milliseconds—while making the process as intuitive as snapping together toy blocks.
Duke PhD Candidate Bálint Pató, working with June Vanlerberghe, built the toolkit PlanqTN—pronounced "plankton" and cleverly combining Planck, quantum, and Tensor Networks (TN). The tool tackles quantum error correction: the task of protecting fragile quantum information from environmental interference. The name also nods to how individual tensors resemble tiny plankton organisms.
"This framework allows you to construct larger, more complex codes, like the Steane code, from smaller building blocks," Pató explained in his YouTube announcement.
The platform builds on the quantum Lego formalism developed by Charles Chubb and Brad Lackey—a modular mathematical approach to composing complex codes from simpler pieces.
PlanqTN's workflow is highly visual, Pató demonstrates, and extends traditional code concatenation, streamlining tracking logical operators via operator pushing—systematically following how logical operations transform as smaller building blocks are connected. A key payoff is much faster weight enumerator calculations—counting how many codewords exist at each "weight" (number of non-zero elements), crucial for evaluating a code's error-correcting strength.
Building Blocks for Quantum Reliability
Think of error-correcting systems as protective shields for delicate data. Traditionally, creating them meant wading through dense mathematics. With PlanqTN, researchers drag tensors (multi-dimensional mathematical objects) onto a digital canvas.
Users can instantly inspect stabilizer generators (error-detection rules) and logical operators (the system's ones and zeros). Linking these components creates tensor networks—webs of interconnected parts that work together.
And here's where it gets impressive: distance calculations that once took an hour now finish in 0.1 seconds, according to Pató's benchmarks. In quantum coding, a code's distance is the number of physical-qubit errors required to cause an undetected logical error—higher distance means stronger protection.
Tests used rotated surface codes, a resource-efficient surface-code variant. For example, a distance-5 rotated surface code (corrects up to two errors) completed its weight enumerator in 0.1 s—a 36,000-fold speed-up over brute-force methods.
From Abstract Math to Visual Design
Running calculations is flexible, according to presentation video. For lighter jobs, PlanqTN can run in the cloud via Google Cloud Run. For heavier work, researchers can spin up a local kernel with Docker and Kubernetes for consistent performance.
"The tensor network method used by PlanqTN is far faster than a brute-force approach," Pató noted.
A smaller distance-3 setup—capable of correcting a single error—runs in about 0.03 seconds, Pató reports. The distance-5 version still finishes in 0.1 s, versus roughly an hour using traditional methods.
Open Source, Open Invitation
The PlanqTN team—Pató and June Vanlerberghe, with guidance from Charles Cao, Brad Lackey, and Kenneth R. Brown—built the platform collaboratively. The project is in early days and the team is actively seeking contributors on GitHub. PlanqTN is open source under the Apache 2.0 license and free for all to use.
"We would love to hear your ideas and welcome more contributors," he said. "I hope you'll explore PlanqTN's features, learn something new, and help us grow this project."
By turning error correction from an exclusive mathematical craft into a visual, hands-on process, PlanqTN opens the door for more people to contribute to quantum computing's future.
Note
Watch Bálint Pató's full demonstration and announcement on YouTube to see PlanqTN in action.
