Nvidia has released the code and training tools for the Ising Decoder ColorCode 1 Fast. This AI module pre-processes error signals before passing them to the Chromobius decoder, the company reports.
In simulations, this combination reduced logical error rates by 347.7 times and increased processing speed by 7.3 times compared to using Chromobius alone. The results were achieved on a quantum memory model with a code distance of 31 and a physical error rate of 0.3%. The test was conducted on synthetic data, not on an operational quantum processor.
The Ising Decoder ColorCode 1 Fast is a 17-layer three-dimensional convolutional neural network with approximately 2.9 million parameters. Its receptive field is 13, and it was trained using input arrays sized 13 × 13 × 19.
The model functions not as an independent decoder but as a preliminary processing stage. It analyzes local error signals, reduces their number, and passes the remaining sparse map to the classical Chromobius decoder.
Color Codes Move Closer to Real-Time Application
Quantum error correction allows for the combination of multiple unstable physical qubits into more reliable logical ones. The decoder analyzes the results of control measurements to determine which errors need correction.
Surface codes are often used for storing quantum information due to their relatively high error threshold and ease of decoding.
Color codes enable more efficient execution of certain logical operations, but their error signals are more complex to process. According to Nvidia, the lack of fast and accurate decoders has long hindered the real-time application of such codes.
The Ising Decoder ColorCode 1 Fast is expected to reduce the load on the main algorithm. The authors of the study stated that the advantage of the combination with Chromobius increases as the code distance grows.
However, the speed comparison was conducted on different types of hardware. The neural network ran on Nvidia DGX GB300, while Chromobius was executed on a Grace Neoverse-V2 processor. Therefore, the 7.3-fold speed increase reflects not only algorithmic differences but also the use of a GPU instead of a CPU.
Nvidia has published the framework and training recipes in an open repository under the Apache 2.0 license.
The Ising family of open models was introduced by the company in April. It includes tools for calibrating quantum processors and error correction.
In June, IBM unveiled an updated roadmap, according to which the company plans to create the world’s first large-scale fault-tolerant quantum computer by 2029.
