Abstract
Analogue computing uses the physical behaviours of devices to provide energy-efficient arithmetic operations. However, scaling up analogue computing platforms by simply increasing the number of devices leads to challenges such as device-to-device variation. Here we report scalable analogue computing and neural networks in the synthetic frequency domain using an integrated nonlinear phononic platform on lithium niobate. This synthetic-domain computing is robust to device variations, as vectors and matrices are concurrently encoded at different frequencies within a single device, achieving a high throughput per area. Leveraging inherent nonlinearities, our device-aware neural network can perform a four-class classification task with an accuracy of 98.2%. The nonlinear phononic computing hardware also maintains consistent performance over a wide operational temperature range (characterized up to 192 °C). Our synthetic-domain computing combines single-device parallelism, inherent nonlinearity and environmental stability, and could be of use in edge computing applications in which power efficiency and environmental resilience are crucial.
| Original language | English |
|---|---|
| Pages (from-to) | 698-708 |
| Number of pages | 11 |
| Journal | Nature Electronics |
| Volume | 8 |
| Issue number | 8 |
| DOIs | |
| State | Published - Aug 2025 |
Funding
We thank Rohde & Schwarz for support with the microwave instrumentation. Device fabrication was conducted at the Center for Nanophase Materials Sciences (CNMS2022-B-01473 and CNMS2024-B-02643, L.S.), which is a US Department of Energy, Office of Science User Facility. Research was partially supported by the Air Force Office of Scientific Research (AFOSR) under grant no. W911NF-23-1-0235 (L.S.) and award no. FA9550-22-1-0548 (W.X.), and by Commonwealth Cybersecurity Initiative in Virginia (W.X.). Development of the optical vibrometer was partially supported by the Defense Advanced Research Projects Agency (DARPA) OPTIM program (HR00112320031, L.S.). Development of the nonlinear phononic device and material calculation were partially supported by DARPA SynQuaNon DO program under agreement no. HR00112490314 (L.S.). The work at the University of Texas at Dallas is supported by the Office of Naval Research (ONR) under grant no. N00014-23-1-2020 (W.G.V.). The views and conclusions contained in this document are those of the authors and do not necessarily reflect the position or the policy of the United States government. No official endorsement should be inferred. Approved for public release.