Weather QML · FPGA inference
Hybrid VQC Weather Forecasting with FPGA Inference
7-qubit variational circuit for 24-day temperature forecasting, with a custom kernel on an AMD Alveo FPGA for low-latency edge inference.
Published · IEEE TQCEBT 2026
- 7qubits
- 50parameters
Interactive abstract
Play with the idea
This interactive diagram needs JavaScript. The schematic and write-up below explain the same idea.
Layman view
In plain words
Draft — to be verified. Written in plain language from the paper and project notes on this site.
Weather forecasting means predicting the future from today's measurements. This project asked whether a small quantum circuit could learn temperature patterns 24 days ahead from daily reanalysis records (NCMRWF data), and whether the trained model could answer quickly enough to run on a small device at the edge.
A variational quantum circuit works like a tunable recipe. Weather measurements are written into the rotation angles of 7 qubits, a fixed pattern of entangling gates mixes them, and 50 adjustable settings are tuned during training until the circuit's measured outputs match the recorded temperatures.
Once trained, the recipe is frozen. The same calculation was then written as a custom circuit for an FPGA, a chip whose wiring can be programmed for one job, so that forecasts come back with low delay.
Expert view
The work, step by step
-
01
Problem
24-day-ahead temperature forecasting is hard: atmospheric dynamics are nonlinear and labelled data is limited. The question was whether a compact variational quantum circuit, trained on daily NCMRWF reanalysis data, could learn useful long-range patterns, and whether its inference could run fast enough for edge use.
-
02
Approach
The model is a 7-qubit variational quantum circuit with 50 trainable parameters, using angle encoding to map meteorological variables onto the qubits and two StronglyEntanglingLayers as the ansatz. It was trained, validated and tested in PennyLane on a CPU-based quantum simulator, reaching a test RMSE of 3.01 K and R² of 0.801.
-
03
Acceleration
For low-latency edge inference, the frozen model runs on a custom Vitis HLS kernel on an AMD Alveo U55C FPGA; the kernel exchanges batches of input features, pre-trained weights and expectation values with the host over HBM.
-
04
Result
A hybrid VQC forecaster with low-latency FPGA inference for edge use. Published at IEEE TQCEBT 2026.
Schematic
How it works
Contribution
My role
Co-author (with first author N. Khan), IEEE TQCEBT 2026.
- NCMRWF reanalysis data—
- Angle encoding—
- VQC: 7 qubits, 50 parameters—
- Training in PennyLane—
- FPGA kernel on Alveo U55C—
Publication
The paper
2026 · Conference paper
Hybrid variational quantum circuits based temperature forecasting with low-latency FPGA inference
N. Khan, S. Siyanwal, K. Islur, A. Tiwari, H. Gupta
IEEE TQCEBT 2026, pp. 1–6
Abstract
A hybrid quantum machine learning approach for 24-day temperature forecasting from daily NCMRWF reanalysis data, using a 7-qubit variational quantum circuit with Strongly Entangling Layers and 50 trainable parameters, trained on a CPU-based quantum simulator, with low-latency inference on an AMD Alveo U55C FPGA kernel.