Telstra and Silicon Quantum Computing (SQC) reported results from a year-long collaboration that applied quantum-enhanced machine learning to predictive analytics in telecommunications networks. The project focused on forecasting network performance metrics such as latency and bandwidth—an operational challenge that underpins proactive fault prevention, resource optimization, and personalized connectivity services.
The joint team evaluated SQC’s quantum reservoir system, known as Watermelon, which generates quantum-derived features for use in AI models. Engineers assessed whether these features could forecast network metrics effectively and compared the results with a recently developed deep learning model used by Telstra. According to the companies, the quantum-enhanced approach matched the accuracy of the deep learning model while significantly reducing training and fine-tuning time—from weeks to just days.
The trial also highlighted infrastructure implications. The quantum reservoir operated without the GPU-intensive hardware typically required for deep learning, pointing to potential reductions in computational cost and energy consumption. Telstra and SQC said the results support further investigation into how quantum technologies could complement existing AI systems in live network environments and broader digital infrastructure applications.
- Evaluated quantum machine learning for time-series network prediction
- Used SQC’s Watermelon quantum reservoir to generate AI features
- Achieved accuracy comparable to deep learning models
- Reduced model training time from weeks to days
- Avoided GPU-heavy hardware requirements
- Established a foundation for future quantum-enabled network analytics
Shailin Sehgal, Telstra Group Executive of Global Networks and Technology, said: “This trial shows how quantum capabilities could complement our existing systems and technology to deliver faster insights and better outcomes for our customers. The collaboration with SQC demonstrates how Australian industries and homegrown innovation can work together to shape the nation’s digital future.”
🌐 Analysis
The Telstra–SQC results add to growing evidence that hybrid quantum–classical approaches may deliver near-term value before large-scale fault-tolerant quantum computers arrive. Similar quantum machine learning experiments are emerging globally, as telecom operators and cloud providers explore lower-cost, faster-training alternatives to GPU-intensive AI for network optimization.
Silicon Quantum Computing (SQC) is an Australian quantum technology company focused on building commercial-scale quantum processors using silicon-based qubits fabricated with atomic precision. The company develops quantum chips by placing individual phosphorus atoms in silicon, a manufacturing approach designed to deliver high-fidelity qubits and scalable architectures compatible with established semiconductor processes. SQC has demonstrated integrated quantum circuits at the atomic scale and reported high-efficiency implementations of core quantum algorithms, including Grover’s search. Alongside its long-term hardware roadmap, SQC is advancing near-term applications through systems such as its Watermelon quantum reservoir, which applies quantum dynamics to machine-learning workloads in real-world industry environments.
