Telstra and Silicon Quantum Computing (SQC) have demonstrated progress in applying quantum machine learning to real-world challenges, with a joint team training SQC’s ‘Watermelon’ quantum reservoir system in just a few days. The 12-month effort moved quantum computing beyond laboratory settings and into the telecommunications industry, focusing on predictive analytics to improve network performance. Telstra engineers evaluated Watermelon’s ability to predict key metrics such as latency and bandwidth, with the aim of proactively managing resources and improving customer experience. The quantum-enhanced model ultimately matched the accuracy of a deep learning model trained over several weeks. Shailin Sehgal, Group Executive, Global Networks and Technology at Telstra, said: “Combining Telstra’s experience in complex connectivity management with SQC’s quantum systems proves that combining deep domain knowledge and engineering can lead to innovations that potentially have real-world customer impact.”
Watermelon Reservoir accelerates network metric prediction
A quantum reservoir system named Watermelon has significantly accelerated the training of machine learning models used to predict network performance at Telstra, demonstrating a potential pathway to real-world quantum applications. During a 12-month collaboration, Telstra and Silicon Quantum Computing (SQC) engineers evaluated SQC’s quantum repository, Watermelon, against recently developed deep learning models tasked with predicting network metrics such as latency and bandwidth. The experiment went beyond a theoretical exploration to test quantum computing within the telecommunications industry. The results revealed that Watermelon achieved comparable accuracy to existing deep learning models, while significantly reducing training time. The quantum system completed training in a few days, compared to weeks for traditional models. This efficiency stems from Watermelon’s ability to generate quantum features used within AI models, and it was also able to operate without the heavy GPU hardware demands typically associated with deep learning.
Michelle Simmons, CEO of Silicon Quantum Computing, called the results “an exciting and important step forward in the commercial implementation of quantum technology,” adding, “Watermelon’s quantum feature generation helps uncover complex relationships in classical data while significantly reducing training time.” This trial establishes the foundation for further exploration of the role of quantum technology in enhancing digital infrastructure and improving customer outcomes.
Quantum storage dynamics is different from deep learning
Telstra’s quest to enhance predictive analytics within its network infrastructure led to a comparison of traditional deep learning and new quantum approaches. Specifically, Silicon Quantum Computing (SQC)’s “Watermelon” quantum reservoir system. While traditional deep learning models are effective, they require significant training periods, often several weeks, and significant processing power from graphics processing units. SQC’s quantum reservoir offered markedly different dynamics. This speed advantage comes from the reservoir’s reliance on internal quantum mechanics rather than incremental statistical learning, a property that also makes it potentially more robust to incomplete or noisy data. This efficiency is critical as artificial intelligence becomes increasingly resource-intensive and suggests a potential pathway to more sustainable and accessible AI applications within complex systems like communication networks.
We have always believed that the key to unlocking quantum potential lies in building systems with atomic precision and purity. This partnership demonstrates how quantum processors are moving beyond theory to practical, scalable solutions that power Australia’s digital infrastructure.
Professor Michelle Simmons, Silicon Quantum Computing CEO
Telstra and SQC demonstrate training reduction with quantum systems
Silicon Quantum Computing (SQC) has partnered with Telstra to demonstrate a significant reduction in the training time required for artificial intelligence models used in network management, advancing quantum computing applications beyond theoretical research. The collaboration, which will last over 12 months, will focus on applying quantum machine learning to Telstra’s predictive analytics systems to proactively optimize resource allocation by predicting network metrics such as latency and bandwidth. Telstra is now employing deep learning and AI to predict network performance, detect anomalies and enable pre-emptive maintenance and automated adjustments. SQC’s ‘Watermelon’ quantum reservoir system was at the heart of the trial. The system generates quantum features aimed at improving the accuracy of AI models.
From increasing personalization to preventing problems, we’re always looking at technology that can help create smarter connected experiences for our customers. Quantum computing is a promising frontier that we are exploring.
Quantum computing leverages superposition and entanglement for complex problems
The successful trial between Telstra and Silicon Quantum Computing (SQC) was more than just a demonstration of quantum computing. It highlighted how fundamental quantum mechanics, superposition and entanglement, can be used to address real-world challenges in network management. Unlike classical computers, which are limited to bits that represent 0 or 1, quantum computers utilize qubits that exist in multiple states simultaneously through superposition, allowing for vast possibilities of parallel processing. This, coupled with the instantaneous information sharing of entangled qubits, provides a fundamentally different approach to computing particularly complex problems. SQC’s ‘Watermelon’ quantum reservoir system, which is at the heart of the Telstra collaboration, operates on different principles than traditional deep learning.
This is an exciting and important step forward in the commercial implementation of quantum technology. In collaboration with Telstra, we were able to test our quantum reservoir system Watermelon in a real-world telecommunications environment. This is something that only a few quantum companies have achieved.
