Telstra, SQC Train Quantum AI in Days, Matching Deep Learning for Network Forecasts
Updated
Updated · Quantum Zeitgeist · Jun 6
Telstra, SQC Train Quantum AI in Days, Matching Deep Learning for Network Forecasts
1 articles · Updated · Quantum Zeitgeist · Jun 6
Summary
A 12-month Telstra-Silicon Quantum Computing trial found the “Watermelon” quantum reservoir could forecast network latency and bandwidth with accuracy comparable to a deep learning model, but trained in days rather than weeks.
The speed gain came from Watermelon generating quantum features through its internal quantum dynamics, avoiding the heavy GPU demands typically needed to train conventional deep learning systems.
Telstra tested the system on predictive analytics used to manage network resources proactively, aiming to improve performance, prevent issues earlier and lift customer experience.
SQC and Telstra cast the result as a step from lab research toward commercial quantum use in telecoms, and a basis for further work on digital infrastructure applications.
Australia's SQC is proving quantum's real-world value. How will this accelerate the global race for commercial quantum advantage?
Quantum AI trained in days, not weeks. How will this speed advantage transform industries beyond just telecommunications?
Telstra and SQC’s Quantum Reservoir "Watermelon" Delivers Deep Learning-Level Network Forecasts in Days, Not Weeks
Overview
Telstra and Silicon Quantum Computing (SQC) partnered in a year-long project that moved quantum computing from the lab into real-world telecommunications. Their main goal was to see if SQC’s quantum reservoir system, called Watermelon, could predict network performance as well as Telstra’s deep learning models. The results were impressive: Watermelon matched the accuracy of traditional AI models while working much faster. This breakthrough marks a major step for Australia’s digital infrastructure, showing how quantum machine learning can bring practical benefits to network analytics and set the stage for future innovation in the industry.