"Nvidia's new quantum models, including Ising Calibration, are designed to help developers minimize noise in quantum systems, which currently generate errors roughly once in every thousand operations."
"The Ising Calibration model, with 35 billion parameters, can be integrated into an automated framework to adjust settings and reduce error rates significantly, akin to quantum autotune."
"Ising Decoding models, available in two sizes, are designed to detect and correct errors in real time, employing a convolutional neural network architecture for enhanced performance."
Nvidia introduced new AI models to assist quantum hardware developers in reducing error rates in quantum processors. Current quantum systems experience errors approximately once per thousand operations, necessitating a reduction by a factor of a billion for practical use. The Ising Calibration model helps optimize settings to minimize noise, while Ising Decoding models detect and correct errors in real time. These models utilize a convolutional neural network architecture and are designed to operate efficiently on Nvidia hardware, improving error detection speed significantly.
Read at Theregister
Unable to calculate read time
Collection
[
|
...
]