AudioSourceRE trains AI models 20% faster on OCI compared to cloud competitors
The professional audio company turns to Oracle Cloud Infrastructure (OCI) to create cutting-edge audio AI models.
“OCI allowed us to train our models faster and better, and at a lower cost.”
Originally part of the Nvidia Inception Program, AudioSourceRE is an audio separation company that’s pioneering advancements in sound engineering through cutting-edge AI models and algorithms. Audio separation involves splitting an audio recording or stream into its underlying sources. Because the company’s audio separation technology is based on deep-learning models, it needed to train AI models to be able to deliver the best audio separation quality to customers. AudioSourceRE found that Oracle Cloud Infrastructure (OCI) with OCI Compute Virtual Machines (VMs) provided the best balance of reduced training time and costs in comparison to other cloud offerings. After the implementation, AudioSourceRE was able to train models 20% faster and for a better price compared to its previous cloud platform. Now, the company can innovate more quickly, set new trends, and create new sound solutions across the music and film industries.