Nvidia Expands Voice AI Toolkit Amid Booming Enterprise Demand
Nvidia releases Riva 2.0 to help enterprises build speech recognition and voice generation applications as the voice AI market rapidly expands.
Nvidia highlights its commitment to voice AI by releasing the Riva 2.0 SDK and a managed Riva Enterprise offering during its GTC 2022 conference. These tools allow enterprises to develop and deploy speech recognition and text-to-speech applications. The release accompanies major hardware launches like the Grace CPU Superchip and the H100 GPU, but specifically targets the quickly growing voice technology sector.
The speech and voice recognition market experiences rapid growth as companies seek new ways to improve operational efficiency. Surveys show that a large majority of executives see significant value in AI voice technologies, while major brands like Snap and RingCentral already integrate Riva for live-captioning and developer tools. Nvidia makes the technology more accessible by integrating Riva 2.0 with TAO, a low-code solution that helps data scientists customize speech applications without heavy programming.
Nvidia also pushes into voice generation with Riva Custom Voice, a toolkit that creates human-like cloned voices using only 30 minutes of audio data. This technology allows brands to develop consistent, custom voices for various digital interactions. As speech AI expands into new enterprise applications, Nvidia plans to bring these workflows to embedded platforms and continue simplifying the development process for future customers.