Ziad Asghar, Qualcomm's vice president of product management, responsible for Snapdragon product technology and planning, said that Qualcomm's launch of processor products combined with self-generative artificial intelligence will not only transform the user experience of mobile phones, PCs and other devices, but will also drive new application development opportunities, thereby creating a win-win situation for users, industry providers of large-scale natural language models, and Qualcomm.
At the Democratization of Artificial Intelligence (AI) workshop on the final day of the Snapdragon Tech Summit 2023, Ziad Asghar explained that both the Snapdragon 8 Gen 3 and Snapdragon X Elite processor designs incorporate large-scale natural language models from numerous industry vendors. Leveraging the processor's inherent computing power and pre-trained results, these processors enable devices using these processors to handle a wide range of automated AI computations.
At the same time, the processor's low power consumption, design flexibility for use in various devices, and its compatibility with 5G, high-speed Wi-Fi, and Bluetooth connectivity not only provide sufficient device-side AI computing performance, but also integrate with the cloud for more complete collaborative computing. For many developers, startups, or small and medium-sized enterprises, this can actually save more costs on purchasing AI computing equipment or building automatically generated AI, and can more quickly introduce various types of AI computing into their services and product applications.
Even through the development tool resources provided by Qualcomm, many companies and startups can invest in the deployment of self-generated artificial intelligence application services without having to possess extensive technical development capabilities, which means that the subsequent self-generated artificial intelligence will be able to accelerate its popularization like mushrooms after rain.
Currently, deploying various types of automatically generated artificial intelligence usually requires purchasing artificial intelligence servers at a higher cost and also requires certain technical construction capabilities. For many developers, startups and small and medium-sized enterprises, there is a relatively high entry barrier.
Even if public clouds or application service providers are used to provide resources to quickly invest in the development of automatically generated artificial intelligence applications, this also means that sensitive content such as data will be stored in an external environment, or must be executed through collaborative computing in the cloud, resulting in a situation where one must be highly dependent on others.
Ziad Asghar believes that Qualcomm's newly released processor design will benefit developers, startups, and small and medium-sized enterprises, helping them more quickly integrate generative AI computing resources into their services. Furthermore, for companies providing various large-scale natural language models, their accuracy will improve with increased usage and scale. Qualcomm will profit from providing this solution, creating a win-win situation for all three parties.
Ziad Asghar also emphasized that the solutions provided by Qualcomm are based on an inclusive and trustworthy open source model, so there will be no ethical issues such as discrimination and injustice. He also emphasized that Qualcomm will work with industry players who provide various open source models to ensure the security of using this solution.
