By 
Qualcomm’s ventures into pushing the boundaries of mobile processing technology has been rather impressive to say the least, and it seems that the company is further improving its efforts into working with AI. As such, the company recently showcased the world’s first demonstration of Stable Diffusion on an Android device, a feat unheard of, at least until now.
For those not in the know, Stable Diffusion is a deep learning, text-to-image foundation model first developed in 2022, commonly used to generate detailed images based on text descriptions, as seen on the likes of “DALL-E” and other similar online platforms – think of a highly-intelligent piece of software that can paint you a picture upon hearing your command. With that said however, a certain limitation that Stable Diffusion faces is its confinement to running via a cloud system – or in other words, “online.”
Qualcomm’s AI Research team recently performed full-stack AI optimizations using the Qualcomm AI Stack to deploy Stable Diffusion on an Android smartphone for the very first time, specifically one powered by the company’s Snapdragon 8 Gen 2 platform. Qualcomm’s AI stack software uses artificial intelligence with support for a wide range of intelligent devices with broader AI software access and compatibility, such as mobile, automotive, IoT and cloud platforms, to name a few.
Qualcomm’s detailed article on how this procedure was made possible is a rather technical read, but this demonstration essentially allowed a smartphone to generate an image render based on text descriptions in under fifteen seconds, a feat that even took longer on most cloud-based Stable Diffusion programs on the internet. It also opens the door for stable diffusion to run locally on hardware easily accessible to the public.
What does this all mean, then? Ideally, the optimizations for Stable Diffusion to run efficiently on phones means that this same technology can also be used for other platforms like laptops, XR headsets, and virtually any other device powered by Qualcomm Technologies, reducing costs which otherwise would be significantly higher on cloud-based alternatives. All of this points towards the enhancement of AI and its integration into Qualcomm products, which the company hopes will enrich how users live.
Source: Qualcomm
Comments