OnQ Blog

Snapdragon and Google Cloud Vertex AI NAS are committed to making your AI faster and smarter [video]

Dec 16, 2021

Qualcomm products mentioned within this post are offered by Qualcomm Technologies, Inc. and/or its subsidiaries.

Hopefully you had a chance to follow all the latest and greatest announcements during our Snapdragon Tech Summit earlier this month. We showcased a depth of advancements coming from our Snapdragon 8 Gen 1 mobile platform and our new 7th Generation Qualcomm AI Engine. The Snapdragon 8 Gen 1 is an AI powerhouse, delivering boundless AI to your favorite devices with a suite of new AI use cases and AI performance boosts.

AI software is the center of making AI faster and this year we are thinking outside of the box, exploring new methods on taking AI performance to the next level. One of the more notable examples of this is the strategic collaboration with Google Cloud, on Vertex AI NAS (Neural Architecture Search) to accelerate neural network deployments.

What is NAS?

The mobile industry is constantly moving forward with AI everywhere. It is crucial that we create and optimize new networks without wasting time. Neural networks are powerful and adaptable models that excel at a variety of tough learning tasks such as image, audio, and natural language processing. Despite being vastly available, neural networks remain difficult to construct. NAS was designed to use artificial intelligence to help create and optimize networks automatically by creating the best neural topology for your specific hardware or performance targets.

NAS automates the design of neural network models. For example, consider a video bokeh model within a video conferencing app. By looking at the hardware that it’s going to use along with other design parameters, NAS can help engineers quickly reach an optimal design. 

7th Generation AI Engine NAS Google

Dec 13, 2021 | 0:52

NAS considers design targets, network types, frames per second, size and accuracy. In addition, NAS is hardware aware, meaning it basically learns and takes a blueprint of the hardware, and using AI, NAS leverages the hardware characteristics to produce an optimal result. NAS can take your initial network as a starting point and search through and optimize it, enabling it to run faster on Snapdragon 8 Gen 1 within your targeted constraints. 

All this means is your apps using AI will run faster, smoother and your experiences will get a whole lot better with devices powered by Snapdragon. (FYI, pretty much all of your apps today have some sort of AI running.)

Qualcomm Technologies will be the first SoC manufacturer to offer Google Cloud Vertex AI NAS

Working with Google, we will be integrating Vertex AI NAS into the Qualcomm Neural Processing SDK to enable OEMs/Ecosystem to create, test, and deploy unique and efficient experiences on the edge — in just weeks — while using significantly fewer resources. We will also maximize performance for AI application on Snapdragon platforms with Vertex AI NAS optimized neural nets and other Snapdragon products including XR, automotive, Windows with Snapdragon technology and IoT.

We are pleased to bring NAS to Snapdragon technology and integrate it into our 7th generation Qualcomm AI Engine – to help engineers deliver the best possible AI solutions, and so that you can have the mobile experiences you deserve.

 

Snapdragon, Qualcomm Neural Processing SDK, and Qualcomm AI Engine are products of Qualcomm Technologies, Inc. and/or its subsidiaries.

 

Opinions expressed in the content posted here are the personal opinions of the original authors, and do not necessarily reflect those of Qualcomm Incorporated or its subsidiaries ("Qualcomm"). Qualcomm products mentioned within this post are offered by Qualcomm Technologies, Inc. and/or its subsidiaries. The content is provided for informational purposes only and is not meant to be an endorsement or representation by Qualcomm or any other party. This site may also provide links or references to non-Qualcomm sites and resources. Qualcomm makes no representations, warranties, or other commitments whatsoever about any non-Qualcomm sites or third-party resources that may be referenced, accessible from, or linked to this site.

Jeff Gehlhaar

Vice President of Technology, Qualcomm Technologies

Vinesh Sukumar

Senior Director, Product Management, Qualcomm Technologies