In case you have not seen, synthetic intelligence in a single type or one other has emerged as a key cog in all types of machines, from smartphones and blended actuality headsets, to a plethora of IoT gadgets and vehicles (and all the pieces in between). AI has even discovered its solution to toasters, a seemingly easy kitchen equipment.
The purpose we point out all it’s because it is the driving drive behind Qualcomm’s new AI Stack portfolio, which is basically an end-to-end AI software program basis to develop an assortment of merchandise, each large and small. Or put one other method, Qualcomm invitations builders to ‘come one, come all in your native AI frameworks and our instruments will settle for your stuff’, to paraphrase the announcement.
“This comprehensive Qualcomm AI Stack portfolio is a truly revolutionary step in improving AI development for OEMs and developers and achieving high performance across our product portfolio,” stated Ziad Asghar, vp of product administration, Qualcomm Technologies, Inc. “We’re delighted to extend our one technology roadmap with this new portfolio and expand our leadership in the Connected Intelligent Edge.”
So what precisely is at play right here? Let’s take a look…
What this entails is a complete portfolio providing builders direct entry to Qualcomm’s AI Engine and devoted AI cores on Qualcomm Cloud AI 100. And as a part of this new providing, the Qualcomm AI Engine can now scale throughout each AI accelerator throughout a variety of the corporate’s applied sciences.
“Qualcomm AI Engine direct is an AI library that delegates and deploys existing models directly to the AI accelerators on Qualcomm Technologies platforms and will now provide OEMs and developers the ability to develop one feature, then move the same model across different products and tiers,” Qualcomm explains.
Circling again to what we said earlier, AI is all over the place today, however it’s not similar assortment of applied sciences throughout the board. Your smartphone, for instance, processes speech recognition to show instructions into reminders or to lookup instructions, whereas an XR headset has sensors to trace your actions. And within the car sector, there’s a complete completely different set of sensors, resembling LIDAR.
From high to backside, AI Stack helps widespread AI frameworks resembling TensorFlow, PyTorch, ONNX and runtimes together with TensorFlow Lite, TensorFlow Lite Micro, ONNX runtime. It additionally consists of inferencing SDKs resembling Qualcomm’s Nueral Processing SDK for Android and, extra just lately, Windows as nicely.
Part of the proposition right here is {that a} developer can construct a framework for one sort of system or platform, after which have the liberty to extra simply move throughout nearly of Qualcomm’s merchandise and tiers utilizing a singular AI software program bundle. It’s a easy idea, however complicated in execution—it took Qualcomm a number of years to develop its AI Stack, in addition to cross collaboration between its software program and {hardware} groups.
“Qualcomm AI Stack can be used directly by OEM and developers to develop on mobile and compute products. It is also a building block for our domain specific SDKs that are available today, for example, the Snapdragon Ride SDK, the Intelligent Multimedia SDK and the Snapdragon Spaces SDK,” Qualcomm says.
Qualcomm AI Stack is commercially accessible instantly.