Linux offering for IoT development with LTS Kernel, multi-SoCs support, focused on upstream compliance. Qualcomm Linux is a comprehensive package of software, tools, and documentation meticulously designed for Qualcomm's IoT platforms. It offers a unified Linux distribution that caters to multiple System-on-Chips (SoCs), featuring essential components such as the UEFI bootloader, the latest Long-Term Support (LTS) kernel, Yocto, and carefully selected drivers. This ensures a consistent and seamless developer experience across all targets. The Qualcomm Linux software stack extends support to all processors, subsystems, and components within the platform, encompassing the CPU, GPU, VPU, DPU, NPU, PMICs, and more. It extensively utilizes upstream open-source software (OSS) while incorporating

Qualcomm Technologies, Inc. has published the abstract syntax tree (QAST) dataset used to support the experiments described in the paper, “Simulating Execution Time of Tensor Programs Using Graph Neural Networks.” We hope this new dataset, comprising 12 unique conv2d workloads, will benefit the graph research community and raise interest in optimizing compiler research.Find out more about theQAST Dataset of Tensor Programs Execution Timesand see how you can use it in your research projects.For any questions or technical support, please contact us atresearch.datasets@qti.qualcomm.com

Sound design is an important part of creating the six degrees of freedom (6DoF) required for the best immersive experience for your users. The Qualcomm® 3D Audio Plugin is a binaural spatial audio plugin for Unity, ideal for XR(AR/VR) experiences and games.

The Qualcomm® Autonomy and Advanced Driver Assistance (ADAS) SDK (previously known as FastADAS) is designed to allow your Snapdragon Ride™ platform and ADAS systems to operate at peak efficiency and minimum latency. Developers can use this SDK to optimize and reduce the end-to-end latency of their most critical pipelines to accelerate their algorithms and systems. The SDK is a C/C++ programming API comprised of binary libraries and header files.The Qualcomm ADAS SDK is comprised of many features and is continually growing. Details of current features can be found in theAPI Referencesection.

Edge Impulse is an end-to-end edge AI platform that empowers developers to build, train, and deploy machine learning models directly on embedded devices rather than relying on cloud infrastructure. It offers a comprehensive toolchain for Sensor data collection, Model training using algorithms like neural networks, Optimization for low-power hardware, and Deployment to edge devices such as microcontrollers

A portfolio of Windows software that is supported on our IoT platforms to accelerate and run your Windows-based apps on devices powered by our Dragonwing IoT chipsets.

Android on Dragonwing is an enterprise-ready Android OS, backed by a board support package (BSP) with exclusive optimizations that work across our portfolio of Qualcomm DragonwingTM processors. Use Android on Dragonwing to take advantage of 5G, Wi-Fi, and Bluetooth® wireless connectivity, and its high-performance CPU, GPU, and NPU, to build responsive, power-efficient solutions for industrial, retail, and edge AI.

The Qualcomm Automotive Connectivity team has developed a Connectivity Thermal Tool to support customers in evaluating thermal performance under various conditions. Contact Lance Watson and Santosh Kumar Tirunagari who supported this tool. His team can investigate why the customer got the response for support on this.

The Snapdragon Game AI SDK is Qualcomm’s first dedicated gaming toolkit, enabling developers to build real‑time, low‑latency, on‑device AI experiences on Snapdragon‑powered Android and Windows devices. It provides a set of local AI inference features optimized for Qualcomm’s Hexagon NPU and integrated with the Unreal Engine™ Andy Michel

A plugin for Unity to easily migrate your Snapdragon Spaces™ project to Android XR

These Graphics Drivers help unlock the full potential of the Adreno GPUs found in Qualcomm’s Snapdragon X Series Windows PCs. They are regularly updated to contain the latest bugfixes and performance enhancements for games and other applications.

WinDbg is a tool used to debug and analyze issues with the Windows platform

Generative AI models such as large language models (LLMs) and large vision models (LVMs) contain multiple binaries after optimization due to their complexity and larger size. These binaries require a specific order of execution to utilize the processing power of a Neural Processing Unit (NPU). With Gen AI Inference Extensions (GENIE), you can streamline on-device Gen AI models in a single execution job. GENIE simplifies Gen AI inferencing at the edge by providing software library commands and resources integrated with Qualcomm®AI Engine direct. Qualcomm AI Hub supports GENIE, offering optimized models and sample apps on how to use Gen AI Inference Extensions (GENIE) to executeLLMs. A single SDK download includes the Qualcomm Neural Processing SDK, Qualcomm AI Engine direct SDK, and

The platform for on-device AI. Any model, any device, any runtime. Deploy within minutes.”

Qualcomm Technologies contributed the ExecuTorch repository for developers to access Qualcomm® Hexagon™ NPU directly. This streamlines the developer workflow and unlocks the benefits of local AI inference, from personalization to performance, privacy, and reduced reliance on cloud infrastructure

QACT™ provides a comprehensive tuning and calibration environment tailored towards the audio and voice algorithms from Qualcomm Technologies, Inc. and other 3rd parties.

Qualcomm® Intelligent Multimedia Product Software Development Kit (QIMP SDK) is a powerful and unified solution designed to empower application developers in the Internet of Things (IoT) space. Product SDKs combine several functional SDKs making it easy for you to build your applications.

Welcome to use the Qualcomm® Intelligent Robotics Product SDK (QIRP SDK) 2.0. This SDK is applicable to the Qualcomm® Linux releases. QIRP SDK is a collection of components that enable a developer to realize robotics features on Qualcomm Platforms.

IoT Solutions Microservices brings containerized IoT to edge development so you can quickly build on-device IoT applications that are easy to deploy and manage.

FoundriesFactory is a SaaS platform that uses GitOps to build and customize Yocto-based Linux images and containerized apps and deploy them onto embedded devices such as the QualcommDragonwing™ RB3 Gen 2. The comprehensive platform is built with Foundries.io’s open source Linux microPlatform (LmP) operating system, giving your embedded devices the benefit of a DevSecOps-based workflow.

AIMET is a library that provides advanced quantization and compression techniques for trained neural network models.

For developers, Qualcomm Technologies provides the Snapdragon® Telematics Application Framework (TelAF), a unified development environment including SIM management, peripheral device connections, and APIs for building cloud-connected apps and services. The framework is a set of build/debug/code-generation tools and support designed to simplify application development. Open source and scalable, the application framework allows developers to use telematics functionality across all platforms in the Snapdragon® Digital Chassis™, reducing costs and complexity when developing cloud-based applications and services. Based on the Yocto Linux distribution, this feature-rich set of tools provides developers with common APIs for integrating technology features such as 5G, location services, memory

Ubuntu on Qualcomm® IoT Platforms emerges from a strategic collaboration between Qualcomm Technologies and Canonical, delivering optimized Ubuntu images specifically for Qualcomm IoT platforms. This initiative is designed to cater to a wide array of device applications at the intelligent edge.

MRTK (Mixed Reality Toolkit) is a trusted framework for building immersive, cross‑platform XR applications using shared input, spatial UI, and interaction components. Now also available for Android XR, MRTK supports XR devices like Samsung Galaxy XR while enabling developers to continue using both MRTK3 and existing MRTK2 workflows.

The connected car is a broad frontier of wireless possibilities, and the Qualcomm® Telematics SDK (TelSDK) is your path for developing telematics applications.Connected-car developers can use the Telematics SDK for rapid development and deployment of applications that run on telematics control units (TCU). The SDK abstracts the modem and hardware in the modem and includes full documentation, sample applications, header files and stub code.What Can You Build into Your Applications with the Telematics SDK?With the Telematics SDK you can build out features for a wide range of connected-car scenarios including identity management, ridesharing, operator flexibility, fleet management, data analytics and firmware-over-the-air (FOTA) updates.The SDK provides a set of backward-compatible APIs to

The PlotTwist dataset enables you to study visual reasoning with Vision-Language Models (VLMs) to answer human-readable questions about images of mathematical graphs.

The AirLetters dataset has over 161,000 video-label pairs showing video clips of humans drawing letters and digits in the air to evaluate a model’s ability to classify articulated motions.

Qualcomm AI Research presents the Qualcomm Exercise Video Dataset (QEVD) which explores human-AI interaction in the challenging real-world domain of fitness coaching – a task which intrinsically requires monitoring live user activity and providing timely feedback. Our dataset includes corrective feedbacks to address potential user mistakes and steer them towards successful workout completion.Our dataset contains 474+ hours of videos and includes the following:short-clip videos (∼5 seconds in length) annotated with 1M+ question-answer pairsshort-clip videos (∼5 seconds in length) annotated with 650k+ live feedbacks including corrective feedbackslong-range videos (>3 minutes in length) annotated with 7.5k+ live feedbacks including corrective feedbacks

ClevrSkills includes 330,000 robot episodes/trajectories including videos (from multiple views), corresponding actions, and other annotations including text, bounding boxes, cameras poses etc., which were generated from over 33 tasks in ClevrSkills environment suite. It includes a carefully designed curriculum of tasks which can be used for training robotics models to perform tasks ranging from simple pick and place to more complicated manipulation, such as sorting, stacking etc

The Qualcomm Rasterized Images for Super-resolution Processing dataset was created to facilitate the development and research of super-resolution algorithms for gaming. The dataset consists of parallel captures of various scenes in different modalities and resolutions. It is designed to be diverse, with a variety of backgrounds and models, to better generalize to new video games.

The Something-Something dataset (version 2) is a collection of 220,847 labeled video clips ofhumans performing pre-defined, basic actions with everyday objects. It is designed to train machine learning models in fine-grained understanding of human hand gestures likeputting something into something, turning something upside downandcovering something with something.

Redefining the car of the futureThe era of the intelligently connected, autonomous driving car is upon us. The Snapdragon Ride Software Developer Kit (SDK) - a key pillar of the Snapdragon® Digital Chassis™ solution - assists automotive suppliers and automakers to deploy virtually all the autonomous driving features they want now — with the ability to scale in the future. Our full suite of automotive products was developed with more than 30 years of mobile leadership and 20 years of automotive experience. The Snapdragon Ride Platform and SDK are supporting automakers to build cars that are connected to networks, the cloud, other cars, pedestrians, traffic, and infrastructure systems.Simply put: Snapdragon Ride is redefining the car of the future.TheSnapdragon Ride Platformis at the

Your model recognizes certain simple, single-frame gestures like a thumbs-up. But for a truly responsive, accurate system, you want your model to recognize complex gestures too, even when the differences between them are subtle. Is the person pointing to something or wagging their index finger? Is the hand cleaning the display or rubber-banding an image with two fingers? Given enough examples, your model can learn the difference.The Jester gesture recognition dataset includes 148,092 labeled video clips ofhumans performing basic, pre-defined hand gesturesin front of a laptop camera or webcam. It is designed for training machine learning models to recognize human hand gestures likesliding two fingers down, swiping left or right and drumming fingers.

Keyword spotting (KWS) is widely used for detecting specific keywords in personal devices like mobile phones and home appliances. A keyword may consist of multiple words, where “Hey, Siri,” “OK, Google,” and “Hi, Bixby” are well-known examples.Qualcomm Technologies, Inc. has published a keyword dataset for the Snapdragon® mobile platform. The Hey Snapdragon keyword dataset contains 4,270 audio clips of four English keyword classes spoken by 50 people. The four keyword classes are:“Hey, Android”“Hey, Snapdragon”“Hi, Galaxy”“Hi, Lumina”Find out more about the keyword speech dataset and see how you can use it in your research projects.For any questions or technical support, please contact us at research.datasets@qti.qualcomm.com

Massive MIMO Spatial Channel Model Dataset is available for research purposes.

The Wireless Indoor Simulations dataset contains a large set of channels to enable a better understanding of the interplay between the propagation environment (e.g., materials, geometry) and corresponding channel effects (e.g., delay, receive power).The dataset is in two parts:Wi3RoomsWhere channels are simulated in various random configurations of 3-roomed indoor layouts in a 10⨉5⨉3m hull using the simulator PyLayers [1]WiIndoorWhere the 10⨉10⨉3m indoor configurations are based on the RPLAN dataset [2] and typically contains 4-8 rooms. The simulations were performed using Wireless InSite with ray tracer X3D [3]The dataset contains the geometry of the propagation environment (3d mesh, in.obj file format) and the wireless channels (multi-path components, in an hdf5 file format)

CausalCircuit is a dataset designed to guide research into causal representation learning – the problem of identifying the high-level causal variables in an image together with the causal structure between them.

Enables the study of visual understanding with Vision-Language Models (VLMs) to answer questions about events occurring in videos.

A comprehensive set of ready-to-use AI functions and tools provided as both standard OpenAI APIs, and an easy-to-use Python SDK. It also includes sample applications, agents, tools, and libraries for developing and deploying AI inference on premises or via cloud deployments.

Cloud AI SDKs enable developers to optimize trained deep learning models for high-performance inference on Qualcomm® Cloud AI 100 accelerators.The Cloud AI SDKs provide workflows to optimize the models for best performance, provides runtime for execution and supports integration with ONNX Runtime and Triton Inference Server for deployment.

The Qualcomm Broadcast SDK gives developers the power to bring Broadcast connectivity and content to their apps. This resource includes the APIs, tools, and documentation to develop and test Broadcast applications for Android-based mobile devices with broadcast service capabilities.

The Snapdragon® LLVM Toolchain for Arm® Technology is a C/C++ toolchain that enables a broad range of configurations, providing better performance, code-size and power use for your non-high-level operating system (HLOS) as well as Android and Windows on Snapdragon applications. This product is specific to QIK Installers.

The Qualcomm® Hexagon™ NPU device is an AI inference accelerator integrated with Qualcomm Technologies client CPUs. It enables energy-efficient execution of artificial neural network tasks. The Hexagon NPU Driver for Windows* includes preview support for Windows ML. See the release notes for supported platforms, operating system info, fixed and known issues, as well as a section on how to install/update the NPU driver.

An all-in-one SDK that supports heterogenous computing, system-level configurations and designed to direct AI workloads to all accelerator cores on our platforms. Provides developers flexibility, including inter-core collaboration support and other advanced features.

The WLAN QSPR Subsystem provides automated testing through QSPR WLAN chipset enabled DUT and associated test equipment. See the QDART Help file for the full list of supported test equipment.

Product Configuration Assistant Tool (PCAT) provides set of functionalities like build loading, dump collection, EFS, QCN backup/Restore & NV browser.

Provides lower-level, highly customizable unified APIs that speeds up AI models on all AI accelerator cores with individual libraries. Can be used directly to target a specific accelerator core or delegate workloads from popular runtimes including Qualcomm Neural Processing SDK, TensorFlow Lite and ONNX runtime.

The Qualcomm® Broadcast SDK gives developers the power to bring Broadcast connectivity and content to their apps. The package is a DASH player for optional use which allows developers to build applications for playback of DASH encoded content with the media player on Android-based devices.

This section contains various sample applications to demonstrate how Qualcomm Technologies' broadcast services can be used by the application. Developers can refer the sample source code and see how the Android MBMS and Qualcomm® SDK APIs are being used.

Allows Developers to easily get started developing Applications and Middleware for Qualcomm® Telematics products using APIs from the Telematics SDK

The Qualcomm® Hexagon™ SDK provides Windows and Linux based tools, software, and documentation to assist developers to build, run, debug, and profile applications on Qualcomm Technologies targets including Mobile, Embedded Linux (LE) and XR devices leveraging the Hexagon DSPs.

Windows Adreno GPU driver for Snapdragon X Platform

Qualcomm Fast Computer Vision (CV) SDK provides a low-level computer vision library that is accelerated on Snapdragon® platform and target for optimized computer vision applications

Qualcomm® Software Center manages Qualcomm® software and simplifies the development process. It offers features such as 1) Browse and discover Qualcomm® tools, SDKs, and chip software in the web portal or the Qualcomm Software Center GUI desktop application. 2) The desktop application is a guided UI to find, install, compile, and manage software resources for Qualcomm® devices (referred to as “Launcher” on Qualcomm® Software Center). You can also manage installation and updates of tools, SDKs, and chip software for your development environment. 3) qsc-cli command line interface (CLI) downloads and compiles Qualcomm software builds. qsc-cli uses docker environment at the backend and simplifies host setup, sync, and build to a few simple commands.

The Qualcomm® Profiler is a powerful performance profiling system designed to detect, measure, and enhance application scalability across various components of Qualcomm System-on-Chip (SoC) devices. It provides insights into the performance of CPUs, GPUs, DSPs, and other essential IP blocks, enabling developers to optimize their applications effectively.

Snapdragon® Profiler is profiling software that runs on the Windows, Mac, and Linux platforms and connects with devices powered by Snapdragon processors. Snapdragon Profiler allows developers to analyze CPU, GPU, DSP, memory, power, thermal, and network data, so you can find and fix performance bottlenecks.

Application Software Development Kit for Cloud AI products

Platform Software Development Kit (SDK) for Cloud AI products

The LPAI Add-on for Hexagon SDK contains libraries and tools to provide the following (It is built on top of the base Hexagon SDK) : ML runtime targeting deeply embedded use cases Optimized for a deeply embedded environment Low Power Usage Low Memory Usage : option to share memory between models Ability to select Operators between Low Power Island (LPI) memory and DDRAM Inference runtime – currently not for model training The Hexagon SDK is a software development kit supporting LINUX for programming the Hexagon Processor embedded in all Snapdragon® Platforms. The Hexagon SDK enables code generation, simulation, profiling, debugging and logging of DSP code for execution in a dynamic environment, where user DSP code can be converted into a dynamically loadable .so file for execution on a

Qualcomm USB host drivers providing logical representations of Qualcomm chipset-enabled mobile devices over USB connection