# Snapdragon X2 Linux AI/ML software overview documentation Snapdragon X2 adds two new AI accelerators and a layered AI/ML software stack that lets you run general AI, audio always-on AI, and on-device generative AI workloads across the CPU, GPU, and NPU. ## Overview New hardware AI accelerators on Snapdragon X2 Explore the new Snapdragon X2 hardware accelerators for AI workloads. https://docs.qualcomm.com/doc/80-A0399-65/topic/introduction.html#ai-accelerators Hexagon DSP evolution: Hexagon Tensor Processor Learn how the Hexagon DSP evolved into the Hexagon Tensor Processor for AI acceleration. https://docs.qualcomm.com/doc/80-A0399-65/topic/introduction.html#dsp-evolution Hexagon Tensor Processor software improvements Explore the software improvements for AI workloads on the Hexagon Tensor Processor. https://docs.qualcomm.com/doc/80-A0399-65/topic/introduction.html#hexagon-tensor-processor-software-improvements Low power AI (eNPU) on Snapdragon X2 Learn how the eNPU supports low-power AI workloads on Snapdragon X2. https://docs.qualcomm.com/doc/80-A0399-65/topic/introduction.html#low-power-ai-with-the-enpu Hexagon Tensor Processor and LPASS AI accelerator Compare the benefits of the Hexagon Tensor Processor and LPASS AI accelerator for AI workloads. https://docs.qualcomm.com/doc/80-A0399-65/topic/introduction.html#hexagon-tensor-processor-and-lpass-ai-accelerator-benefits AI inferencing: Compute platforms Explore the compute platforms available for AI inferencing on Snapdragon X2. https://docs.qualcomm.com/doc/80-A0399-65/topic/introduction.html#ai-inferencing-compute-platforms AI inferencing: GenieX > > > Learn about GenieX for running generative AI inference on-device. https://docs.qualcomm.com/doc/80-A0399-65/topic/introduction.html#geniex-on-device-generative-ai-runtime ## AI/ML software solutions AI software stack architecture Explore how the Qualcomm AI software stack routes AI/ML workloads across the CPU, GPU, NPU, and LPAI. https://docs.qualcomm.com/doc/80-A0399-65/topic/software-solutions.html#ai-sw-stack-arch Qualcomm AI Runtime SDK 2.X Accelerate AI/ML workloads and develop custom Hexagon Tensor Processor operators using Qualcomm AI Runtime SDK 2.X. https://docs.qualcomm.com/doc/80-A0399-65/topic/software-solutions.html#qualcomm-ai-runtime Qualcomm AI Model Efficiency Toolkit Pro Fine-tune PyTorch and TensorFlow models for Qualcomm AI Software SDKs using AIMET Pro compression, quantization, and visualization. https://docs.qualcomm.com/doc/80-A0399-65/topic/software-solutions.html#aimet Last Published: Aug 27, 2026 [Next Topic AI/ML software introduction](https://docs.qualcomm.com/bundle/publicresource/80-A0399-65/topics/introduction.md)