The SRM949 module is based on the Qualcomm QCM6650 platform and is equipped with an 8-core 4nm advanced process Kryo CPU, which achieves an excellent balance between efficiency and performance. The module features a dual-core AI processor built-in.Dual Hexagon Tensor Processor）， Equipped with 6 TOPS AI computing power, it can efficiently run end-side AI models such as speech recognition, driver behavior analysis, and occupant state monitoring. Through multimodal integration and the autonomous scheduling capabilities of the Agent framework, it can effectively create autonomous cockpit intelligences.

In graphics and video frequency processing, SRM949 built-in Adreno GPU, support OpenGL ES3.2, Vulkan 1.x, OpenCL 2.0 And DirectX FL12 and other mainstream graphics interface standards, can smoothly drive the control of the large screen HD interface rendering and 3D dynamic interaction. At the same time, the module integrates Adreno VPU video processing units and supports 4K HMX (hardware-grade HEVC encoding acceleration), up to 4K60 video encoding and 4K120 video decoding, and is fully compatible with mainstream video codec formats such as H.264, HEVC, VP9 and other major video codecs.

In terms of communication capabilities, the SRM949 module supports the 5G NR Sub-6GHz band, supports downside 4x4 MIMO and upside 2x2 MIMO antenna configurations, and is compatible with NSA and SA networking modes, providing high-speed, low-latency wide-area connectivity for vehicles.In addition, the module integrates 2 × 2 mimo Wi-Fi 6E and BT5.4, taking into account the high-speed wireless local area network access in the car and the peripheral Bluetooth connection, fully meeting the diversified communication needs in the car scene.

In terms of positioning capabilities, SRM949 integrates L1 + L5 dual-band GPS, which can significantly improve the positioning accuracy and stability in complex environments such as urban canyons and viaducts. The dual-frequency positioning solution meets the stringent requirements of AECS (Vehicle Emergency Calling System) and provides highly reliable position information support for vehicle navigation and emergency rescue services.

The SRM949 module has a rich functional interface configuration, including LCM, touchscreen, 4 path CSI camera interface, UART, USB, I2C, SPI, PCIe and other interface types, allowing customers to flexibly extend various peripherals. Among them, 4 CSI interfaces can simultaneously access multiple camera inputs such as AVM circling, DMS driver monitoring, OMS occupant monitoring, etc.Combining Module 6’s TOPS AI computing power, SRM949 enables efficient visual perception processing. Leveraging its comprehensive interface capabilities, it can simultaneously support various key cockpit functions such as instrument display, central console entertainment, AVM panoramic surround view, DMS driver monitoring, OMS passenger monitoring, and T-BOX vehicle network communication. By effectively covering various aspects of cockpit agent multimodal perception and efficient execution, with the collaborative support of MeiG Smart AI Engine and AI Toolchain, we can help customers implement cockpit agent solutions efficiently and at a low cost, thereby gaining a competitive edge in the market.

The modules are developed in strict accordance with the AEC-Q104 automotive specifications, have an ultra-wide operating temperature range, and can withstand harsh automotive application environments and maintain stable and efficient computing and communication performance.

With AI modules as the core, MeiG Smart has built a complete product matrix from 4G / 5G automotive-grade communication modules to integrated communication + computing intelligent modules, and the 5G intelligent modules have shipped more than one million units in the field of new energy vehicle in-cabin. The launch of SRM949 further improves the product line layout of MeiG Smart in the field of cockpit computing and cockpit smart body, providing vehicle manufacturers and Tier1 customers with diversified choices covering different computing power levels and different cost ranges, and promoting the acceleration of cockpit smart body.