→ Back to Home
Edge Computing

SolidRun's New NXP i.MX 95 Portfolio Integrates NVIDIA AI for Industrial Edge

SolidRun, a prominent developer of high-performance System-on-Modules and industrial edge computing solutions, announced today its NXP i.MX 95 product family at electronica 2026. This new portfolio includes the i.MX 95 SoM, the HummingBoard i.MX 95 IIoT industrial carrier, extensive software support, and a camera integration ecosystem. A key highlight is the demonstrated end-to-end NVIDIA Holoscan implementation, signaling a strong push towards integrated AI capabilities at the edge. The platform is specifically engineered for demanding applications in Industry 4.0, real-time robotics, energy storage systems (ESS), and industrial networking infrastructure, promising deterministic multi-gigabit networking, real-time safety control, and scalable Physical Edge AI. This announcement is particularly significant for practitioners grappling with the complexities of deploying AI at the industrial edge. By bringing together NXP-based embedded platforms and NVIDIA-based AI systems under a single vendor, SolidRun aims to simplify development and reduce engineering risk. The integration of all Ethernet PHYs, including 10GbE, directly on the i.MX 95 SoM, along with pin-compatibility with previous i.MX8 lines, directly addresses hardware design friction, a common pain point for OEMs. This unified approach means that developers can leverage a single hardware partner for both edge processing and advanced AI inference, which is crucial for accelerating time-to-market in rapidly evolving industrial sectors. The broader trend in cloud and DevOps is a continuous push towards decentralization and bringing compute closer to the data source, often termed "edge computing." This is driven by the need for lower latency, increased data privacy, and reduced bandwidth consumption, especially with the proliferation of IoT devices and AI workloads. The integration of AI capabilities directly into edge hardware, as seen with SolidRun's offering, aligns perfectly with the rise of "Physical AI" where machines are designed to perceive, understand, and act safely in real environments. This move from software-defined to AI-defined platforms, leveraging specialized AI chips for local inference, is a defining characteristic of 2026's technological landscape. Companies are increasingly recognizing that processing data at the edge is not just an option but a necessity for real-time decision-making in critical applications. In practice, this means that industrial OEMs and system integrators should closely evaluate solutions like SolidRun's i.MX 95 family. The platform's focus on integrated connectivity, real-time control, and safety features, powered by Arm Cortex-A55 cores and an integrated NPU, positions it as a strong contender for applications requiring high-performance, low-latency AI at the edge. The support for Time-Sensitive Networking (TSN) across all Ethernet ports is vital for deterministic industrial communication. Practitioners should consider how such integrated solutions can reduce their bill of materials, simplify supply chains, and mitigate the risks associated with multi-vendor ecosystems. The ability to upgrade existing hardware without complete redesign due to pinout compatibility also offers a significant advantage in terms of cost and effort. This development underscores the importance of selecting edge platforms that offer a comprehensive, integrated hardware and software stack to effectively deploy and manage complex AI-driven industrial applications.
#industrial edge#edge ai#nxp#nvidia holoscan#system-on-module#industry 4.0
Read original source