Innodisk has launched the APEX-E300 and APEX-P300 Edge AI systems, which utilize Intel Core Ultra Series 3 processors to deliver up to 180 TOPS of performance. These platforms are designed to enable local, real-time execution of Vision Language Models and LLMs for industrial and retail video intelligence applications.
The shift toward local execution of Vision Language Models marks a critical transition for real-time video intelligence in retail and manufacturing. By moving inference from the cloud to the edge, operators eliminate the latency and connectivity costs that previously hindered immediate automated responses in smart infrastructure. This launch highlights a broader industry move toward heterogeneous computing, where specialized NPU and GPU silicon are consolidated to handle multi-model workloads on a single device. As streaming professionals integrate more AI-driven metadata and vision analysis into edge nodes, the ability to balance 180 TOPS of performance against a 25W power envelope will be a key benchmark. Watch for how quickly these localized VLM deployments scale in quick-service retail environments to replace cloud-based recommendation engines.
Innodisk's APEX-E300 launch arrives amid intensifying competition among edge AI hardware vendors targeting video intelligence workloads. The APEX-P300, Innodisk's PCIe-based companion product, targets a different deployment model where existing servers or gateways need GPU-class acceleration without a full system redesign. Innodisk has positioned both APEX platforms as part of a broader edge AI portfolio that includes industrial-grade storage and networking modules, though the company's primary differentiation lies in pairing Intel's NPU silicon with optimized thermal designs for fanless enclosures. Intel's Core Ultra Series 3 processors, which integrate CPU, GPU, and NPU on a single die, represent the company's push to consolidate heterogeneous compute for edge inference without discrete accelerator cards.
The business case for edge-based video AI inference is strengthening as enterprises seek to reduce cloud egress costs and meet data-residency requirements. Bitmovin's 2026/2027 Video Developer Report found that 98 percent of video professionals now use AI or ML in their workflows, with visual quality and optimization cited by 30 percent of respondents, underscoring demand for inference capacity closer to the camera. For retail and industrial operators evaluating Innodisk's APEX platforms, the calculus centers on whether local VLM execution at 180 TOPS can replace cloud API calls for tasks like shelf monitoring, defect detection, and real-time content moderation without sacrificing model accuracy.
On the competitive landscape, Mux has moved aggressively into AI-assisted video workflows that could complement or compete with edge inference deployments. Mux launched Robots in early 2026, a first-party API that runs video analysis jobs natively inside its platform using the @mux/ai engine, eliminating the need for developers to manage external AI provider keys. Meanwhile, a 2026 industry analysis of managed video APIs found that Mux now ships Claude auto-chaptering, semantic search, and GenAI clips, while Cloudflare Stream offers per-title AI encoding and Hive moderation, illustrating how cloud-based video platforms are absorbing AI functions that edge devices like Innodisk's APEX-E300 aim to handle locally. Buyers evaluating edge versus cloud inference for video intelligence must weigh Innodisk's 25W power envelope and on-premises data control against the operational simplicity of managed APIs that bundle encoding, delivery, and AI analysis in a single platform.
Innodisk has launched its APEX-E300 and APEX-P300 Edge AI platforms, powered by Intel Core Ultra Series 3 processors. These systems deliver up to 180 TOPS of performance, enabling industrial and retail operators to run complex Vision Language Models and LLMs locally, reducing cloud dependency, latency, and operational costs for real-time video analysis.
The APEX-E300 and APEX-P300 systems provide up to 180 total platform TOPS by combining CPU, GPU, and NPU resources.
Local execution allows operators to eliminate cloud egress costs and latency, enabling immediate automated responses for tasks like shelf monitoring and defect detection while maintaining data residency.
The integrated compute resources in the APEX systems allow for power-efficient AI execution at levels as low as 25W.
The APEX-P300 model includes PCIe expansion slots, which allow users to scale compute resources by adding discrete GPU acceleration.
Add StreamingMeme as a preferred source on Google to see more of our streaming news at the top of your Search results.
Add as preferred source
