Edge artificial intelligence
PulseAugur coverage of Edge artificial intelligence — every cluster mentioning Edge artificial intelligence across labs, papers, and developer communities, ranked by signal.
9 day(s) with sentiment data
-
Goke unveils GK6950 edge AI chips for embodied robots
Goke is developing the GK6950 series of edge AI chips, slated for release next year. These chips are designed for embodied robots and will offer 10 TOPS of INT8 performance. They will also support mixed precision for in…
-
Edge AI enhances IoT by enabling faster, localized decisions
The Internet of Things (IoT) is increasingly integrating into various aspects of the physical world, connecting devices in homes, factories, healthcare, and infrastructure. Edge AI is enabling more localized decision-ma…
-
Edge AI's real bottleneck: Model distribution, not model size
The primary challenge for on-device AI is not model optimization but rather the complex process of distributing, versioning, and updating these models across a fleet of devices. Unlike cloud-based models treated as serv…
-
Edge AI framework enhances 6G resource allocation with multi-view semantic approach
Researchers have developed a new framework for semantic resource allocation in multi-user Edge AI, aiming to address resource scarcity in 6G networks. The approach extends the Opportunistic Information Bottleneck (OIB) …
-
Edge AI and Physical AI to Power Cognitive Cities
This white paper explores the integration of edge AI and physical AI to transform smart cities into cognitive cities. It details how urban infrastructure can move from passive monitoring to active decision-making and ac…
-
TinyML enables autonomous fault detection in optical wireless communication
Researchers have developed an autonomous optical wireless communication receiver node that integrates a TinyML classification engine for real-time fault detection. This system, built around an ESP32-S3 microcontroller, …
-
Edge AI Security Explored in New Technical Video Series
Shawn Hymel has released the third part of his "Let's Talk Technical: Edge AI" video series. This installment focuses on the security aspects of edge artificial intelligence, drawing on insights from various engineers a…
-
Small edge AI models outperform large ones on specific tasks
Small language models, specifically those around 3 billion parameters, can outperform larger 70 billion parameter models when deployed at the edge, provided the task is narrowly scoped. These smaller models offer advant…
-
Edge AI and Mistral Offer Alternatives to Cloud AI Subscriptions
The article discusses alternatives to cloud-dependent AI services, focusing on edge AI, Mini RAG, and models from Mistral AI and Ensu. It suggests these options offer greater data control and reduce reliance on subscrip…
-
Connectivity gaps hinder Edge AI adoption for infrastructure
Edge AI's potential is being hampered by connectivity limitations, particularly for infrastructure systems that require reliable, persistent communication rather than just high speeds. Traditional wireless networks, opt…
-
MLOps for Edge AI: Preparing Models for Deployment
This cluster of posts discusses the practice of MLOps, specifically focusing on its application to edge AI. The content highlights the process of preparing and deploying artificial intelligence models to run on edge dev…
-
Tech M&A trends point to future of AI, agents, and data platforms
Technology mergers and acquisitions can offer insights into future trends, particularly in areas such as AI routers, agents, edge AI, data platforms, and coding tools. Following the financial investments in these sector…
-
Banana Pi launches new Edge AI modules with Rockchip processors
Banana Pi has released two new computing modules, the BPI-CM5 Pro and the BPI-AIM7, both designed for Edge AI applications. The BPI-CM5 Pro utilizes the Rockchip RK3576 processor, while the BPI-AIM7 is powered by the Ro…
-
Edge AI inference on smartphones is less sustainable than cloud, study finds
A new study published on arXiv investigates the environmental impact of running large language models (LLMs) on mobile devices, challenging the assumption that edge AI is inherently more sustainable than cloud-based inf…
-
ADATA expands B2B memory and storage for AI servers and edge AI
ADATA is expanding its business-to-business memory and storage offerings to support AI servers and edge artificial intelligence applications. This move aims to cater to the growing demand for specialized hardware in the…
-
IoT, AI, and Blockchain Convergence Promises Data Integrity and Edge Computing
The convergence of three key technologies—Internet of Things (IoT), Artificial Intelligence (AI), and Blockchain—is poised to revolutionize the digital landscape. IoT devices will gather real-world data, AI will analyze…
-
Edge AI for Military Needs Offline Functionality and Resilience
The article discusses the necessity for AI systems deployed in military environments to function effectively without constant network connectivity. This requirement highlights the critical importance of low latency, eff…
-
Edge AI research focuses on optimizing inference throughput and efficiency
Two new research papers from arXiv explore methods for optimizing AI inference on edge devices. The first paper introduces a channel-adaptive AI algorithm designed to maximize inference throughput by adjusting computati…
-
On-device AI gains traction with new SDKs and CCTV integration
The NobodyWho library is enabling developers to integrate large language models (LLMs) directly into applications for on-device AI, offering benefits like offline functionality, enhanced privacy, and reduced latency. Th…
-
Edge-AI framework enhances task allocation in smart manufacturing
A new research paper proposes an Edge-AI-driven framework for decentralized task allocation in circular smart manufacturing. This approach utilizes lightweight decision intelligence deployed at the machine level, incorp…