Manta Network Partners with Aethir for Enhanced GPU Access

IoTeX, the decentralized network bridging the physical and digital worlds, announced a significant expansion into artificial intelligence. This move introduces a unified ecosystem for Physical AI, a new category of intelligence powered by real-world data from machines, devices, and decentralized infrastructure. The transition of AI systems from digital environments to real-world applications emphasizes the importance of access to live, trustworthy data from the physical surroundings. IoTeX aims to enable a new class of AI grounded in reality by creating a shared infrastructure where machines and AI agents can coordinate via verified, real-time data, transforming decentralized networks into a foundation for responsive, real-world AI systems. Since its establishment in 2017, IoTeX has been at the forefront of technologies bridging physical devices with blockchain networks. The company's latest AI expansion aligns with its vision of linking the demand for real-time, verifiable data from AI developers with the supply of connected devices and DePIN participants on IoTeX. The AI infrastructure is powered by IoTeX Layer 1 Blockchain for secure coordination, the ioID Protocol for decentralized machine authentication, Quicksilver AI framework for transforming raw machine signals, and the IOTX Token for governance and incentives. Central to IoTeX's AI architecture are Realms, evolving knowledge bases of real-time, real-world data that synthesize insights from machines, people, and sensors to generate actionable intelligence in various sectors like mobility, health, energy, and robotics. Realms serve the dual purpose of providing real-time situational awareness for AI agents and offering dynamic environments for training the next generation of Physical AI models. As data flows into a Realm from connected devices and DePIN nodes, its intelligence becomes more precise, adaptive, and valuable, evolving into living knowledge bases that scale with participation. IoTeX's vision is to build dynamic economies of intelligence where machines, people, and AI collaborate to generate continuously evolving knowledge, forming a global constellation of verified environments where physical intelligence becomes a scalable and monetizable resource for real-world AI applications.

Theta EdgeCloud has unveiled an innovative feature: a large language model (LLM) inference service that provides distributed verifiability. This groundbreaking service allows AI agents and chatbots to perform LLM inference that is both trustworthy and independently verifiable, a first in the industry. By leveraging blockchain-backed public randomness beacons, Theta EdgeCloud combines advanced AI technology with the decentralized and trustless nature of blockchain, making it the only platform to offer such capabilities among both crypto-native and traditional cloud services. This development is particularly significant for sectors like enterprise and academia, where the integrity of computational outputs is crucial. The need for trust in AI outputs has become increasingly important as modern AI agents depend heavily on LLMs for generating responses and executing complex tasks. Traditionally, many frameworks have relied on centralized LLM APIs or hardware-enforced security, which can compromise the reliability of results. The introduction of DeepSeek-V3/R1, an open-source alternative to proprietary LLMs, paves the way for full-stack verification. Theta EdgeCloud has taken this a step further by implementing a Distributed Verifiable Inference system, ensuring that LLM outputs are reproducible and tamper-proof, even from the service provider itself. The newly released industry-grade verifiable LLM inference engine is integrated into Theta EdgeCloud, achieving public availability through deterministic token probability and verifiable sampling. By utilizing open-source models, users can independently verify the next-token probability distribution. The incorporation of a publicly verifiable random seed ensures that the sampling process is reproducible and transparent. This two-part design guarantees that no party can tamper with the model output, thereby enhancing the integrity of AI outputs. As AI continues to permeate various aspects of life and commerce, the ability to verify and trust these outputs will be essential for safety and security, marking a significant advancement in the field of AI.

In May, CUDOS made significant strides in the cloud computing landscape, particularly in the realm of decentralized infrastructure. The highlight of the month was the participation in Consensus 2025, where the emphasis was placed on the necessity of open, permissionless, and privacy-preserving compute infrastructure. Developers and teams from various sectors, including DeFi and AI, expressed their frustrations with centralized cloud solutions, indicating a growing demand for alternatives like CUDOS Intercloud, which offers digital wallet-authenticated GPU access without the need for accounts or KYC. Additionally, CUDOS introduced new one-click deployment templates, enhancing the ease of AI experimentation. The latest additions, including Llama 3.1 and Llama 3.3, complement existing templates like Dify and JupyterLab, allowing users to launch full-stack deployments with minimal effort. Furthermore, the integration of Squid Router enables payments across over 70 chains, simplifying the process for users to pay for compute resources in their native tokens without the need for bridging. The metrics for May reflect CUDOS's rapid growth, with a record $239,300 in on-demand revenue and over 330,000 GPU compute hours consumed. The ecosystem has expanded to include more than 18,000 users and a total revenue exceeding $2 million. Moreover, the first containerized compute hardware for the Artificial Superintelligence Alliance was deployed, marking a pivotal step towards distributed computing solutions. CUDOS continues to innovate and scale, reinforcing its commitment to building a robust and accessible cloud infrastructure for the future.

In the realm of personal health technology, the BrushO AI-Powered Mining Toothbrush emerges as a groundbreaking innovation that redefines oral care. This intelligent toothbrush is not just a tool but a dedicated coach that transforms daily brushing into a precision health session. Equipped with Fully Smart Brushing (FSB) technology, the BrushO meticulously tracks every movement, pressure point, and coverage area in real-time. This advanced oral health tracking, powered by AI and IoT principles, provides users with instant, actionable feedback through a user-friendly app, effectively acting as a personal dental hygienist that guides users towards optimal oral health. The benefits of the BrushO extend beyond just improving brushing techniques. Over time, the AI analyzes users' habits and identifies trends, offering a holistic view of their oral health ecosystem. This data-driven approach allows for proactive prevention of dental issues by pinpointing potential problem areas before they escalate. Additionally, users receive personalized recommendations tailored to their unique habits, optimizing their brushing technique for maximum effectiveness. This continuous, AI-guided care translates into significantly improved gum health and reduced plaque buildup, ensuring long-term oral health benefits. What sets BrushO apart is its innovative 'brush-to-earn' model, which rewards users for contributing anonymized data to the BrushO network, part of the Decentralized Physical Infrastructure Network (DePIN). Built on Solana Blockchain technology, BrushO emphasizes data ownership and control, allowing users to maintain sovereignty over their health information. This model not only empowers users with tangible rewards in the form of $ORAL tokens but also fosters a collaborative health management system without compromising personal privacy. BrushO is not just a toothbrush; it’s a smart companion dedicated to enhancing well-being and transforming oral health practices for a healthier future.

In the rapidly evolving landscape of AI-powered cryptocurrencies, Bittensor (TAO), Mind of Pepe (MIND), and FloppyPepe (FPPE) are emerging as frontrunners poised to capture the attention of Silicon Valley venture capitalists. These projects are not only innovating within the blockchain space but are also laying the groundwork for an AI-driven Web3 economy. Bittensor allows users to earn by enhancing AI models on a decentralized network, while Mind of Pepe integrates AI into meme culture. FloppyPepe stands out by combining intelligent tools with crypto content creation, sentiment analysis, and community engagement, making it a strong candidate for exponential growth in the coming years. FloppyPepe (FPPE) is particularly noteworthy as it develops a comprehensive AI toolkit tailored for the next generation of crypto communities. Unlike its competitors, FloppyPepe boasts two AI agents and a chatbot, enhancing its influence in the digital space. Its innovative features include the Meme-o-Matic Machine, which generates viral memes, FloppyX, an AI video bot that revolutionizes crypto content creation, and FloppyAI, a personal chatbot that tracks trends and engages users. This multifaceted approach positions FloppyPepe as a significant player in the AI crypto market, with the potential for substantial returns on investment. As the market for AI cryptocurrencies heats up, FloppyPepe's presale offers a unique opportunity for early investors. With a low entry price and a promotional bonus for presale participants, the project aims to attract significant interest from both individual and institutional investors. Meanwhile, Bittensor is gaining traction with analysts who see it at a pivotal development stage, akin to Bitcoin's early days. Mind of Pepe has also made a strong debut, launching its AI agent to positive reception. As these projects continue to evolve, they represent a promising intersection of AI and blockchain technology, appealing to those looking for the next big opportunity in the crypto space.

Theta Labs has announced the beta release of its innovative hybrid edge cloud architecture for the Theta EdgeCloud network, aimed at providing enterprise-grade AI computing at significantly reduced costs. This platform integrates traditional cloud-based GPUs with a decentralized network of over 30,000 community-operated edge nodes, allowing for cost-effective access to high-performance computing resources. The introduction of a decentralized GPU marketplace ensures competitive and transparent pricing, addressing the challenges posed by the rising costs and limited availability of specialized hardware for AI and machine learning tasks. The Theta EdgeCloud serves as a dynamic marketplace that connects GPU computing power supply and demand. It enables individuals with idle GPUs to contribute their resources and earn rewards, while offering developers and AI teams a scalable platform for running containerized workloads. Users can select the most suitable infrastructure for their computing tasks, whether it be powerful cloud GPUs for large AI model training or community-operated gaming machines for parallelizable workloads. This market-driven approach promotes fair pricing, allowing node operators to set their rental rates while users can choose nodes based on their performance needs and budget constraints. Currently, Theta EdgeCloud supports a range of academic and enterprise clients, including prestigious universities and major sports teams. The platform's beta release includes features such as persistent storage for AI model training and improved job prioritization, responding to customer requests. With over 80 PetaFLOPS of distributed GPU compute power, the hybrid architecture supports various applications, including AI model training, video encoding, and financial simulations. By leveraging unused GPU power from community members, Theta Labs aims to democratize access to high-performance computing, enabling organizations to conduct more experiments and advance their AI research without financial limitations.