Gate News bot message, according to the official announcement from Gate on September 5, 2025: Gate will launch the 311th Launchpool project Gata (GATA) token airdrop event. This airdrop event will take place from 11:00 on September 5, 2025, to 11:00 on September 15, 2025, with a total airdrop amount of 2,000,000 GATA. Users can participate in the airdrop by staking USDT, GT, or GATA, and the system will issue token rewards every hour based on the proportion of the user's staked assets. The Gata project focuses on building decentralized AI inference and training technology, aiming to achieve efficient collaboration of large-scale AI models on a globally distributed GPU network. During the event, new users who register after September 3, 2025, and participate in the USDT staking pool while maintaining a continuous stake for more than 48 hours with an average effective staking amount of no less than 100 USDT will receive a reward of 100 USDT position experience vouchers, limited to 3000 slots.

Gate News bot message, according to the official announcement from Gate on September 3, 2025: Gate will launch the GATA token with the GATA/USDT trading pair on September 4, 2025, at 20:00 ( UTC+8 019283746656574839201. At the same time, the platform will initiate the 290th HODLer Airdrop event, where a total of 250,000 GATA tokens will be airdropped to users holding no less than 1 GT, proportionate to their GT holdings. The airdrop event will last until September 4, 2025, at 18:00 ) UTC+8 019283746656574839201. Gate is a project dedicated to building decentralized AI inference and training technologies, aimed at achieving efficient collaboration of large-scale AI models on a globally distributed GPU network. The GATA token is issued based on the BEP-20 standard, with a total supply of 1 billion, of which 45.7% is used for ecosystem incentives, 21% is allocated to the team and advisors, 13.5% is used for exchanges and marketing, and the remainder is allocated to investors and the foundation.

According to the Gate News bot message, as per the official announcement from Gate on September 1, 2025: Gate will launch NodeOps( spot trading on September 1, 2025, at 19:00 ) UTC+8. NODE was previously launched on Gate Alpha on August 7, 2025. NodeOps Network is a platform that supports large-scale verifiable computing, providing GPU/CPU orchestration and RPC services. The project has secured $5 million in funding from institutions such as L1D, BFF, and Finality Capital, currently boasting over 700,000 users and managing assets exceeding $150 million. The platform connects more than 89,000 machines and over 24,000 suppliers, generating cumulative revenue exceeding $4.1 million, ranking among the top five in the revenue growth rate of DePIN projects.

ChainCatcher news, EdgeX Labs officially launches EdgeX Cloud. EdgeX Cloud is a GPU cloud computing platform specifically designed for AI Agents and large language models. EdgeX Cloud focuses on building cost advantages, low operational costs, and multi-model support. Core capabilities include: Pre-configured LLM and Agent runtime environment; Automatic Elastic Scaling + Cloud-Edge Hybrid Computing; Pay-as-you-go billing to avoid wasting computing power; Local/private deployment to meet sensitive scenarios; This release means that AI infrastructure is becoming faster, lighter, and more inclusive.

Gate News Bot message, August 20, according to CoinMarketCap data, as of the time of writing, PHA (Phala Network) is currently priced at $0.12, with a rise of 11.31% in the last 24 hours, reaching a high of $0.12 and a low of $0.10. The current market capitalization is approximately $95.1 million, an increase of $9.67 million compared to yesterday. Phala Network is a new type of privacy AI cloud platform designed to build AI that people can trust. The platform offers a hardware-secure computing environment that can deploy enterprise-level privacy-protected AI models in just a few minutes. Phala Network supports pre-configured secure AI models and provides hardware-secure GPU servers for instant, private AI computation. The platform also offers a comprehensive privacy computing platform suitable for AI.

Golden Finance reports that Nasdaq-listed company VCI Global has announced the establishment of a sovereign-level infrastructure joint venture backed by 18,000 BTC. VCI Global will hold a 70% stake in the company. The development roadmap for the joint venture includes: exploring BTC-backed tokenization of RWA, providing sovereign-level encryption storage, building "AI GPU as a Service" infrastructure, and offering plug-and-play encryption reserve infrastructure for institutions.

Injective has partnered with Squaretower to launch the first on-chain rental trading market for NVIDIA H100 GPUs, with real-time updates on rental prices. This innovation transforms GPU rental prices into tradable on-chain assets, expanding the Decentralized Finance space.

Golden Finance reports that NVIDIA claims the RUBIN GPU chip is "progressing smoothly."

On August 13, Jin10 reported that Jintian Co., Ltd. stated on the interactive platform that the company has a good customer base and technical reserves in the chip computing power field, and is also one of the earliest companies in the world to achieve bulk supply of copper-based materials to leading enterprises in the aforementioned field. Among them, the company's high-precision shaped oxygen-free copper busbar products have successfully achieved mass production in the new AI cooling structure of 3DVC, and strategic cooperation has been established with many global first-tier cooling module companies, being applied in several top GPU cooling solutions; the company's self-developed copper heat pipes, liquid-cooled copper pipes, and other products have been successfully integrated into the computing power server products of many leading enterprises.

Gate News Bot message, on August 13, according to CoinMarketCap, as of the time of writing, IO (io.net) is currently priced at $0.68, having pumped 12.70% in the last 24 hours, with a peak of $0.69 and a low of $0.54. The current market capitalization is approximately $128 million, an increase of $14.4 million compared to yesterday. io.net is a decentralized GPU ecosystem focused on AI workloads. It provides AI teams with a smart stack that allows for instant access to enterprise-level computing resources at 70% lower cost than centralized alternatives. io.net has over 30,000 verified GPUs across more than 130 countries, deployable within minutes without approvals, contracts, or waiting lists. The platform supports containerized AI workloads, Ray clusters, and bare-metal deployments, offering users flexible options.

Aethir announced the optimization of the core parameter K value, dropping the cloud host stake requirements, applicable to new GPU providers, while existing Nodes are not affected. This will release more GPU supply, enhance AI and gaming customer satisfaction, and drive network utilization above 95%. The new mechanism will take effect on August 14, aiming to strengthen the Decentralization of GPU infrastructure.

ChainCatcher news, EdgeX Labs announced the launch of a brand new official website, fully presenting the hybrid AI + DePIN technology architecture that integrates cloud and edge, along with the global Computing Power network layout. This upgrade focuses on PaaS and SaaS level AI services, deeply integrating multimodal large language model capabilities, hybrid AI Computing Power scheduling protocols, and the global Node network, providing developers and enterprises with full-link support from model deployment, personalized fine-tuning to real-time inference. At the same time, the newly integrated TTS voice service and multimodal interaction system enable text, images, and voice to achieve low latency and efficient collaboration within the same intelligent network. EdgeX Labs Smart Hardware Matrix: PoS GPU-Server for enterprise-level high concurrency inference, PoW AI Dual

Jinse Finance reports that Nasdaq-listed Bitcoin mining company Hut 8 has released its financial performance report for the second quarter, disclosing that the company's revenue reached $41.3 million, net income was $137.5 million, and strategic Bitcoin reserves expanded to 10,667 coins, with a holdings market capitalization of $1.1 billion as of June 30, 2025. In the second quarter, Bitcoin mining, GPU as a service, and data center cloud operations generated $34.3 million in revenue, and in collaboration with American Bitcoin, they began offering ASIC hosting services with a capacity exceeding 130 MW.

Gate News Bot news, August 6, according to CoinMarketCap data, as of the time of writing, RDO (Reddio) is currently priced at $0.00096, rising 33.95% in the last 24 hours, reaching a high of $0.00143 and a low of $0.00045, with a 24-hour volume of $9.02 million. The current market capitalization is approximately $1.82 million, an increase of $461,800 compared to yesterday. Reddio is a high-performance parallel Ethereum-compatible layer 2 network that utilizes zero-knowledge technology to achieve unparalleled computational scale with Ethereum-level security. The project employs multi-threaded parallel execution technology, supporting all transactions within the same block to be executed in parallel across multiple threads, fully utilizing multi-core servers to achieve high throughput and low transaction fees. Reddio has also undergone GPU acceleration optimization, using

Gate News Bot news, on August 1st, according to CoinMarketCap data, as of the time of writing, SPON (Spheron Network) is currently priced at $0.05, having risen 95.11% in the last 24 hours, with a high of $0.06 and a low of $0.02. Current market capitalization data is temporarily unavailable. Spheron Network is a decentralized GPU network designed to provide revolutionary GPU computing power for AI and ML workloads. The network aggregates idle resources from gaming equipment, data centers, and mining farms, connecting them through a permissionless decentralized network. Spheron Network is committed to providing developers with a scalable, decentralized cloud computing alternative, optimized specifically for AI, ML, and next-generation workloads. Recent important SPO

ChainCatcher news, the Bitcoin Layer 2 network GOAT Network announces the official launch of BitVM2 Testnet Beta, and has implemented the first real-time proving system running in a production environment, becoming the first Bitcoin zkRollup project to support instant withdrawal initiation and full-chain visual proof. According to the introduction, the system is based on a pipelined parallel architecture and a distributed GPU prover network, combined with ZKM's zkVM "Ziren", to achieve three-phase parallel processing of block proof (2.6s), aggregate proof (2.7s), and Groth16 proof (approximately 10.4s), supporting real-time proof capability of 3.4 seconds for block generation, allowing users to initiate withdrawals without waiting. GOAT Network states that all

The RWAiFi project GAIB recently completed a $10 million financing round, led by Amber Group. The funds will be used for GPU asset tokenization to promote the development of the AI economy. It is expected that by 2025, the platform will attract $28 million in deposits and will have partnerships with Morpho and Pendle.

According to Gate News bot, on July 25, 2025, Gate will launch the spot trading of Spheron Network (SPON) token for the first time and will start the 275th HODLer Airdrop activity. The airdrop event will end on July 29, 2025, at 15:00 ( UTC+8, with a total airdrop amount of 266,667 SPO. The SPO/USDT trading pair will start trading on July 29, 2025, at 17:00 ) UTC+8. Spheron Network is a decentralized computing network that integrates data center-grade and retail-grade GPU/CPU resources, with over 50,000 active nodes. The SPON Token is issued on the BASE network, with a total supply of 1 billion tokens, and the contract address is 0x080d43c2164afdbc3712422ce78ab902ccab5ca1.

According to a survey by TrendForce, the overall server market is stabilizing, with ODM manufacturers focusing on the development of AI servers. The Blackwell series GPUs are expected to account for over 80% of NVIDIA's high-end dump. At the same time, the application rate of liquid cooling solutions in high-end AI chips is continuously increasing.

Jin10 data reported on July 10, this morning, Microsoft open-sourced the latest version of the Phi-4 family, Phi-4-mini-flash-reasoning, on their official website. The mini-flash version continues the Phi-4 family's characteristic of small parameters and strong performance, specifically designed for scenarios limited by Computing Power, memory, and latency. It can run on a single GPU, making it suitable for edge devices such as laptops and tablets. Compared to the previous version, mini-flash utilizes Microsoft's self-developed innovative architecture SambaY, achieving a 10-fold increase in reasoning efficiency, with average latency reduced by 2 to 3 times, resulting in a significant overall improvement in reasoning performance.

ChainCatcher message, ZKM has released the next generation zkVM (zk-SNARKs Virtual Machine) Ziren (formerly zkMIPS), which opens new possibilities for the deployability and scalability of zero-knowledge (ZK) systems in real-world application scenarios through GPU acceleration, a distributed proof architecture, and a deeply optimized constraint system. Ziren achieves a core proof speedup of 30 times, an aggregation proof speedup of 15 times, and an improvement in state preparation time from 5 minutes to 30 seconds, and has been put into use in the GOAT Network Bitcoin native zkRollup project. The naming inspiration for Ziren comes from the mythological variant "Siren", representing the transition from illusion to reality - technology is no longer just attractive, but a genuine promise that can be implemented. Ziren aims to become the true tool in the hands of mainstream developers.

Gate News bot reports that according to the Gate announcement on June 30, 2025, Gate will launch the decentralized creative AI reasoning network project Sogni AI (SOGNI)'s token. It is reported that the SOGNI/USDT trading pair will begin trading on July 2, 2025, at 20:00 ( UTC+8. The platform also announced the launch of the HODLer airdrop event for the SOGNI Token, with a total airdrop amount of 5,263,158 SOGNI. The event will last until July 2, 2025, 18:00 ) UTC+8 (, and users holding GT can participate in the distribution of airdrop rewards according to their holding proportion. Sogni AI is a decentralized creative AI inference network that supports parallel batch processing of over 100 models and 512 images. The project uses a dual-layer DePIN GPU architecture, and the SOGNI Token will be used for transactions, staking, governance, and model royalties. The project has received support from several well-known institutions including Comma3 Ventures and Republic Ventures, with a total token supply of 10 billion, deployed on both the BASE and Etherlink chains.

Jin10 data reported on June 27th, on June 27th, Tencent released and Open Source the HunYuan-A13B model. According to the introduction, as a large model based on the Mixture of Experts (MoE) architecture, it has a total of 80 billion parameters and 13 billion active parameters. While its performance is comparable to top open source models, it significantly drops inference latency and computational costs; under extreme conditions, it can be deployed with just one mid-range GPU card.

PANews reported on June 25th that Matter Labs, a company from ZKsync, unveiled a new type of Zero-Knowledge Proof tool called 'Airbender' at the Permissionless conference. This proof tool is built on the RISC-V architecture and can complete ETH Block verification in 35 seconds using a single GPU, significantly faster than similar products. The CEO of Matter Labs stated that this breakthrough will reduce Transaction Cost to less than 1 cent, suitable for high-frequency trading, small payments, and Decentralization social scenarios. Currently, Airbender is still in the early stages of release, but Matter Labs has deployed an application for easy testing. If everything progresses as planned and the ZKsync governance process is approved, Airbender will be included in the protocol upgrade by the end of this summer.

Jinse Finance reported that the second layer of ETH, ZKsync, announced the launch of the Open Source RISC-V zkVM 'ZKsync Airbender', which supports rapid verification of ETH transactions on a single GPU. For example, 'ZKsync Airbender' can achieve a basic verification speed of about 21.8 MHz on the H100 GPU, about 3.45 MHz on the SP1 Turbo, and about 1.1 MHz on the RisZero. This means that ZKsync OS (ZKsync's zkEVM execution environment) can achieve end-to-end ETH Block proof on a single GPU within 35 seconds, thereby reducing on-chain costs. It is reported that ZKsync Airbender has completed internal audits and is currently in the early testing phase.

Spheron Network's latest data shows that its annual recurring revenue has surpassed 4 million USD, with over 27,463 nodes deployed worldwide. The launched Spheron Compute 2.0 provides cost-effective GPU computing services and collaborates with multiple DePIN projects and AI model hosting platforms to promote decentralization in cloud computing.

Gate News bot news, ionet has launched full network earnings (TNE), allowing all GPU network payments and earnings to circulate on-chain through Solana. Since its launch, it has recorded over 14.5 million dollars in revenue. Now, each transaction is hosted by smart contracts, and IO is repurchased daily from the earnings.

The DPIN Foundation has reached a strategic cooperation with Nanyang Biotechnology to utilize its stable and efficient GPU Computing Power network to support the research projects of Nanyang Biotechnology, promoting the integration of AI and DePIN technology, with the aim of building a global open computing network.

Introduction Bitcoin mining, a key term in this digital gold rush, is attracting the attention of countless investors. From choosing the best mining machines to joining quality mining pools, this article will reveal the entire process of Bitcoin mining. Whether you are a beginner or an expert, you can find practical strategies and tips here to enhance your profits. Let us delve into this world of digital gold filled with opportunities together. Unveiling Bitcoin Mining: Earning Digital Gold from Scratch Choose the best mining machine: a perfect balance between performance and budget Bitcoin mining has become a highly specialized field, and choosing the right mining machine is crucial. Currently, the mainstream Bitcoin mining machines on the market are mainly divided into two categories: ASIC (Application-Specific Integrated Circuit) and GPU (Graphics Processing Unit). ASIC miners are typically more efficient due to their design specifically for Bitcoin mining, but they are also more expensive. On the other hand, GPU miners, while less efficient, offer multifunctionality.

Gate News bot News: Optalysys, a British startup focused on secure computing, has launched what is said to be the world's first blockchain server that can process data at scale without decryption. The company's LightLocker node is a server with fully homomorphic encrypted (FHE), a mathematical technique that allows computations on encrypted data without compromising encryption. In the last year, several companies have raised funds to explore the use of FHE in the cryptocurrency space. Optalysys says its server hardware is designed specifically for blockchain encryption, providing a more economical alternative to expensive and inefficient GPU systems that consume 40% less energy.

BlockBeats News, on June 16, according to CoinDesk, Optalysys, a UK-based startup focused on secure computing, announced the launch of what is said to be the world's first server designed specifically for blockchain, capable of large-scale data processing without the need for decryption. The company's LightLocker node is a server with fully homomorphic encryption (FHE) technology. This is a mathematical method that allows computations to be made directly on encrypted data without breaking the encryption itself. Over the past year, several companies have raised capital in the cryptocurrency space and explored the application of fully homomorphic encryption (FHE). Optalysys says its server hardware is purpose-built for blockchain encryption, offering a more cost-effective alternative to expensive and inefficient GPU systems, with 40% less energy consumption

Gate News bot message, cybersecurity company Kaspersky reported a mining attack incident targeting Russian industrial equipment. The hacker organization Librarian Ghouls (also known as Rare Werewolf) has been launching phishing email attacks against Russian industrial enterprises and engineering schools since December 2024, affecting Belarus and Kazakhstan as well. The attackers infiltrated hundreds of devices through disguised official documents. The organization remotely disabled the security systems of the target devices between 1 AM and 5 AM, utilizing the devices' RAM, CPU, and GPU to conduct cryptocurrency Mining. According to a Kaspersky investigation, the organization used legitimate third-party software instead of homemade malware to carry out the attack, displaying characteristics of hacker activism. Source: Cointelegraph.

Golden Finance reports that the hacker group Librarian Ghouls (also known as Rare Werewolf) has infiltrated hundreds of Russian devices and is using them for crypto mining. The group spreads malware through phishing emails disguised as legitimate organizations, establishing remote connections after infecting the devices and disabling security systems such as Windows Defender. Hackers collect information on the devices' RAM, CPU cores, and GPU to optimize the configuration of the crypto mining programs. This hacking incident began in December 2023, with the attack primarily affecting industrial enterprises and engineering schools in Russia, while there were also victims in Belarus and Kazakhstan. Kaspersky speculates that Librarian Ghouls may be hacktivists, as they rely on legitimate third-party tools rather than developing their own malware.

NVIDIA CEO Jen-Hsun Huang stated that the company is developing a revolutionary storage platform, including the new query system IQ/NVIDIA AIQ, which will use stacked GPUs to replace traditional CPUs for processing meaning discovery of unshared data, emphasizing that future storage needs to be equipped with GPU computing nodes to address the challenges of data transformation.

Jin10 Data reported on May 19 that Nvidia CEO Jen-Hsun Huang stated that we have launched the new Blackwell RTX Pro 6000 series motherboard, which connects eight GPUs and serves as a switching network called CX8, marking the emergence of a new class of chips. The CX8, which has entered mass production, facilitates communication between GPUs through PCI Express, showcasing an astonishing network bandwidth of 800Gbps. Each GPU is equipped with its own network interface, allowing for excellent east-west high-speed communication between GPUs, exceeding performance expectations.

NVIDIA CEO Jen-Hsun Huang announced the launch of NVLink Fusion, which supports multiple hardware and aims to strengthen the construction of enterprise AI infrastructure, with customizable configurations for CPU, GPU, and ASIC, addressing the complexity challenges of building enterprise AI environments.

The co-founder of Kava Labs stated that decentralized AI has practicality and potential, and is not a short-term bubble. Kava launched an AI platform and marketplace, including blockchain-native chat Bots and decentralized GPU infrastructure, winning the market. He emphasized that the key to future AI development lies in Open Source transparency, rejecting black-box models.

Golden Finance reported that OpenAI CEO Sam Altman posted on the X platform, stating that the company is currently facing serious GPU Computing Power pressure, which may lead to delays in the release of new versions and occasional service interruptions or slowdowns. He called for anyone with 100,000-level GPU resources to contact them as soon as possible, as the team is working hard to resolve the issue.

PANews, April 1 news, OpenAI CEO Sam Altman posted on the X platform stating that the company is currently facing severe GPU Computing Power pressure, which may lead to delays in the release of new versions, occasional service interruptions, or slowdowns. He called on anyone with GPU resources in the range of 100,000 to contact them as soon as possible, as the team is working hard to resolve the issue.

ChainCatcher news, OpenAI founder Sam Altman stated that it is really interesting to see people enjoy using the (generated) image feature in chats. However, our GPUs are overloaded, and while we are working to improve efficiency, we will temporarily introduce some rate limits on the image generation feature of ChatGPT.

According to a message from the Gate.io News bot, OpenAI founder Sam Altman recently stated that due to heavy GPU load, the company will implement temporary rate limits on the image generation feature of ChatGPT. Altman stated that while the enthusiasm of users for the image generation feature during chat is encouraging, this limiting measure will have to be implemented while the team works to improve system efficiency.

Jin10 data reported on March 28th, OpenAI founder Sam Altman stated that it is very interesting to see people enjoy using the (generated) image feature in chats. However, our GPUs are overloaded, and while we strive to improve efficiency, we will temporarily introduce some rate limits on the image generation feature of ChatGPT.

On March 20, Jinshi Data reported that Morgan Stanley stated that Tencent Holdings (00700.HK) is expected to benefit from returns brought by artificial intelligence (AI) investments, along with sustainable advertising revenue growth, demand from enterprise (2B) clients, and monetization of consumer market (2C) client applications. They are committed to capital returns and improving profit margins, making it their preferred stock. The target price was raised by 13.5% or HKD 75 from HKD 555 to HKD 630, and the "Buy" rating is reaffirmed. In the best-case scenario, the target price could rise to HKD 790, while in the worst-case scenario, it could drop to HKD 430. Morgan Stanley expects Tencent's capital expenditure this year to be around RMB 94 billion, accounting for 13% of total revenue, which is consistent with market expectations, primarily for GPU procurement. Additionally, it is expected that enterprise service revenue will grow by 20% year-on-year this year, and the overall revenue and operating profit growth forecast under non-International Financial Reporting Standards has been raised to 10% and 16%, respectively, with an operating profit margin expected to reach 38.