deployment

The co-founder of the on-chain perpetual contract platform Hyperliquid, Jeff, stated that after the next network upgrade, HIP-4 will increase the deployment limits. The maximum number of active HIP-4 results that a single deployer can run simultaneously will increase from 100 to 200, and the maximum number of results that a single deployer can deploy daily will also increase from 500 to 1000.Jeff added that this adjustment was made based on feedback from the community Builders. In addition, the funding rate cap will also be adjusted from 4% per hour to 0.5% per hour in the next upgrade.

SpaceX has confirmed the deployment of 27 Starlink satellites.

The privacy-oriented public chain Midnight is preparing to open smart contract deployment on the mainnet.

Elon Musk stated that the Starlink V3 constellation will begin deployment this month, and the bandwidth ultimately provided will exceed more than 100 times that of the current constellation of 11,000 satellites.

Investor Joanna Lichter published a post outlining the working hypothesis of value accumulation in the robotics field, which she believes has a barbell structure: one end consists of horizontal robotic foundation models (RFM) and infrastructure platforms that achieve compound growth through data, computing power, and ecosystems; the other end consists of category-leading application companies that capture value by transforming intelligence into useful work. The middle layer lacks attractiveness, including interchangeable models, easily replicable tools, and deployments lacking software leverage.She pointed out that, unlike large language models, robotics faces distribution costs, scarce physical data, real-time computing power limitations, and security constraints, presenting a gated diffusion rather than immediate large-scale adoption. If deployment can form a learning feedback loop, reducing the manpower required per robot and accumulating reusable data and operational knowledge, it can become a moat. Vertical entry points can accelerate this path, allowing value to ultimately accumulate at both ends of the technology stack.

Cerebras released its fourth-generation AI inference system CS-4 this week, based on the same 5nm WSE-3 wafer, achieving double the performance by doubling the clock frequency and power consumption. A single CS-4 cabinet accommodates 3 wafers (CS-3 has 2), featuring a modular "backpack" design that simplifies manufacturing and deployment, with a TDP of approximately 125 to 135kW. The CS-4 can provide an inference speed of nearly 4000 tokens/second/user, about twice that of the CS-3, and supports decomposed inference with heterogeneous systems such as AMD and AWS Trainium.Cerebras claims that the CS-4 offers about 2000 times the on-chip memory bandwidth of NVIDIA's Rubin (43PB/s), but the 44GB SRAM capacity remains unchanged, and long-context inference still requires multi-wafer stacking. For example, with the DeepSeek V4 Pro (1.6T parameters), approximately 20 systems are needed for a 1M context window, and about 40 systems are required for 256 concurrent users, corresponding to a CAPEX exceeding 20 million USD. Cerebras is collaborating with clients such as OpenAI and plans to achieve approximately double performance improvements each year, aiming for a 20-fold throughput increase by 2027. The "backpack" cabinet design of the CS-4 will continue into the next-generation "Nexus" platform.

Eclipse founder Neel Somani published an analysis stating that by leasing diesel generator sets, AI data centers can bypass grid access bottlenecks and complete deployment within months. Taking a 50MW inference data center as an example, about 40 leased diesel units are needed (monthly rent of approximately 2 million USD), consuming about 2.5 million gallons of diesel each month (approximately 12.5 million USD), with the cooling system costing about 1 million USD, and the total hardware investment being about 1.9 billion USD (including approximately 21,000 GPUs and supporting network equipment). Based on similar leasing agreements like Colossus 2, monthly revenue could reach approximately 180 million USD.

Uniswap announced that it will complete the deployment when the September Arc mainnet officially goes live, providing developers with deep on-chain liquidity infrastructure to support the construction of DeFi applications such as lending, structured products, LP strategies, and token issuance. Arc is a public chain under Circle, using USDC as the Gas token, and features sub-second deterministic finality.

According to SemiAnalysis, AMD reached a performance-based warrant agreement with Meta and OpenAI from the end of last year to the beginning of this year. The warrants will vest in batches based on the two companies' procurement and deployment of up to 6GW of AMD Instinct GPUs, while AMD's stock price must exceed a tiered target (up to $600).As AMD's stock price has risen from around $200 in early 2026 to the current $500-600, SemiAnalysis estimates that the GPU capital costs for Meta and OpenAI have decreased from about $2.15/hour and $2.50/hour to approximately $1.60/hour and $1.75/hour, respectively. If AMD's stock price rises to $600-700, the capital cost of the GPUs themselves will approach zero; if it further exceeds $1000, the total capital cost of the entire cluster (including servers, networks, and CPUs) will effectively be zero for both companies. This analysis is based on the assumption of linear vesting and the premise that the entire 6GW capacity is deployed, and the specific terms of the warrants have not been disclosed.

According to Jinshi reports, SpaceX has confirmed the successful deployment of 24 Starlink satellites.

Hyperliquid News posted on platform X that the HIP-4 permissionless deployment has gone live on the testnet.

According to Aave founder Stani Kulechov, after completing a comprehensive review, Aave will gradually phase out 50 low adoption rate asset reserves across multiple deployment environments and orderly stop deployments on Sonic, Scroll, zkSync, Metis, Soneium, and Aptos, involving an additional 25 asset reserves.In addition, 21 expired Pendle PT assets will also be eliminated and replaced by new expiring assets. This adjustment involves approximately $98.1 million in supply assets and $15.6 million in debt. Stani stated that this move aims to reduce the economic and technical risks of the protocol based on the new Aave risk framework and technical asset listing framework, and Aave will continue to assess the risks of assets in each deployment environment.

According to The Block, Tether's compliant stablecoin USA₮ (USAT) has officially launched on the Celo mainnet, marking its second mainnet deployment after Ethereum. The token is issued by Anchorage Digital Bank and can be natively minted and redeemed on Celo, and can be directly used for on-chain gas payments through Celo's fee abstraction mechanism.USAT launched in January this year, with a current market capitalization of approximately $185 million.

According to SolanaFloor, the Solana ecosystem perpetual contract platform Bulk has launched BIP-1, allowing anyone to deploy their own perpetual market on the mainnet with almost zero upfront costs before going live.Successful markets can subsequently join BULK's portfolio margin system, allowing traders to use up to 70% less collateral.

According to CNBC, AMD has officially launched its first rack-scale AI system for artificial intelligence, Helios, which is seen as AMD's important product to compete with NVIDIA's Grace Blackwell and Vera Rubin AI systems. It has been adopted by enterprise customers such as Microsoft, Meta, OpenAI, and Oracle, attempting to initiate more direct competition with NVIDIA in the AI infrastructure market.Microsoft announced that it will deploy the AMD Helios system in its Azure data centers, becoming the latest customer to adopt this platform. AMD stated that Helios is expected to start delivering to customers later this year and will be used to support cutting-edge AI model inference, Azure AI services, and enterprise-level AI applications.Market analysis firm Futurum Group data shows that NVIDIA currently holds over 95% of the data center GPU market share, while AMD accounts for about 4.5%. Analysts believe that if Helios is deployed successfully, AMD has the opportunity to capture 20%-25% of the market share in the future, corresponding to a potential market space worth hundreds of billions of dollars.

According to The Block, Hyperliquid announced that it will introduce a permissionless deployment mechanism for the HIP-4 outcome markets in future upgrades. According to the disclosure plan, validators will vote on outcome templates to ensure market quality and clarity of expression.In addition, HIP-4 market deployers are required to stake 500,000 HYPE and can receive up to 50% of the fee share from the markets they create.

Aave V4 has been deployed to the Avalanche network, marking the first multi-chain deployment of the Aave V4 Hub & Spoke architecture. This deployment includes a core liquidity hub and three sub-markets: the main market supports lending of assets such as WAVAX, BTC.b, USDC, USDT, and WETH.e; the AVAX-related market focuses on AVAX liquidity staking strategies, allowing users to deposit sAVAX with a 95% collateral rate and borrow WAVAX; the foreign exchange market supports mutual collateralization and lending among three stablecoins: EURC, USDC, and USDT.The core liquidity hub adopts a shared liquidity pool design, with all sub-markets sharing the same pool to avoid liquidity fragmentation. Aave stated that since its deployment on Avalanche in 2021, the historical cumulative inflow for V2 and V3 has exceeded $15 billion, and the current stablecoin utilization rate in the V3 market is over 90%.

According to KPMG's latest survey report involving 2,145 executives from 20 countries, as technology companies like Anthropic, OpenAI, and GitHub recently shifted some of their AI services from fixed subscription models to usage-based billing, businesses are facing challenges in cost forecasting and management during the scaling of AI deployment.The report indicates that 29% of corporate executives find it difficult to understand and control operational costs when scaling AI deployment, and one-third of executives believe that insufficient understanding of AI economics hinders the deployment of AI entities. Due to costs exceeding expected value, nearly half (about 49%) of corporate organizations have chosen to delay or readjust their AI deployment plans; meanwhile, low-cost, high-fidelity large models are accelerating their impact on corporate AI strategies.In addition, tech giants are increasing capital expenditures to build AI capacity. Amazon plans to spend about $200 billion on capital expenditures this year and is investing $1 billion in its AWS frontline engineering organization to assist customers in adopting AI entities; Microsoft's total capital expenditure is expected to reach $190 billion this year, with $2.5 billion allocated to the new entity Microsoft Frontier Company. KPMG emphasizes that, in addition to cost pressures, accountability in AI governance, employee engagement rules, and the prevention of system "hallucinations" remain core challenges faced by businesses today.

According to Jinshi reports, Broadcom is expected to deploy OpenAI chips faster than anticipated.

Samsung Electronics announced the deployment of ChatGPT Enterprise and Codex to all its employees in South Korea and employees in the global Device Experience (DX) department. This collaboration is one of OpenAI's largest enterprise-level deployments to date.It is reported that Samsung Electronics plans to widely apply these AI tools in core business areas such as software development, product research and development, marketing, and manufacturing, covering both technical and non-technical positions, to enhance employee productivity and problem-solving capabilities. This deployment also marks the expansion of Samsung's partnership with OpenAI from the early stage of AI infrastructure (advanced memory semiconductor supply) to the enterprise-wide AI office application level.