Most large-cap U.S. tech stocks declined in pre-market trading, with Tesla and Nvidia falling around 1%.

According to monitoring by OnchainLens, Cumberland has deposited assets worth approximately $6.65 million into trading platforms. The holdings include 108,090 HYPE tokens valued at around $5.95 million transferred to Bybit, and 700,000 USDT transferred to Binance.

According to PeckShield monitoring, the address marked as the attacker of Across Protocol has returned 331.8 ETH, valued at approximately $623,900, to the multi-signature address of the Across Protocol Hub Pool Owner. Earlier, Across Protocol suffered an attack on Solana, resulting in the theft of around $3.6 million worth of crypto assets.

South Korea’s largest cryptocurrency exchange Upbit has designated STORJ as a trading warning project and suspended deposits for the token. According to HTX market data, STORJ’s price has seen a continuous gradual decline following its bankruptcy announcement. Its parent company, Storj Labs, announced it has voluntarily filed for Chapter 11 bankruptcy reorganization under U.S. bankruptcy code to resolve historical debts and continue operations.

Citrini analyst Jukan, citing TrendForce’s latest industry survey, noted that the three leading MLCC (multi-layer ceramic capacitor) manufacturers in South Korea and Japan posted their highest monthly shipments in five years in June: Murata hit 140 billion units, Samsung Electro-Mechanics 98 billion units, and Taiyo Yuden 40 billion units. The growth was mainly driven by strong AI demand, a trend that has continued into July. Cloud service providers’ procurement momentum for self-developed ASICs (such as Google TPU and AWS Trainium) remains robust, boosting demand for high-capacitance, low-voltage, small-sized MLCCs. However, AI-grade production capacity is diverting supply from consumer applications: South Korean and Japanese manufacturers are rapidly shifting their consumer-grade X5R MLCC production capacity to higher-spec X6S/X7R products for AI use, putting pressure on medium-to-high capacitance consumer MLCC (1μF to 22μF) capacity, with inventories of key products generally falling to 30 days or less. Jukan pointed out that supply tightness has triggered a knock-on effect, leading to a surge in distributors’ emergency orders. Order spillover has benefited Taiwanese and Chinese mainland suppliers, driving channel prices sharply higher: Distributors that stocked up in advance raised prices by an average of 20% to 25%, while spot market prices rose to two to three times their original levels. The market shows a structural divergence of “weak end demand but firm channel prices”: On one hand, consumer electronics remain sluggish due to factors like renewed inflation concerns from US-Iran geopolitical tensions and Apple’s recent price hikes for MacBook and iPad; on the other hand, AI bottlenecks are spreading from GPUs and memory to the smallest on-board components, with production capacity reallocation and channel scarcity being the main drivers of current price strength.

According to market data from BIT (bit.com), US pre-market optical communication concept stocks extended their losing streak, with declines as follows: Marvell Technology (MRVL) fell 2.85%; Applied Optoelectronics (AAOI) dropped 3.11%; Lumentum Holdings (LITE) decreased 2.24%; Coherent Corp (COHR) fell 3.31%; and Ciena Corporation (CIEN) dropped 2.7%.

Hyperliquid officially responded to the price pinning incident in its SK Hynix (SKHYNIX) perpetual contract market, explaining how the HIP-3 market operates. Hyperliquid is a permissionless blockchain, where different teams can deploy and run markets on it as an infrastructure layer. The SKHYNIX perpetual contract was deployed and operated by the Trade.xyz team, which is currently investigating the incident and will share updates once a conclusion is reached. Secondly, HIP-3 deployers push their market’s mark price, oracle price, and external perpetual price inputs. Based on today’s posts and support tickets, it may be necessary to walk through this mechanism. Deployers can choose to adopt a mark price method similar to that used for perpetual contracts operated by validators (such as BTC). In this case, the protocol contributes one of three median components: the median of on-chain data (latest trade price, best bid, best ask). The other two components are pushed by the deployer and influence the final mark price. For a simplified example: if the median of on-chain data (latest trade price, best bid, best ask) is 100, but the deployer pushes (150, 151), the mark price will be 150. Note: As of press time, Hyperliquid officials have not provided further clarification. For more related reports, see "The Truth Behind Hyperliquid’s Pinning Incident: $868 Leveraged $500M Hynix Contract Market".