Ark Invest increased its holdings in Tesla and Circle yesterday, while reducing its stakes in Figma and Robinhood.

A whale staked 2.93M $HYPE ($172M)! Over the past 24 hours, 19 wallets (likely belonging to the same whale) deposited 2.93M $HYPE ($172M) into Hyperliquid and staked it. The 2.93M $HYPE was accumulated 9 months ago at an average price of $44 and is now up ~$44.5M.

Optical communication chip supplier MaxLinear has released its Q2 FY2026 financial results, reporting total revenue of $169 million, up 55% year-over-year. Diluted earnings per share (EPS) came in at $0.02, compared to -$0.52 in the prior-year period, marking a 106% YoY increase. Net profit stood at $2 million, up 107% YoY. Non-GAAP net profit reached $34 million, surging 1854% YoY, while Non-GAAP EPS was $0.35, versus $0.22 in the same period last year. MaxLinear projects its next fiscal quarter revenue will range from $210 million to $220 million. GAAP gross margin is expected to hit 57.0% to 60.0%, and Non-GAAP gross margin is forecast at 58.5% to 61.5%. The company anticipates Non-GAAP operating expenses of $66 million to $71 million, with a fully diluted share count of 99 million. According to market data from BIT (bit.com), MaxLinear’s shares plummeted 11.22% in after-hours trading, changing hands at $81.

AMD has launched the Ryzen AI Embedded X100 series processors. The chips utilize the same Strix Halo design as the Ryzen AI Max series but feature a distinct positioning, targeting robotics, industrial automation, healthcare, and other embedded computing systems. AMD previously previewed the X100 series at CES 2026, though no further specific details were disclosed at the time. The newly released X100 series brings the Strix Halo architecture to physical AI applications, integrating Zen 5 CPU cores, RDNA 3.5 GPU, and XDNA 2 NPU AI acceleration units, designed to deliver high-performance edge computing for devices like robots that require real-time perception and AI inference. The series supports up to 16 Zen 5 CPU cores and adopts a unified memory architecture, meeting the demands of robotics and industrial equipment for running large-scale AI models, visual processing, and autonomous decision-making. AMD aims to expand its presence in the physical AI and edge intelligence markets via the X100 series, competing with vendors including NVIDIA for the next-generation robotics computing platform market.

According to Reuters, the U.S. Department of Defense (DoD) announced Thursday it has struck a nearly $7 billion, up-to-10-year deal with Oracle to consolidate the Pentagon’s existing on-premises software licenses into a single unified contract. This marks the latest initiative by the Pentagon’s top technology official to cut costs and eliminate fragmented procurement practices. The so-called "enterprise software agreement" was negotiated under the lead of the U.S. Navy, with the contract signed between the DoD and Oracle. The deal primarily covers the use of Oracle’s on-premises computing resources, aiming to boost software procurement efficiency, lower operational costs, and further optimize the information technology (IT) infrastructure of defense systems.

According to Bloomberg, NVIDIA (NVDA.US) and Amkor Technology (AMKR.US) have signed a $1.5 billion agreement to support the latter in expanding its chip packaging capacity, as part of a broader initiative to grow semiconductor operations in the U.S. Per a Thursday statement, the deal includes an upfront payment from NVIDIA that will help Amkor enhance its manufacturing capabilities in Arizona. The news sent Amkor’s shares surging roughly 15% in after-hours trading. The two companies said the collaboration will focus on chip packaging and testing technologies for the artificial intelligence sector. Packaging is a critical final step in semiconductor manufacturing, during which chips are housed in enclosures to enable connection with other components.

ARK Invest Research Head Lorenzo Valente published a report noting Blockworks data indicates Hyperliquid’s real-world asset (RWA) transactions built on the HIP-3 protocol are growing rapidly. Last week, Hyperliquid’s RWA trading volume surpassed crypto asset trading for the first time, accounting for 54% of the platform’s total trading volume. Individual stock trading emerged as the fastest-growing segment: since June, its volume has outpaced indices and commodities, now making up roughly 61% of Hyperliquid’s total RWA volume, signaling on-chain financial markets are expanding beyond pure crypto assets to traditional assets. Data shows last week, total perpetual contract trading volume across decentralized exchanges (DEXs) hit ~$79 billion, with Hyperliquid contributing ~$50 billion. Of that, HIP-3 RWA volume reached $26 billion—even exceeding the combined crypto perpetual contract volume of all other DEXs. RWA trading may not fully consolidate around existing crypto trading platforms; distinct asset classes could develop their own leading players, and future competition among trading platforms may shift from vying for crypto asset traffic (such as BTC, ETH, SOL) to competing for on-chain trading demand for real assets like stocks and commodities.