gpu

Cryptocurrency platform Bullish announced a $100 million debt financing line to the stablecoin protocol USD.AI to support GPU-collateralized loan operations. This financing will help USD.AI provide liquidity for artificial intelligence infrastructure development, connecting on-chain capital with computing power financing needs through tokenization.USD.AI, developed by Permian Labs, currently has over $225 million in crypto assets locked within the protocol. Bullish Exchange also plans to launch the sUSDai token of USD.AI and open multiple trading pairs, providing investors with access to a secondary market channel for GPU-collateralized debt.David Choi, CEO of Permian Labs, stated that computing power is becoming an independent credit market, and this financing will help USD.AI fund more AI development while creating a more transparent and in-depth computing power credit market.

According to CoinDesk, the cryptocurrency platform Bullish will provide USD.AI with $100 million in stablecoin debt financing to support loans collateralized by GPUs and other high-performance computing assets, promoting the construction of artificial intelligence infrastructure.

NVIDIA's revenue for the second quarter of fiscal year 2027 increased by more than double year-on-year to $96.22 billion, exceeding market expectations of $92.17 billion; adjusted earnings per share were $2.22, surpassing the expected $2.10. Data center revenue grew more than double to $89 billion, higher than the expected $85.08 billion. The company expects third-quarter revenue of $108 billion (with a fluctuation of 2%), above analysts' expectations of $104.19 billion; the adjusted gross margin is expected to be 74% (with a fluctuation of 0.5 percentage points), lower than the market expectation of 74.77%.NVIDIA expects revenue to grow by about 70% in fiscal year 2028 and to expand its collaboration with AWS, with both parties planning to deploy an additional 2 million NVIDIA GPUs in Amazon's global infrastructure in 2027 and 2028. The company stated that revenue opportunities for AI chips could exceed $1 trillion before 2027. The third-quarter guidance does not include any sales from data center chips in China.After the earnings report was released, NVIDIA's stock price rose over 4% in after-hours trading. As of Wednesday's close, the market capitalization was approximately $5.16 trillion, with the stock price having increased by 12.4% year-to-date.

Venture capitalist Tomasz Tunguz pointed out that the narrative of AI infrastructure resembles a slow relay race, with bottlenecks sequentially transmitting from GPUs to memory, CPUs, and storage, each link freezing the supply chain of the next for years and locking in higher baseline costs. At the beginning of 2023, the GPU shock caused H100 rental prices to exceed $9 per hour, and server shipments fell by 22%; subsequently, manufacturers shifted capacity to HBM, leading to an 80% quarterly increase in enterprise SSD prices and over a 60% rise in DRAM.By the end of 2025, the workload of intelligent agents will push the CPU to GPU ratio to about 1:1, with the average price of server CPUs rising by 27% year-on-year; in 2026, nearline HDD annual capacity will be sold out. The construction cost of data centers has risen to about $20 billion per gigawatt, with orders for long-cycle equipment such as transformers and turbines scheduled as far out as 2029 to 2031.Tunguz referred to this as the long whip effect in the hardware sector: years of manufacturing delays amplify downstream demand shocks upstream, and when pressure is relieved at a certain bottleneck, it will be delayed in transmitting to the next link, with transformers scheduled for delivery in 2027 to 2028, NAND wafer fabs, and turbine production lines potentially facing the risk of overcapacity.

According to a Goldman Sachs report on August 20, CoreWeave's Q2 revenue met expectations, with an EBIT margin exceeding market consensus by 200 basis points, and the 2026 revenue guidance surpassing market expectations by 1%. Revenue backlog increased by 5% quarter-over-quarter to $104 billion, with over $25 billion in new committed orders since Q3. Active power installations rose from 1GW in Q1 to over 1.5GW, with contracted power installations reaching 4.2GW. Goldman Sachs raised the 12-month target price from $121 to $139, indicating a 53% upside from the current stock price, maintaining a neutral rating.Goldman Sachs believes that CoreWeave's short-term certainty is clear: demand continues to outpace supply, pricing for new and old GPU generations remains high, and capacity is expanding as expected. The new generation of chips (Blackwell, Vera Rubin) is continuously setting new price highs, and recent A100 contract deliveries have been extended to 2029. The proportion of enterprise customers is increasing (Caterpillar, IBM, Nissan, ZF), and AI computing power demand is spreading from tech giants to the real economy. Goldman Sachs expects EBITDA to grow from $3.1 billion in 2025 to $31.3 billion in 2028. The neutral rating reflects a wait-and-see approach until software and platform services become more certain contributors to profit margins before making a more positive judgment.

According to BBX data, last weekend, publicly listed companies in the U.S. stock market disclosed the latest real accounts regarding digital asset treasury adjustments, staking yields, and computing power infrastructure financial data. The core dynamics are as follows:Hyperscale Data ( NYSE : $ GPUS ) sold 685 BTC for $43 million: AI data center company Hyperscale Data officially announced that it has sold approximately 685 bitcoins in the open market, obtaining about $43 million in cash. After this reduction, the company still holds approximately 275 BTC. The funds will mainly be used for the ongoing development and expansion of its Michigan data center, as well as to optimize its debt and equity capital structure to support its computing power transformation strategy.Solana concept stock HSDT ( NASDAQ : $ HSDT ) recorded 31,200 SOL staking yields in Q2: Nasdaq-listed Solana treasury and ecosystem company HSDT disclosed its performance for the second quarter of 2026. The company's total revenue for the quarter was $2.5 million, mainly contributed by the 31,200 SOL staking rewards obtained during the quarter; the net loss for the quarter was $30.3 million. As of June 30, HSDT's total assets reached $176.1 million, including $147.3 million in long-term digital assets, related positions, and fund investments.Soluna ( NASDAQ : $ SLNH ) Q2 revenue increased by 145% year-on-year, repaid $19.1 million loan early: Green data center listed company Soluna released its Q2 financial report, with total revenue reaching $15.1 million (a year-on-year increase of 145%). Among them, data hosting revenue surged to $12.65 million, while cryptocurrency mining revenue decreased to $1.72 million; the net loss was $22.6 million, with an adjusted EBITDA loss of $1.6 million. As of the end of the quarter, the company held unrestricted cash of $113.4 million. The financial report disclosed that the company fully repaid approximately $19.1 million of Generate debt early on August 10.

Luke Dashjr has randomly determined that BLAKE2b will be the new POW algorithm for the BIP-110 chain, expected to replace the currently used SHA-256d algorithm in Bitcoin. Compared to SHA-256d, BLAKE2b is more suitable for mining with general-purpose CPUs/GPUs, being more CPU-friendly, which can reduce the advantages of existing Bitcoin ASIC miners.Previously, it was reported that Luke Dashjr plans to officially implement the BIP-110 hard fork on September 1, making it an independent blockchain and generating a new token distinct from Bitcoin. Luke Dashjr has had his Bitcoin BIP editing permissions revoked by the Bitcoin BIP editing team.

Lambda, an AI cloud computing service provider supported by Nvidia, is financing $917 million through the leveraged loan market to procure AI chips. As the construction of artificial intelligence infrastructure accelerates, chip financing is becoming a new way for capital investment in the AI industry. Lambda belongs to the rapidly developing "new cloud vendor" camp in recent years, with its main business being to provide GPU computing power and AI infrastructure services to enterprises and developers.This financing plan will be completed through a loan based on GPU asset-related rights, aimed at supporting the company's expansion of AI computing resources. Reports indicate that AI infrastructure companies are actively exploring new financing methods to meet the enormous capital investment required for building large-scale computing clusters. Previously, AI cloud service provider CoreWeave completed the first transaction in the institutional leveraged loan market for chip financing, providing a new financing model for the industry. As the demand for generative AI continues to grow, Nvidia's GPU supply has become a core resource for AI companies' expansion. By using GPU assets as the basis for financing, AI cloud service providers can rapidly scale their computing power without fully relying on equity financing, while also allowing the traditional credit market to participate in the wave of AI infrastructure investment.

The domestic GPU manufacturer Moore Threads released its semi-annual report for 2026, with revenue of 1.736 billion yuan, a year-on-year increase of 147.42%. The net profit attributable to the parent company was a loss of 11.56 million yuan, narrowing by about 259 million yuan compared to the same period last year; the net profit attributable to the parent company after deducting non-recurring gains and losses was a loss of 151 million yuan, narrowing by 52% year-on-year. Revenue in the second quarter was approximately 999 million yuan, a quarter-on-quarter increase of 35.4%. Research and development expenses were 769 million yuan, accounting for 44.3% of revenue. The company stated that the improvement in performance was driven by the demand for AI computing power and the accelerated commercialization of the Kuage Intelligent Computing Cluster.As of the end of the quarter, the book value of inventory was 3.55 billion yuan, an increase of 166.5% compared to the end of last year, reaching a four-year high. The company stated that this was due to proactively increasing stock to meet market demand. On the same day, Moore Threads announced plans to issue H shares and list on the main board of the Hong Kong Stock Exchange, just about 8 months away from its listing on the Sci-Tech Innovation Board in December 2025. If successful, it will achieve an "A+H" layout. The announcement stated that this issuance still requires approval from the shareholders' meeting and regulatory approval, which carries uncertainty. Another domestic GPU manufacturer, Muxi Co., has already launched its H share plan in June, with a shorter interval. Moore Threads is accelerating the development of the new generation "Huagang" architecture and the Huashan and Lushan chips. In the first half of the year, the net cash outflow from operating activities was 2.169 billion yuan, an increase of about 86% year-on-year.

According to the latest supply chain survey by TrendForce, despite the United States implementing export controls on high-end GPUs, China's AI industry has not slowed down; instead, it is accelerating the construction of a domestic AI supply chain. The report points out that the supply structure of high-end AI chips in China is shifting from reliance on imported GPUs to a dual-track model of "domestic GPUs plus self-developed ASICs," with domestic solutions expected to capture nearly 90% of the Chinese high-end AI chip market by 2026.This year, the Chinese government is actively promoting the adoption of domestic AI chips, with policies prioritizing support for high-potential local manufacturers to significantly expand their market share in the high-end AI server market. Meanwhile, the domestic ecosystem is gradually maturing in key areas such as advanced processes, advanced packaging, and thermal management. TrendForce expects that, against the backdrop of an accelerated domestic substitution process, domestic chip manufacturers like Huawei Ascend, Cambricon, and Haiguang are likely to continue benefiting, and the related industrial chain will usher in structural growth opportunities.

On July 29, the U.S. Department of Commerce signed letters of intent with seven companies, totaling up to $874 million, to support seven "post-GPU era" underlying technology routes such as CPO, ferroelectric memory, and 3D packaging in the form of equity investments. This marks a shift in the U.S. chip strategy from "capacity reshoring" to "technology route selection."The seven companies and their technology directions include: GlobalFoundries (CPO silicon photonic integration, $300 million), Kepler Computing (ferroelectric 3D memory, $245 million), Multibeam (multi-electron beam direct-write lithography and advanced packaging, $140 million), Extropic (thermodynamic sampling unit TSU, $75 million), Thintronics (ultra-low loss dielectric materials, $50 million), Aeluma (large-size phosphorus-free optoelectronic device substrates, $30 million), and OBSIDIA (hardware zero-trust chip anti-counterfeiting, $34 million). All companies are required to provide non-controlling minority equity to the U.S. government.This move shows that the funding usage of the CHIPS Act is shifting from subsidizing wafer fabs to directly holding equity in cutting-edge technology companies with national capital, in order to secure rule-making authority in the post-Moore era.

According to The Information, Nvidia is weighing a significant adjustment measure, considering using less high-bandwidth memory for the Rubin Ultra GPU to address the shortage of advanced high-bandwidth memory chips.

According to Jinshi reports, DDN has partnered with NVIDIA (NVDA.O) to advance GPU-driven data access technology.

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 BBX data, yesterday global listed companies disclosed the latest real announcements regarding adjustments in cryptocurrency assets and business payment applications, with the following core dynamics:Hyperscale Data monetizes 100 BTC to build an AI data center: U.S. listed company Hyperscale Data, Inc. (NYSE American: $GPUS) announced the sale of 100 bitcoins to fund the construction of its AI data center located in Michigan. The company's CEO William Horne stated that this move is "converting one balance sheet asset into another," and clarified that the company's underlying belief in bitcoin has not changed.Quantum Solutions sells 1,000 ETH to inject into AIDC business: Japanese listed company Quantum Solutions announced that it sold 1,000 ETH through its merged subsidiary GPT Pals Studio Limited (GPT) for $1.9 million (approximately 311 million yen). The proceeds from the sale will be fully invested in the expansion of artificial intelligence infrastructure (AIDC) and data center business.Vida Global seamlessly accesses bitcoin payroll through the Lightning Network: U.S. listed company Vida Global (NYSE American: $VIDA) announced the adoption of Voltage Credit to pay its global team members in bitcoin via the Lightning Network. This mechanism utilizes a revolving credit line to instantly issue bitcoin and repays the balance in U.S. dollars at the end of each month, similar to processing standard vendor invoices, allowing the company to establish a cryptocurrency payroll settlement channel without directly holding a bitcoin wallet or reflecting cryptocurrency assets on its balance sheet.

Fully Homomorphic Encryption (FHE) company Zama CEO Rand Hindi stated on X on Monday that the company has achieved benchmark results of 1,000 confidential transactions per second on GPUs, an increase of about 5,000 times compared to the 0.2 transactions per second demonstrated at ETHcc in 2022. Hindi mentioned that the cost of running an FHE node is approximately $0.000004 per ERC7984 transaction, significantly lower than zero-knowledge proofs, but this comparison has not been independently verified. The benchmark has not yet been activated on any mainnet, is self-reported data from the company, and the testing methods have not been published, with plans to go live by the end of the year.Hindi also set long-term goals: 10,000 tps by 2028 and 100,000 tps by 2030. Zama has been active recently: on July 24, a confidential RFQ exchange product was launched, which within weeks shielded over $121 million USDT on Ethereum, collaborated with Elliptic to enhance compliance screening for confidential transactions, and co-founded the Confidential Token Association with OpenZeppelin and Inco to standardize encrypted ERC-20 tokens. Zama previously completed a $57 million Series B funding round.

According to a report by the WSJ, Nvidia has invested in the AI laboratory Safe Superintelligence (SSI), founded by former OpenAI chief scientist Ilya Sutskever. The two parties have reached a long-term cooperation agreement aimed at expanding SSI's computing resources while helping Nvidia secure important clients in the AI field. Both companies stated that Nvidia made a "large-scale" investment after understanding some of SSI's research progress, but did not disclose the specific amount.As part of the collaboration, SSI will receive a significant amount of Nvidia's flagship GPU resources, with its computing power expected to increase by an order of magnitude. Previously, SSI primarily relied on TPU chips provided by Google for AI research and development. This collaboration indicates that Nvidia is further expanding its AI chip ecosystem and binding future computing power demands through investments in top AI laboratories.SSI was established in 2024 by Ilya Sutskever, with the goal of developing "Safe Superintelligence." The company has previously raised about $2 billion in funding, with investors including Andreessen Horowitz and Sequoia Capital, and reached a valuation of approximately $30 billion last year. This is not Nvidia's first investment in an AI company founded by former core members of OpenAI. In March of this year, Nvidia also invested in Thinking Machines Lab, founded by former OpenAI chief technology officer Mira Murati, whose first AI model is trained on Nvidia hardware.Ilya Sutskever is one of the important researchers in the field of modern artificial intelligence, having co-founded OpenAI and promoted the development of large model technologies such as ChatGPT. However, after leaving OpenAI, he began to question the approach of solely relying on expanding data and computing power to drive AI progress, turning instead to explore new directions in superintelligence research.