Edgen Stock·Sep 16 2025, 15:16
*The webscale sector posted record revenues in 2Q25, largely *
U.S. equities closed higher on Tuesday, with the S&P
[1] Webscale's Record-Breaking Quarter: $722B Revenue and Soaring Capex | Webscale Market Tracker 2Q 2025[2] Webscale Market Tracker 2Q 2025: Webscale Infrastructure Spending Shatters Records in 2Q25 - ResearchAndMarkets.com - Business Wire[3] Tech giants' capex hits $92b as infra investments rise - Tech in Asia
Related News
Sep 16 2025, 19:43
Nvidia Refocuses Cloud Strategy on Internal R&D and Core AI Infrastructure
Nvidia (NVDA) is reportedly scaling back its external-facing DGX Cloud business, shifting its primary focus to leveraging the service internally for advanced chip design and artificial intelligence model development. This strategic adjustment reduces direct competition with major cloud providers while reinforcing Nvidia's position as a foundational AI infrastructure provider. Strategic Pivot in Cloud Services Nvidia Corp. (NVDA) is reportedly curtailing its direct efforts in the competitive cloud computing market with a significant strategic adjustment to its DGX Cloud offering. Initially positioned as a direct competitor to established hyperscalers like Amazon Web Services (AMZN), DGX Cloud will now predominantly serve Nvidia's internal research and development needs. This shift was first reported by The Information, indicating a move away from actively pursuing external business customers to concentrate on leveraging its powerful AI chip infrastructure for proprietary innovation. The DGX Cloud service, comprising servers powered by Nvidia's advanced AI chips, will be utilized by the company's own researchers for critical tasks, including the design of new semiconductor architectures and the development of sophisticated AI models optimized for these chips. While Nvidia will not entirely cease accepting new customers—as evidenced by the recent announcement that quantum and AI startup SandboxAQ will utilize DGX Cloud—the emphasis has decidedly moved inward. Navigating Competitive Pressures and Pricing Dynamics This recalibration of Nvidia's cloud strategy comes amid challenges in competing with established cloud giants. Reports suggest that initial ambitious revenue goals for DGX Cloud, once projected to potentially reach $150 billion, faced hurdles. A primary factor was the pricing structure, with some AI developers reportedly finding DGX Cloud's costs significantly higher than comparable offerings from traditional cloud providers, which limited demand and led to short-term client churn. Furthermore, Nvidia's foray into direct cloud services created tensions with its crucial partners: Amazon Web Services, Google Cloud, and Microsoft Azure. These hyperscalers are significant customers for Nvidia's GPUs, contributing a substantial portion of its revenue. DGX Cloud's expansion was perceived as a competitive threat, incentivizing these partners to explore developing their own AI chips to reduce reliance on Nvidia. By scaling back its external cloud ambitions, Nvidia aims to alleviate these tensions and strengthen its collaborative relationships with these critical customers. The company had initially pursued a $2 billion software revenue goal, which included DGX Cloud, but has now opted to prioritize its core chip sales and partnerships. Reinforcing the AI Infrastructure Backbone Nvidia's strategic pivot signals a reinforced commitment to its role as the foundational provider of AI infrastructure rather than a direct cloud service operator. This calculated decision aims to position Nvidia as the "backbone of AI infrastructure," enabling cloud providers to deploy its cutting-edge chips across hybrid environments. The company's vision, as articulated by CEO Jensen Huang, emphasizes the "AI network as the computer," focusing on its expertise in GPUs, networking, and software to create composable infrastructure spanning various computing environments. This ecosystem-centric approach is proving financially beneficial. Nvidia's data center revenue is projected to reach $54 billion in Q3 2025, with a significant portion of its recent quarterly revenue coming from three hyperscale customers, highlighting the strength and importance of these partnerships. The upcoming Blackwell GPU, expected to ramp up production in Q4 2025, is projected to generate $210 billion in revenue for the year, further underscoring Nvidia's dominance in high-performance AI hardware. The willingness of AWS and Microsoft Azure to host Blackwell cloud instances reflects the surging demand for AI computing and the ongoing collaboration with Nvidia. The Path Forward for Nvidia and the AI Ecosystem The strategic adjustment of DGX Cloud allows Nvidia to concentrate its resources on advancing its core semiconductor and AI technologies, which are critical to maintaining its market leadership. By reallocating efforts towards internal research—including the allocation of $13 billion to acquire its own AI chips from other cloud providers—Nvidia aims to accelerate the development of new chips and AI tools that enhance chip performance. For investors, the long-term implications hinge on Nvidia's ability to innovate at scale, navigate potential antitrust scrutiny, and sustain its vital partnerships with major cloud providers. While Nvidia holds a commanding 70-95% share in the AI training market, challenges such as production delays and rising competition from AMD and Intel, alongside U.S. export restrictions, remain factors to monitor. This strategic shift is indicative of Nvidia's confidence in its technological edge and its proactive approach to managing competitive and regulatory headwinds in the rapidly evolving AI landscape. The company's success will ultimately depend on the continued adoption of its advanced hardware in hybrid cloud environments and its agility in addressing an dynamic industry.
Sep 16 2025, 19:43
China Initiates Probes into U.S. Semiconductor Sector Amidst Critical Trade Talks
China has launched two investigations targeting the U.S. semiconductor industry, including anti-discrimination and anti-dumping probes, alongside an antitrust inquiry into NVIDIA. These actions, timed with high-stakes U.S.-China trade negotiations in Madrid, underscore escalating geopolitical tensions and strategic leverage in the global technology race. U.S. equities saw a cautious response in the semiconductor sector this week as China escalated its trade offensive, initiating two significant investigations against the American chip industry. These probes, announced immediately preceding pivotal U.S.-China trade negotiations in Madrid, highlight the deepening technological rivalry and the use of regulatory measures as strategic leverage between the two economic superpowers. The Event in Detail China's Ministry of Commerce made the initial move, announcing an anti-discrimination investigation into U.S. policies concerning chip trade. This probe aims to determine if Washington's regulations unfairly target Chinese companies. Concurrently, a separate anti-dumping investigation was launched, scrutinizing suspected dumping of U.S.-manufactured analog chips. These chips are critical components found in a wide array of devices, including hearing aids, Wi-Fi routers, and temperature sensors. The Jiangsu Semiconductor Industry Association claimed that the dumping margin for certain simulation chips from the U.S. has exceeded 300%, with these chips holding a nearly 45% share of the Chinese market as of 2024. Companies implicated in the anti-dumping probe include Texas Instruments (TXN) and Analog Devices (ADI). Adding another layer to the regulatory pressure, China's State Administration for Market Regulation (SAMR) announced further inquiries into U.S. chipmaker NVIDIA. The SAMR is investigating alleged violations of China's anti-monopoly law, stemming from an earlier probe into NVIDIA's 2020 acquisition of Mellanox Technologies. While specifics of the alleged violations remain undisclosed, this action is widely perceived as a retaliatory measure against Washington's export controls on the Chinese chip sector. These actions coincided with the U.S. Commerce Department adding 32 entities, including 23 Chinese firms, to its restricted trade list, some accused of acquiring U.S. chipmaking equipment for China's top chipmaker SMIC. The backdrop to these developments is the fourth round of U.S.-China trade negotiations in Madrid, led by U.S. Treasury Secretary Scott Bessent and Chinese Vice Premier He Lifeng, where tariffs, export limits, and technology conflicts are central discussion points. Analysis of Market Reaction The immediate market reaction in the semiconductor sector was characterized by uncertainty and some downward pressure on key stocks. Following the news of the anti-monopoly probe, NVIDIA (NVDA) shares traded approximately 2.6% lower in pre-market activity. In the preceding week, Analog Devices (ADI) saw its stock decline by 0.75%, while Texas Instruments (TXN) experienced a 2.8% drop, remaining below key moving averages. However, some advanced chipmakers like NVIDIA and Advanced Micro Devices (AMD) had seen gains in the prior week, indicating a complex and often divergent market response influenced by various factors beyond direct regulatory pressure. Despite the headline-grabbing nature of China's investigations, the overall market reaction remained relatively muted. This suggests that investors are increasingly accustomed to such "tech-trade salvos," viewing them as tactical maneuvers in ongoing negotiations rather than immediate fundamental threats to the industry's long-term prospects. Broader Context and Implications These regulatory actions are integral to the ongoing "tit-for-tat" trade dispute between the U.S. and China. The U.S. has implemented bans on exports of advanced AI chips to China and placed restrictions on Chinese chipmakers, while Beijing has retaliated by delaying supplies of rare-earth minerals critical for electronics. China's current probes are widely interpreted as a strategic counter to U.S. export controls, aiming to exert leverage in the Madrid trade talks. Beyond immediate negotiation tactics, China's moves align with its broader strategic objective of achieving semiconductor self-sufficiency. This is particularly evident in the focus on analog chips, which are generally considered less complex to localize compared to advanced AI accelerators. The escalating tensions have also led to significant financial restructuring, with a 2025 U.S. government policy imposing a 15% revenue fee on AI chip sales by NVIDIA and AMD in China. This "tax" is projected to reduce gross margins by 8–10 percentage points for NVIDIA and 5–7 percentage points for AMD, fundamentally recalibrating investor sentiment. Consequently, NVIDIA's price-to-earnings (P/E) ratio has fallen from 50x to 42x, and AMD's from 60x to 52x, reflecting the pricing-in of geopolitical risk into semiconductor valuations. NVIDIA has proactively reduced its exposure to the Chinese market, with management guiding for minimal near-term sales. Expert Commentary Analysts largely view China's recent regulatory actions as strategic leverage within ongoing trade discussions, rather than an immediate existential threat to U.S. chipmakers. > "China's regulatory moves are viewed as strategic leverage in ongoing trade discussions rather than an immediate existential threat to U.S. chipmakers." This perspective underscores the complex interplay between geopolitics and market dynamics in the semiconductor industry. Looking Ahead The outcome of the Madrid trade talks, which are addressing critical issues such as tariffs, export limits, and technology transfers, will be crucial in determining the near-term trajectory for the semiconductor industry. While a complete breakthrough is deemed unlikely by some experts, the negotiations may provide clarity or further exacerbate the current deadlock. The ongoing investigations create a "headline overhang" that could complicate product approvals and market access for U.S. companies operating in China. Furthermore, these tensions are likely to accelerate strategic shifts, prompting multinational corporations to diversify their procurement away from China and encouraging Chinese customers to favor local vendors, thereby reshaping global supply chains in the long run.
Sep 16 2025, 19:43
Navitas Semiconductor Stock Gains on Nvidia Partnership, Faces Near-Term Revenue Challenges
Navitas Semiconductor (NVTS) has seen a significant stock increase following a partnership with Nvidia for AI data center power solutions. Despite this long-term growth catalyst, the company faces projected revenue declines in 2025 and its stock valuation appears elevated relative to near-term financial performance. Technology Sector Focus Shifts as Navitas Semiconductor Stock Surges Navitas Semiconductor (NASDAQ: NVTS) has experienced a notable ascent in its stock valuation, rising to approximately $6 per share from an all-time low of $1.52 in April. This substantial gain is largely attributed to a new strategic partnership with Nvidia (NASDAQ: NVDA), aimed at developing advanced power-efficient delivery systems for AI data centers. The Nvidia Collaboration: A Catalyst for Market Optimism In May, Navitas announced its collaboration with Nvidia, which propelled its stock upward by an estimated 164% to 175%. The partnership positions Navitas as a critical supplier of its proprietary gallium nitride (GaN) and silicon carbide (SiC) power chips for Nvidia's next-generation 800V high-voltage direct current (HVDC) architecture. These advanced chips are designed to enhance power efficiency by approximately 5%, reduce data center maintenance costs by about 70%, and decrease copper wiring needs by around 45% in future AI infrastructure, including Kyber rack-scale systems and Rubin Ultra server racks. Navitas specializes in these cutting-edge power semiconductors, which offer superior speed, power efficiency, and temperature resistance compared to traditional silicon-based alternatives. Beyond AI data centers, these technologies find applications in electric vehicle (EV) chargers, mobile device chargers, industrial motors, and solar inverters. The company expanded its SiC business in 2022 through the acquisition of GeneSiC, further solidifying its position in the EV and data center markets. Analysis of Market Reaction and Valuation Concerns The market's enthusiastic response to the Nvidia partnership underscores the perceived long-term potential for Navitas in the rapidly expanding AI data center market. However, a closer examination reveals a significant disconnect between the stock's current valuation and Navitas's near-term financial trajectory. Analysts project a 42% decline in Navitas's revenue for 2025, expecting sales to fall to $48.6 million, with the company's net loss widening to $116.4 million. Crucially, mass production of Navitas's power chips for Nvidia is not anticipated until 2027. This timeline suggests that the current stock price, trading at approximately 24 times this year's sales with a market capitalization of $1.2 billion, may be significantly inflated. The price-to-sales (P/S) ratio for Navitas is considerably higher than the industry average of 8.6x, and starkly contrasts with competitors such as Wolfspeed (0.47x) and Power Integrations (6.34x). Broader Context, Financials, and Future Projections Navitas has not achieved profitability on a generally accepted accounting principles (GAAP) basis since 2022, and its bottom line is expected to remain in the red for the foreseeable future. For the second quarter of 2025, the company reported revenue of $14.5 million, a decrease from $20.5 million in the same period of 2024. GAAP operating loss stood at $21.7 million, with a non-GAAP operating loss of $10.6 million. Guidance for the third quarter of 2025 projects net revenues of $10.0 million, plus or minus $0.5 million, reflecting ongoing headwinds. The anticipated revenue decline in 2025 is largely attributed to macroeconomic challenges impacting Navitas's core EV, solar, and industrial customers, leading to inventory adjustments. Seasonal softness in the mobile and consumer markets, alongside risks associated with China tariffs on SiC products, further exacerbate these challenges. The company has strategically decided to deprioritize lower-margin China mobile business to sharpen its focus on AI data centers and energy infrastructure. Despite near-term hurdles, the long-term outlook for GaN and SiC technologies in AI data centers is substantial. Navitas estimates that these technologies could support a 100x increase in server rack power capacity and unlock a $2.6 billion market potential by 2030. Analyst projections indicate a 9% revenue increase in 2026, followed by a significant 79% surge to $95 million in 2027, coinciding with the expected start of mass production for Nvidia. Expert Commentary Underscores Speculative Nature Expert analysis frequently highlights the speculative nature of Navitas's current valuation. Many analysts suggest that the stock's current price-to-sales ratio is inflated by the Nvidia partnership, and the company may struggle to grow into this premium valuation. Some even contend that the stock could be halved and still be considered expensive relative to its peers. For investors seeking direct exposure to the AI market, other companies like CoreWeave are often cited as potentially more favorable options due to stronger growth and lower valuations. > "The company's current valuation, at 24 times this year's sales, is considered expensive, with the Nvidia partnership potentially inflating its price-to-sales ratio." Looking Ahead: A Focus on Execution and Market Adoption The future trajectory of Navitas Semiconductor is heavily contingent upon the successful execution of its Nvidia partnership and the widespread adoption of Nvidia's new AI data center architecture. Key milestones include the shipment of initial power chip samples in the fourth quarter of 2025, final selections in 2026, and the crucial commencement of mass production in 2027. The company aims to achieve EBITDA breakeven once quarterly revenues reach the high $30 million threshold, with some analysts anticipating positive adjusted EBITDA in 2026. Navitas holds a strong balance sheet with $161.2 million in cash and cash equivalents as of June 30, 2025, and no debt, providing some financial flexibility during this transitional period. However, navigating ongoing macroeconomic headwinds in its traditional markets, while simultaneously investing in and developing its AI data center and energy infrastructure segments, will be critical to realizing its ambitious long-term growth potential.