Yazılar

Marvell Postpones Investor Day, Narrows Revenue Forecast Amid Trade-Driven Economic Uncertainty

/in /tarafından

Marvell Technology, a major player in networking and custom AI chips, announced Tuesday it is postponing its upcoming investor day due to what it called a “dynamic macroeconomic environment,” citing ongoing global trade tensions and economic uncertainty. The decision spooked investors, sending Marvell shares down more than 6% in after-hours trading.

The company also narrowed its Q1 fiscal 2026 revenue guidance, now expecting approximately $1.875 billion, within a tighter range of ±2%, compared to its prior forecast of ±5%. The midpoint of the outlook remains unchanged.

The announcement comes as semiconductor and computing firms navigate turbulent waters driven by shifting U.S. trade policy under President Donald Trump. Although Trump paused a sweeping new import tariff plan for 90 days starting April 9 to allow negotiations, a baseline 10% tariff and additional duties on key partners remain in place, impacting global supply chains and corporate planning.

Marvell’s COO Chris Koopmans previously stated that tariffs had not yet affected the company’s data center segment, but broader concerns linger industry-wide. Nvidia recently warned of a $5.5 billion impact due to U.S. export restrictions on AI chips bound for China, while ASML raised caution over its future sales outlook.

The postponement of Marvell’s investor day suggests the company may be waiting for greater clarity on trade policies and economic stability before providing long-term strategic updates to shareholders.

Super Micro Shares Fall After Forecasting Q4 Revenue Below Estimates Amid Tariff, Spending Concerns

/in /tarafından

Super Micro Computer (SMCI.O), a leading AI server manufacturer, projected fourth-quarter revenue below Wall Street expectations, causing its shares to drop 5.4% in after-hours trading on Tuesday. The company cited economic uncertainty, tariffs, and delayed customer spending as near-term headwinds.

The San Jose-based firm forecast Q4 revenue between $5.6 billion and $6.4 billion, falling short of analysts’ average estimate of $6.82 billion, according to LSEG data. The company has benefited from surging demand for AI data center infrastructure, leveraging chips from Nvidia, AMD, and others, but has also faced accounting issues in recent months that sparked delisting concerns on the Nasdaq.

Despite some clients delaying purchases, Super Micro expects those deferred deals to materialize in the June–September quarter. However, investor sentiment remains cautious, particularly in light of growing concerns about AI investment slowdowns and tariff-related impacts.

Kim Caughey Forrest of Bokeh Capital Partners suggested the lowered guidance might be self-inflicted, rather than purely market-driven, while D.A. Davidson’s Gil Luria noted the possibility that Super Micro may be losing market share to competitors like Dell, rather than signaling a broader downturn in AI infrastructure demand.

For fiscal year 2025, Super Micro revised its revenue forecast downward to $21.8 billion to $22.6 billion, from a previously expected $23.5 billion to $25.0 billion.

The company had released preliminary results last week, but the lower guidance and uncertain macroeconomic environment continue to weigh on investor confidence.

Huawei Preparing to Ship New AI Chip as China Seeks Alternatives to Nvidia Solutions

/in /tarafından

Huawei Technologies is set to begin mass shipments of its new 910C artificial intelligence chip to Chinese customers as early as next month, according to sources familiar with the matter. These shipments come at a crucial time, as China faces increasing challenges in securing domestic alternatives to Nvidia’s AI chips, which have been restricted due to escalating tensions between the U.S. and China. Some shipments of the Huawei 910C have already been made, with many Chinese AI companies eagerly awaiting a local solution to meet their growing demand for high-performance AI hardware.

The timing of the release is significant, as Chinese AI firms have been scrambling to find alternatives to Nvidia’s H20 chip, which had been widely used in AI development. Recently, the U.S. government announced that sales of the H20 to China would now require an export license, placing additional strain on Chinese tech companies that rely heavily on Nvidia’s advanced GPUs for AI research and deployment. With the Huawei 910C, China is looking to reduce its dependency on foreign technology, particularly in the critical area of AI chip development.

The Huawei 910C, which is a graphics processing unit (GPU), represents an evolution of the company’s previous offerings rather than a revolutionary breakthrough. The 910C combines two 910B processors into a single package using advanced integration techniques, delivering performance that rivals Nvidia’s H100 chip. This architectural design allows Huawei to provide a competitive product without entirely reinventing the wheel, making it an appealing alternative for AI applications in China. While the company has yet to publicly confirm the details of the chip’s capabilities or its shipment schedule, the timing aligns with the urgent need for domestic alternatives to Nvidia’s technology.

The geopolitical context behind the 910C’s development is important, as the U.S. has been restricting the sale of its most advanced AI products to China, citing national security concerns. In addition to the H20 chip, China has also been cut off from Nvidia’s flagship B200 chip, further intensifying the need for local solutions. As Huawei ramps up its efforts to ship the 910C, it is positioning itself as a key player in China’s push to maintain technological independence in the face of foreign restrictions.