Summary: On July 30, 2026, NVIDIA officially launched its next-generation AI accelerator chip B300 at its annual developer conference in Los Angeles, with performance nearly 3 times better than its predecessor, driving a collective surge in the US AI semiconductor sector. This article provides an in-depth analysis of the impact of this event on the industrial chain and identifies key investment targets and ETFs to help investors seize opportunities in the US AI computing power sector.
On July 30, 2026 (US Eastern Time), NVIDIA officially launched its next-generation AI accelerator chip—the B300 at its annual developer conference in Los Angeles. Dubbed a "monster-level" chip by the industry, the B300 uses TSMC's 3nm process and integrates over 250 billion transistors. Its AI inference performance is nearly 3 times better than the previous generation H200, while training performance improves by about 1.8 times. Following the announcement, NVIDIA's stock price surged over 5% in after-hours trading, driving the entire US AI semiconductor sector higher.
Event Core: B300 Chip Performance Leap, Cementing AI Computing Power Dominance
According to presentations at the launch event, the B300's highlight is its innovative "hybrid tensor core" architecture, which efficiently handles FP8, FP16, and MXFP4 precision calculations on the same chip. Paired with the new NVLink 5.0 interconnect technology, multi-card cluster communication bandwidth increases to 3.2TB per second, reducing large-scale model training time by about 40%. NVIDIA CEO Jensen Huang stated in his keynote speech: "The B300 is not just a performance upgrade, it redefines the limits of AI computing."
More exciting for the market is NVIDIA's simultaneous release of the B300's "superchip" version—the GB300 Grace Blackwell Ultra. This combination integrates the B300 GPU with the latest Grace CPU through ultra-high bandwidth packaging, providing an all-in-one computing solution for hyperscale data centers. Cloud service providers, AI startups, and research institutions can use this solution to quickly build AI training clusters ranging from thousands to even hundreds of thousands of cards.
Market Reaction: Industrial Chain Stocks Rally, Capital Flows into Semiconductor ETFs
On the day of the announcement, the US AI chip sector reacted immediately. Besides NVIDIA, TSMC ADR rose 3.2% to a 52-week high; Amkor Technology, which provides advanced packaging for chips, surged 6.8%; while competitor AMD's stock also rose 2.1%, mainly boosted by the sentiment across the entire AI computing power sector. The Philadelphia Semiconductor Index (SOX) overall rose 2.7%, closing at its second-highest level in history.
In terms of capital flows, exchange-traded funds (ETFs) tracking AI and semiconductor themes saw significant net inflows. According to ETF.com, on July 30 alone, the largest AI ETF—Global X Robotics & Artificial Intelligence ETF (BOTZ)—had net inflows of approximately $3.5 billion, while VanEck Semiconductor ETF (SMH) had net inflows exceeding $6 billion. This indicates strong investor confidence in the long-term growth prospects of the AI computing power industrial chain.
Industry Analysis: Surging AI Computing Power Demand, New Chip Opens Growth Ceiling
Industry analysis generally believes that the B300 launch is not just an update to NVIDIA's product line, but a "reassurance" for the entire AI infrastructure investment wave. Currently, global tech giants are accelerating AI data center construction, with demand for high-performance GPUs growing exponentially. According to the latest forecast from market research firm IDC, the global AI server market size will exceed $180 billion by 2028, with a compound annual growth rate of 29%. With the B300, NVIDIA has once again widened the technology gap with competitors, potentially maintaining over 80% market share in the next two years.
Meanwhile, the B300's high energy consumption also brings huge incremental opportunities in areas such as liquid cooling, advanced packaging, and high-bandwidth memory (HBM). Morgan Stanley analysts pointed out that the promotion of NVIDIA's B300 will increase the value per GPU server to over $500,000, with upstream power management, optical modules, and PCB manufacturers all set to benefit significantly.
Investment Opportunities: Focus on Core Leaders and Industrial Chain ETFs
For ordinary US stock investors, there are several approaches to positioning in the AI computing power sector:
- Direct allocation to leading stocks: NVIDIA (NVDA) as the king of chip design remains the top choice, but its valuation is already high, so look for pullback buying opportunities. Additionally, Broadcom (AVGO) has unique advantages in custom AI chips and network chips, making it worth attention.
- Upstream and downstream suppliers: Including SK Hynix (ADR: HXSCL) providing HBM memory for NVIDIA, TSMC for advanced packaging, and liquid cooling solution provider Vertiv Holdings (VRT), among others.
- Index funds and ETFs: If you don't want to bear the volatility of individual stocks, you can allocate to SMH (Semiconductor ETF), SOXX (iShares Semiconductor ETF) or BOTZ (Robotics & AI ETF) to diversify risks and share the overall growth benefits of the sector.
Risk Warning: High Volatility and Valuation Pressure Coexist
Although the AI computing power sector has long-term positive prospects, several risks need attention in the short term: First, valuations of leading stocks like NVIDIA are at historical highs, and subsequent earnings reports that miss expectations could lead to significant corrections; Second, AI chip competition is intensifying, with AMD, Google's custom TPUs, and others actively catching up; Third, a global economic downturn may cause tech companies to cut capital expenditures, affecting demand. Investors should adopt methods like staggered position building or dollar-cost averaging based on their risk tolerance.
In conclusion, NVIDIA's B300 launch sends a clear signal to the market of sustained high-speed growth in AI computing power. For investors focused on the US stock market, the AI industrial chain remains an important theme not to be missed in the coming years.
