Nvidia Pours $1.5 Billion Into US Chip Packaging as Vera Rubin Ships to Hyperscalers
Published on 07/24/2026 at 04:52 | Redaktion boerse-global.de
Nvidia is executing a two-pronged strategy to cement its dominance in artificial intelligence infrastructure: pouring capital into domestic manufacturing capacity while simultaneously shipping its next-generation Vera Rubin platform to major cloud customers. The moves come as the company navigates a delicate balance between surging demand and potential supply-chain bottlenecks.
The chipmaker signed a $1.5 billion multi-year agreement with Amkor Technology on July 23, 2026, to expand chip-packaging operations in the United States. Nvidia is making an upfront payment to help Amkor scale its facilities, including operations in Arizona, with the two companies jointly developing multi-die packaging technologies for AI platforms. Amkor already provides packaging services for Nvidia’s data center processors. The deal sent Amkor shares surging more than 13% in after-hours trading to $74.00, recovering from a regular-session decline of 2.45% that had left the stock at $65.33. Amkor is scheduled to report quarterly results on July 27, with analysts forecasting revenue of roughly $1.80 billion.
Just two days earlier, on July 21, Taiwanese contract manufacturer Wistron opened its first US factory in Fort Worth, Texas. The 324,000-square-foot facility, built with a $700 million investment, is already producing Nvidia’s GB300 Grace-Blackwell-Ultra superchip, with plans to manufacture the Vera Rubin chip there as well. More than 500 jobs have been created so far, with the workforce expected to reach 1,000 by year-end. Nvidia CEO Jensen Huang attended the opening alongside Wistron Chairman Simon Lin, calling the project a key contribution to American reindustrialization. Huang also used the occasion to assert that Taiwan had saved the US semiconductor industry, countering previous political claims that it had stolen it. Nvidia has said it plans to invest up to $500 billion in US-based AI platform manufacturing over the next four years, including expansions at TSMC’s Arizona facilities.
The Vera Rubin Rollout Begins
Nvidia finalized mass production of its Vera Rubin platform in May 2026 and began shipping first rack systems to major North American cloud providers — including Microsoft, Google, and Oracle — in late July. The architecture marks a significant shift away from the Blackwell generation toward an integrated “AI factory” model combining a new Vera CPU with a Rubin GPU. Early testing by partners such as CoreWeave shows the NVL72 platform delivering up to ten times the token throughput per megawatt compared with the previous Grace-Blackwell generation — a critical selling point for hyperscalers facing power constraints in their data centers.
Should investors sell immediately? Or is it worth buying Nvidia?
Nvidia is no longer selling individual chips but complete racks priced at $7 million to $8 million per unit. The company has already shipped “hundreds of thousands” of Grace standalone servers and more than 2.5 million Grace CPUs, including to Meta. The new Vera CPU features 88 custom Arm cores, and Nvidia expects to produce up to 1,000 racks per day. First systems have gone to Anthropic, OpenAI, Oracle, and SpaceX. Nvidia anticipates roughly $20 billion in revenue from the Vera business alone in the current fiscal year.
Memory Bottleneck Looms
The most significant risk to Nvidia’s growth trajectory lies in the supply chain for HBM4 (High Bandwidth Memory). Nvidia has secured about 70% of its HBM4 needs from SK Hynix, but the broader industry is warning of a potential “memory crisis” in the second half of 2026. The transition to HBM4 requires complex 16-layer stacks, and yields at key suppliers Samsung and Micron have come in lower than expected. If the bottlenecks persist, Nvidia may struggle to fulfill delivery commitments for large-scale deployments by OpenAI and other AI labs in the third quarter.
Adding to the pressure, TSMC has raised fabrication prices for its 3-nanometer and 4-nanometer nodes for the 2026 fiscal year, a direct consequence of the AI arms race for capacity. Rising input costs are hitting Nvidia just as its 30-day volatility stands at 34.75%. Critics point to what they call a “geopolitical risk discount,” with macro uncertainty and energy policy tensions keeping the stock closer to its 50-day moving average of $181.07 than to its recent highs.
Stock Under Pressure Despite Fundamentals
Despite the positive manufacturing news, Nvidia shares have come under pressure amid a broader sector rotation. After Alphabet raised its 2026 investment forecast to between $195 billion and $205 billion, investors rotated into memory-chip stocks such as Micron and SK Hynix. The stock closed Thursday at $183.60, down 1.38% from the prior session. That leaves the shares 9.33% below their 52-week high of $202.50, reached in May, though they remain up 14.55% year-to-date.
The operational picture remains robust. In the first quarter of fiscal 2027, Nvidia generated revenue of $81.6 billion, with $75.2 billion coming from the data center segment. Earnings per share of $1.87 beat the consensus estimate of $1.76. The company guided for second-quarter revenue of roughly $91 billion, excluding Chinese revenue. Analysts have set an average price target of about $304, with most rating the stock a buy.
Nvidia at a turning point? This analysis reveals what investors need to know now.
Competition is intensifying. AMD has launched its Helios platform with the MI455X graphics processor, claiming higher bandwidth and more memory than Nvidia’s Vera Rubin NVL72 — a sign that the battle for AI infrastructure is heating up. Nvidia’s dominant roughly 90% market share in data center chips faces challenges from customers like Google and Amazon, both of which are developing their own silicon.
What’s Next
OpenAI is scheduled to begin large-scale deployments on the Vera Rubin platform in the third quarter of 2026. If the promised tenfold throughput gains are confirmed in production environments, it could trigger a revaluation of Nvidia’s growth story. Conversely, if HBM4 supply constraints lead to delivery shortfalls, the stock could test its 200-day moving average of $166.01, particularly if market volatility remains elevated. The average analyst price target of $265.89 implies upside of 44.4% from current levels, but achieving that depends on Nvidia maintaining its market position while navigating the most complex supply-chain transition in its history.
Ad
Nvidia Stock: New Analysis - 24 July
Fresh Nvidia information released. What's the impact for investors? Our latest independent report examines recent figures and market trends.
Disclaimer regarding our articles: No investment advice, no buy or sell recommendation. Information on prices, companies, and markets is provided without guarantee; changes are possible at any time. Stock market transactions can lead to substantial losses. Our articles are created and reviewed in whole or in part automatically with the support of AI.
