Welcome to the AGI era.
That's how OpenAI's Greg Brockman closed the GPT-6 Astra launch. The headlines: → Trained on 100,000+ GPUs at Stargate, Texas → Claims the lead in coding, science, reasoning and cybersecurity → Works inside software: spreadsheets, 3D worlds, circuit boards, tax forms → First model to hit OpenAI's "Critical" cyber threshold — access restricted via Daybreak → Priced at 2.5x GPT-5.6 Why this matters for AI hardware & memory chips: 1) 100k GPUs for one model = the compute race is accelerating, not plateauing. 2) Agentic models that build, reason and act burn far more inference tokens. Inference = sustained demand for HBM and DRAM, not one-off training spend. 3) 2.5x pricing shows customers will pay for capability → hyperscaler capex stays funded. The picks-and-shovels trade remains alive and well.
$NVDA now trades under 10x GS’s updated 2028 EPS estimates, per GS’s Peter Callahan
Chart below is the relative performance of Nvidia relative to the broader semiconductors SOX index. Source: Bloomberg, GS, Negligible capital on X
Nvidia Expands Into AI Infrastructure Finance
Nvidia is reportedly forming a major AI infrastructure investment partnership with some of the world’s largest private-capital firms, including Apollo, Blackstone, BlackRock, Brookfield, Goldman Sachs and KKR. The objective: mobilise huge pools of private capital to finance the chips, data centres and power infrastructure required for the AI build-out. The scale is enormous: Morgan Stanley estimates hyperscalers will spend $3.5 trillion between 2026 and 2028, while Apollo believes total AI infrastructure investment could eventually exceed $8 trillion. Nvidia is moving beyond selling GPUs. It is increasingly helping customers finance the infrastructure needed to buy and deploy its chips—including providing financial backing and potentially guarantees. Nvidia is separately discussing a massive guarantee for a 10GW Ohio data-centre project leased to OpenAI, highlighting how deeply it could become involved in financing the AI ecosystem. Private capital is becoming critical. AI investment requirements are now too large to be financed through corporate cash flows alone, pushing tech companies toward private credit, bonds, securitisation, project finance and equity. The risk: Nvidia financing or guaranteeing customers that ultimately purchase Nvidia chips creates concerns around “circular financing” and increasingly concentrated financial exposure across the AI ecosystem. Bottom line: Nvidia is evolving from the AI boom’s dominant chip supplier into something closer to its financial architect. Partnering with private-capital giants could unlock trillions for AI infrastructure—but also further intertwines Nvidia’s financial fortunes with the customers driving its extraordinary revenue growth. Source: FT
In case you missed it
Apple $AAPL surpasses Nvidia $NVDA as the World's Most Valuable Company ($4.92 Trillion vs. $4.86 Trillion in early hours trading) Source: Barchart
Trump's updated portfolio
heavy on semiconductors (10% NVIDIA, 4.5% Qualcomm, 4% Micron, and 5% Intel) Source: Rand Group
Nvidia is planning to sell at least $20bn of investment-grade debt in the US its first bond sale in more than five years in a test of investor appetite for further exposure to the AI sector.
Source: FT
Nvidia
Source: Investing.com
Nvidia will launch a PC “superchip” this year as the semiconductor giant goes head-to-head with Apple, Qualcomm, Intel and AMD for the first time.
Computer makers including Dell, Asus and HP will use what Nvidia claims is “the most efficient PC chip ever built”, paired with Microsoft’s Windows software, Nvidia chief executive Jensen Huang announced on Monday. The launch marks a major competitive shift in the consumer PC industry and a new business line for $5.1tn Nvidia. The company, best known for its dominance of semiconductors for AI infrastructure, is pushing beyond its traditional graphics cards into integrated chips that power the whole PC. Lenovo, Microsoft, Acer and Taiwan’s MSI will also use the product. Source: FT
