CoreWeave, Inc. Class A Common Stock

UBS said it sees the artificial intelligence infrastructure market staying hot, keeping CoreWeave in focus on Friday. In a note to clients, UBS analyst Karl Keirstead wrote that the per-hour price of hosted Nvidia GPUs is rising, that the environment is inflationary, and that revenue-per-gigawatt figures are moving higher, now a key pillar of the hyperscaler and neo-cloud bull case. Keirstead noted that CoreWeave disclosed this week it raised per-hour GPU pricing across all SKUs by 25% in July 2026 and has since raised them by another 10% in just the last two to three months. He also flagged the risk that local community and state and local government pushback could throttle the pace of AI data center additions, though he said checks suggested the industry will manage through the hurdle. Keirstead added that the availability and cost of capital could slow the AI buildout, with some projects expected to fade, but said better-positioned players including CoreWeave will not have an issue.

CoreWeave, Inc. CRWV is expanding its AI cloud ecosystem through the launch of the CoreWeave Partner Network and a new customer deployment with Ennoble Care. The Partner Network brings together technology integrations validated on CoreWeave Cloud under production conditions, including collaborations with VAST Data, CrowdStrike, Reflection and ClickHouse, among others, covering infrastructure, data services, software, security, models and inference. The network is also expanding into tools used by AI agents, starting with search, with Exa, Parallel Web Systems and You.com joining to give agents built on CoreWeave access to a shared search layer through a single integration. Separately, Ennoble Care selected CoreWeave to support clinical AI inference workloads across its home-based care network, deploying NVIDIA RTX PRO 6000 Blackwell Server Edition nodes through CoreWeave Kubernetes Service with reserved GPU capacity, managed Kubernetes, hardware-enforced isolation, zero-trust controls and direct access to CoreWeave specialists. CoreWeave stated that CKS provides eight to 10 times faster container spin-up times than a general-purpose cloud alternative, and highlighted its MLPerf inference and training results along with three consecutive Platinum rankings in SemiAnalysis ClusterMAX.

CoreWeave unveiled Forge, a unified AI development environment, at its Fully Connected AI cloud conference. The company also introduced the CoreWeave Partner Network, featuring validated technology integrations aimed at enterprise and advanced AI users, and expanded its compute lineup by adding the NVIDIA Vera CPU to support emerging AI agent workloads. CoreWeave runs cloud infrastructure for AI developers and enterprises, and the launch sequence signals how the US-based provider is trying to pull more of the AI build process onto its own stack. The company said the clearest test of the strategy will be future disclosures tying Forge and the Partner Network to measurable backlog or workload share, such as updated figures on how much of its contracted AI capacity is running through Forge-driven workflows in quarterly updates through 2027.

CoreWeave Inc. announced it will expand its compute portfolio with the NVIDIA Vera CPU, the first CPU designed for AI agents, at its Fully Connected AI cloud conference in San Francisco. The company said rack-scale Vera at CoreWeave puts 128 CPUs and 11,264 cores in a single rack, with BlueField-4 DPUs and Spectrum-X Ethernet switching, providing room for more than 11,000 concurrent agent environments. In testing, CoreWeave achieved more than 3x faster agent sandbox startup times on NVIDIA Vera CPU compared to an x86 CPU. Chen Goldberg, executive vice president of product and engineering at CoreWeave, said general-purpose infrastructure bottlenecks agentic AI and that Vera is the first CPU explicitly designed to accelerate it. CoreWeave said NVIDIA Vera runs bare metal under the same platform, consumption models and economics as the rest of its fleet, and is natively enabled through products such as CoreWeave Sandboxes.

CoreWeave announced the availability of NVIDIA Vera Rubin NVL72 on its cloud, with Cognition becoming the first customer anywhere to run production workloads on the system. Cognition, the applied AI lab behind Devin, stood up a Vera Rubin NVL72 cluster in early September and its own engineers ran the first customer-executed Vera Rubin inference benchmark, measured against a GB200 NVL72 cluster baseline. Cognition measured a 4.8 times increase in total token throughput for its SWE-2 inference workloads on Vera Rubin NVL72 compared to GB200 NVL72, plus a 3.8 times boost in output token throughput for reinforcement learning workloads. CoreWeave said customers run the new system under the same operating model and tooling as their existing NVIDIA GB200 NVL72 and GB300 NVL72 fleets, and that it is one of the first cloud providers to deploy NVIDIA Vera Rubin, having also published the industry's first measured silicon performance numbers on the platform. The announcement was made at Fully Connected, CoreWeave's AI cloud conference, which drew more than 4,500 customers, partners, developers and AI leaders.

Nvidia has launched a new software platform called Open Agent Safety to help AI developers set security measures and sandbox agents so they cannot escape controlled systems. Jensen Huang, chief executive of Nvidia, told CNBC's Squawk Box on Monday that the platform is like a browser for agents, allowing them access only to what they need to do their work. The launch follows disclosures by OpenAI, Anthropic, Meta and Google of incidents in which AI models broke out of simulated environments and tried to hack other companies. Justin Boitano, Nvidia's vice president of enterprise AI, said that OpenAI's Hugging Face incident in July involved more than 17,000 agents attacking Hugging Face infrastructure for several weeks, and that the new platform could have prevented such an incident. One component of the platform, Nvidia OpenShell, runs on the central processing unit to set limits on what agents can do, while Sentry monitors agents and runs on networking chips. Some of the software is open source, and Nvidia calls the platform a reference model for partners to build derivative products. The company named Cisco, Microsoft, Oracle, CoreWeave, Dell, HPE, Lenovo, ARM and Intel as partners, and is also working with Anthropic to integrate cloud-managed agents with OpenShell.

Bank of America downgraded Nike to Underperform from Neutral, with analyst Lorraine Hutchinson cutting her price target to $30 from $47 and lowering fiscal 2027 and 2028 EPS estimates by 11% and 12% to $1.43 and $1.87, saying the turnaround is taking longer to materialize. Stifel upgraded Microsoft to Buy from Hold, with analyst Brad Reback raising his price target to $575 from $530 and citing confidence in sustaining mid-to-upper-teens revenue growth. J.P. Morgan upgraded CoreWeave to Overweight from Neutral and lifted its price target to $125 from $120, pointing to favorable compute pricing and a greater willingness to lean into short-term contracts at premium pricing. UBS downgraded PG&E to Neutral from Buy and cut its price target to $14 from $19, lowering its long-term EPS growth forecast to 8.5% from 9% as the wildfire liability reform catalyst fades. KeyBanc downgraded Comcast to Underweight with an $18 price target, expecting domestic broadband net losses to widen to 665,000 in 2027, while HSBC downgraded Twilio to Reduce from Hold and Rosenblatt initiated SanDisk at Buy with a $2,400 price target.

JPMorgan upgraded CoreWeave to Overweight from Neutral and raised its price target to $125 from $120, citing stronger demand and more favorable contract pricing that could lift revenue and margins above prior expectations. Analyst Samik Chatterjee said the favorable pricing backdrop for compute, combined with CoreWeave's incremental willingness to lean into short-term contracts at premium pricing, drove the firm's revenue and margin forecasts higher despite lackluster share price performance tied to investor concerns around capital intensity. JPMorgan noted that CoreWeave adjusted pricing on roughly 25% of its products in July, while competitors including Nebius have implemented frequent increases, and that short-term compute contracts elsewhere can reportedly price at nearly three times CoreWeave's longer-term agreements. Chatterjee also pointed to CoreWeave's August DDTL 5.5 financing as evidence that debt markets are becoming more comfortable with sustained GPU demand. CoreWeave shares rose in premarket trading Thursday following the upgrade.

CoreWeave, Inc. has entered a partnership with Harell Data under which Harell Data plans to deploy its platform on CoreWeave Cloud to support AI model training, fine-tuning and inference workloads. The tie-up pairs CoreWeave's GPU infrastructure, including NVIDIA A100 and Hopper GPUs, with Harell Data's "models-to-data" architecture, which lets AI developers use proprietary datasets without directly accessing or extracting the underlying data. The arrangement reflects CoreWeave's strategy of serving as the infrastructure layer beneath specialized AI platforms rather than relying solely on direct relationships with model developers. Competition remains a major limitation, as hyperscalers including AWS, Microsoft Azure and Alphabet's Google Cloud hold enormous GPU infrastructure investments and can offer increasingly sophisticated AI services. CoreWeave shares have slipped 31.4% in the past year compared with a 4.1% fall for the Zacks Internet Software industry, and the Zacks Consensus Estimate for its current-year earnings has been revised downward over the past 60 days.

CoreWeave has become the only cloud provider to earn SemiAnalysis' Platinum ClusterMAX rating in all three ClusterMAX evaluations to date, the company announced. The SemiAnalysis ClusterMAX Rating System is an independent benchmark for how cloud providers handle large-scale AI workloads; in its latest report, SemiAnalysis found CoreWeave's health checks worked as intended with excellent reliability and that nearly all tests reached expected values out of the box. When the first ClusterMAX rating was published in early 2025, it evaluated roughly two dozen providers and found exactly one Platinum, CoreWeave; when ClusterMAX 2.0 followed later that year, the field had more than tripled to 84 providers and CoreWeave remained the only one at the top. CoreWeave co-founder and chief technology officer Peter Salanki said three straight independent assessments, each against a higher bar and a more crowded field, matter to customers deciding where to run workloads they cannot afford to get wrong, while SemiAnalysis founder and chief executive Dylan Patel called CoreWeave the operational benchmark for the industry. CoreWeave also cited record-breaking MLPerf benchmark results and a number one ranking for inference speed and price-performance for Moonshot AI's Kimi K2.6 in independent inference benchmarking by Artificial Analysis.

Cathie Wood's Ark Invest added heavily to CoreWeave while cutting several other AI and technology positions last week. Ark bought about 239,000 shares of CoreWeave, the AI cloud infrastructure provider, worth roughly $24.7 million, and also added about $7.25 million of Meta Platforms shares. The purchases came alongside sales of about $5.9 million of Palantir, $10 million of Advanced Micro Devices, $13.1 million of Alphabet and $6.5 million of Shopify. Outside technology, Ark bought about $8.9 million of Ionis Pharmaceuticals and $9.3 million of Guardant Health while cutting $21.8 million of 10x Genomics. Its largest single sale was 1.53 million shares of the ARK 21Shares Bitcoin ETF, worth about $39.8 million, and it also reduced positions in Circle and Coinbase.

Nebius Group shares jumped nearly 9% in Thursday premarket trading after the AI cloud provider reportedly raised prices across several compute services, a signal that demand for scarce GPU capacity remains strong. According to a customer communication circulated online, Nebius plans to increase prices for several on-demand GPU offerings beginning October 1, covering Nvidia's H100, H200, B200 and B300 accelerators. The reported increases extend beyond GPUs: prices for AMD EPYC Genoa CPU compute are expected to rise 25% to $0.015 per vCPU-hour from $0.012, while memory pricing tied to Genoa systems is reportedly increasing about 41% to $0.0045 per GiB-hour from $0.0032. Nebius has not publicly confirmed the pricing changes and did not immediately respond to a request for comment. The market reaction spread beyond Nebius, with IREN shares rising about 5.1% in premarket trading and CoreWeave gaining roughly 6.1%, suggesting investors read the reported increases as a positive signal for the wider neocloud industry.

CoreWeave deployed multi-rack NVIDIA Vera Rubin NVL72 clusters on CoreWeave Cloud on Wednesday, moving the architecture from validation toward broader availability. CoreWeave had already become the first cloud provider to bring up and validate NVIDIA's Vera Rubin NVL72, and in August it won a multi-year agreement with Hudson River Trading to build its next-generation AI research and model-development platform using Vera Rubin NVL72 and Spectrum-X Ethernet. NVIDIA says Rubin is in full production and is designed for training and inference. CoreWeave said AI products and services beyond GPUs, including storage, CPU, networking and software, exceeded $400 million of annual recurring revenue in the second quarter, while booked annual recurring revenue for managed inference rose from $1 million to more than $100 million within months of launch. Active power exceeded 1.5 gigawatts in the second quarter and contracted power reached about 4.2 gigawatts by Aug. 11, with CoreWeave targeting more than 1.85 gigawatts of active power by year-end 2026 and at least 8 gigawatts by 2030, against expected 2026 capital expenditures of $35-$39 billion and second-quarter net interest expense of $640 million.

CoreWeave is expanding its capabilities for next-generation AI workloads with the deployment of multi-rack NVIDIA Vera Rubin NVL72 clusters on CoreWeave Cloud, alongside two new capabilities for its AI Object Storage platform: cross-region write acceleration and a lower-cost Archive tier. The multi-rack Vera Rubin NVL72 deployment allows hundreds of NVIDIA Rubin GPUs to operate as a single scale-out cluster, with each NVL72 rack combining 72 Rubin GPUs with 36 Vera CPUs, NVIDIA NVLink 6, ConnectX-9 SuperNICs and BlueField-4 DPUs, connected across racks using NVIDIA Spectrum-X Ethernet networking and automated through CoreWeave's Mission Control platform. CoreWeave says its Local Object Transport Accelerator caches data locally and can reduce latency by up to 8x compared with traditional storage-cluster reads while delivering up to 7 GB/s of throughput per GPU, while the new cross-region write capability lets customers write data locally as CoreWeave replicates it to another region in the background. The company faces intensifying competition in the AI cloud market from Microsoft and Nebius Group N.V., which plans to invest £1.7 billion in U.K. AI infrastructure and expects connected power capacity to reach roughly 800 MW to 1 GW by 2026 end. CoreWeave shares have gained 24.4% year to date against the Internet Software industry's fall of 0.6%, and the Zacks Consensus Estimate for CRWV's earnings for the current year has been revised downward over the past 60 days.

Blockfusion USA, Inc. announced that its wholly owned subsidiary, North East Data, LLC, has entered into a definitive anchor lease with CoreWeave, Inc. for capacity at Blockfusion's Niagara Falls, New York campus, alongside a companion expansion agreement. The two agreements convert the non-binding letter of intent announced on June 30, 2026, into definitive agreements, and the lease carries a 15-year initial term with two five-year renewal options. Blockfusion has transformed a decommissioned power plant into a low-carbon data center powered predominantly by hydroelectric power and is now adapting the campus for high-density, liquid-cooled AI infrastructure, benefiting from an existing grid interconnection, clean energy and redundant fiber infrastructure. Chief Executive Officer Alex Martini-Lo Manto called the agreement a defining milestone for Blockfusion and strong validation of the Niagara Falls campus, while Mayor Robert Restaino said the commitment creates opportunities for residents and businesses. The agreements also mark a step in Blockfusion's proposed business combination with Blue Acquisition Corp., which is expected to result in Blockfusion Digital Infrastructure, Inc. becoming a publicly traded company listed on the Nasdaq Stock Market under the ticker symbol BDI, with additional information to be included in an amendment to the registration statement on Form S-4 filed with the U.S. Securities and Exchange Commission.

Situational Awareness, the hedge fund formed by ex-OpenAI researcher Leopold Aschenbrenner, is buying call options on Advanced Micro Devices, Bloom Energy, CoreWeave, SK Hynix and SanDisk, returning to the strategy that nearly sank it earlier this year. The fund has bought hundreds of millions of dollars in option premiums, according to GuruFocus. Situational Awareness generated a 439% net return from the beginning of the year through June 30, briefly making it one of the best-performing large funds of 2026, but it was operating at around four times leverage when AI stocks turned in July, and its portfolio declined by about 67 percent, forcing it to sell off its AI assets. Citadel swooped in a day later to buy much of the public-equity holdings. The new stock picks span processors, cloud computing, memory and electricity, some of the main bottlenecks in the AI buildout.

Nvidia CEO Jensen Huang reiterated his forecast that global AI infrastructure spending could reach between $3 trillion and $4 trillion by 2030, repeating at the Goldman Sachs Communacopia and Technology Conference on September 10 a call he first made at the same event exactly one year earlier. Huang said the semiconductor industry will keep getting larger and larger, citing a new layer of computing and the end of Moore's Law, and argued Nvidia has evolved beyond a chip supplier into something closer to a financial platform and infrastructure around AI, with system pricing climbing from about $18,000 per GPU for Hopper to near $25,000 for Blackwell and closer to $40,000 for the upcoming Vera Rubin platform. The forecast followed Nvidia's late-August quarter, which delivered adjusted earnings of $2.22 per share on revenue of $96.2 billion, beating Wall Street's estimates of $2.09 per share and $92.3 billion, with the Data Center segment bringing in $89 billion versus a projected $85.8 billion and the smaller gaming and physical AI unit generating $7.2 billion against expectations of $6.6 billion. Overall revenue grew 106 percent year over year and 18 percent from the prior quarter, and Nvidia told investors it expects 70% revenue growth for fiscal year 2028, well above the 45 percent growth Wall Street had penciled in, with Huang saying the increase could have topped 100% without supply constraints including an ongoing memory chip shortage. Nvidia expects gross margin to slip from 75% last quarter to roughly 74% in the current one, and its guidance near $108 billion for the October quarter assumes zero data center chip sales to China, while CoreWeave CEO Michael Intrator said at the same conference that his company is struggling to meet demand every day and that every GPU it has could be sold to multiple different clients.

Anthropic has lined up a $517 billion compute commitment ahead of its confidential IPO, according to a September 6, 2026 report from The Information, up from the roughly $180 billion in server-leasing spend the company previously told investors it expected through 2029. The figure aggregates contracted and reported deals stretching mostly across the next decade rather than a single contract, and actual outlays will hinge on how much capacity is delivered and utilized. Amazon and Alphabet together account for roughly 11 GW and more than $300 billion over about 10 years, Microsoft committed at least $30 billion for roughly 1 GW of Azure capacity, and the SpaceX/Colossus deal runs up to about $45 billion, with neoclouds including CoreWeave, Lambda, Nscale, and Fluidstack also inside the total. AMD captured a 2-gigawatt MI450 Series deal in Helios racks, complementing rather than replacing Anthropic's NVIDIA footprint. NVIDIA reported Q2 FY2027 revenue of $96.22 billion, up 105.85% year over year, with Data Center at $89.02 billion, up 117%, and guided Q3 to $108.0 billion plus or minus 2%, while CFO Colette Kress said the company expects revenue growth of approximately 70% in fiscal 2028.

Oracle's blowout AI cloud results lifted shares of fellow AI infrastructure names CoreWeave, Nebius Group and IREN in premarket trade Friday. CoreWeave climbed around 2.5%, Nebius advanced approximately 2.4% and IREN gained about 1.4% as investors reacted to Oracle's disclosure that cloud infrastructure sales rose 121% year-over-year to $7.4 billion and that the business signed more than $30 billion in new cloud contracts connected to AI. The read-through matters because CoreWeave, Nebius and IREN are all investing heavily in Nvidia GPUs, electricity and data centers on the thesis that clients need ever more computing capacity. Recent company updates have added momentum: CoreWeave announced a field engineering solution to help companies deploy AI models, Nebius partnered with Palantir, and IREN moved ahead with its planned 2-gigawatt Texas data-center project. Wall Street still sees substantial upside, with average analyst estimates from TipRanks pointing to roughly 73% for IREN, 56% for CoreWeave and 26% for Nebius, though the real test remains whether returns on capital can outpace the huge cost of building the infrastructure.

Nvidia CEO Jensen Huang dismissed concerns that the chip giant is fueling its own demand by financing the companies and infrastructure projects that deploy its technology, telling Goldman Sachs' Communacopia + Technology Conference on Thursday, "Is that circular? If that is, let's do more of that." Pressed by Goldman Sachs analyst Jim Schneider on Nvidia's guarantees and financial backing for AI labs and cloud providers, Huang pointed to the $500 billion infrastructure financing platform Nvidia launched with six financial institutions including BlackRock to mobilize third-party capital, and said Nvidia commits relatively little of its own money, adding, "We put in one and 100 comes back in." Huang said Nvidia can see around $100 billion of contracted customer demand behind the projects it helps finance, describing it as "lined-up contracts" and "real stuff," and called the strategy necessary to build a global network of newer cloud providers that serve as distribution channels for Nvidia's architecture. Nvidia disclosed $99 billion of equity investments as of July 26 and another $25 billion of future investment commitments, including a $2 billion investment in cloud provider CoreWeave in January, and has agreed to guarantee up to $105 billion so OpenAI can lease an Ohio data center campus that will run on Nvidia chips, with total guarantees potentially reaching $108.5 billion. Skeptics including short seller Jim Chanos, who challenged Huang directly this week, and Michael Burry, who warned in July that circular spending had reached "biblical proportions," argue the growing financial role makes it harder to separate underlying chip demand from demand enabled by Nvidia's own capital, while Polymarket gives Nvidia a 79% chance of finishing 2026 as the world's largest company by market capitalization. Nvidia's second-quarter revenue jumped 106% year over year to $96.2 billion, with data center revenue up 117% to $89 billion.