--- title: "Will the AI chip market landscape change? CoreWeave issues a warning: abandoning exclusive use of NVIDIA chips will come at a high cost" type: "News" locale: "en" url: "https://longbridge.com/en/news/295719756.md" description: "CoreWeave warns investors that if it is forced to give up exclusive use of NVIDIA AI chips and turn to other solutions, it will face high costs in terms of time and money. As a major cloud computing service provider that relies on NVIDIA GPUs, its clients, such as OpenAI, are attempting to develop their own chips to reduce dependency, which may affect CoreWeave's ability to provide the necessary products to its customers" datetime: "2026-08-12T23:17:03.000Z" locales: - [zh-CN](https://longbridge.com/zh-CN/news/295719756.md) - [en](https://longbridge.com/en/news/295719756.md) - [zh-HK](https://longbridge.com/zh-HK/news/295719756.md) --- # Will the AI chip market landscape change? CoreWeave issues a warning: abandoning exclusive use of NVIDIA chips will come at a high cost According to the Zhitong Finance APP, CoreWeave (CRWV.US), an artificial intelligence (AI) cloud infrastructure provider, has warned investors that if the company must abandon its exclusive use of NVIDIA (NVDA.US) AI chips and switch to other solutions, it may require significant time and investment. CoreWeave stated in a regulatory filing on Wednesday that changing customer demands could force the company to invest "time, capital, and resources" to obtain alternatives to NVIDIA graphics processing units (GPUs). The company warned in the filing that ultimately, this could affect its ability to provide customers with the products they need. Known as the "favorite child of NVIDIA," CoreWeave is part of the emerging group of "new cloud computing service providers." These companies operate by renting out access to leading cloud-based AI computing power infrastructure using NVIDIA GPUs. As the world's largest AI chip manufacturer, NVIDIA is a close partner of CoreWeave and holds a 10% stake in the company. All GPUs currently used by CoreWeave's customers are provided by NVIDIA. However, many customers with significant AI computing needs are working to reduce their reliance on NVIDIA AI chips by developing their own chips. OpenAI, one of CoreWeave's largest customers, has already announced its first AI accelerator built using Broadcom (AVGO.US) technology. Nevertheless, for now, NVIDIA remains the undisputed market leader. The new statement from CoreWeave indicates that some individuals within the company are envisioning a scenario where NVIDIA chips may lose their dominant position in AI work. This content has been included in the "Risk Factors" section of the company's quarterly performance report. Companies typically list hypothetical scenarios that could impact their business in this section. It is worth mentioning that in December of last year, CoreWeave executive Nick Robbins stated that customer demand for NVIDIA technology still overwhelmingly dominates. However, he also emphasized, "If one day we start hearing different voices from customers on a large scale—not just an occasional phone call asking what we think about a certain technology—then that could change our behavior." ## AI Cloud Demand Continues to Surge! Backlogged Orders Exceed $104 Billion, Full-Year Guidance Raised CoreWeave's latest financial report released on Tuesday shows that the company achieved revenue of $2.58 billion in the second quarter, a year-on-year increase of 112%, exceeding the average analyst expectation of $2.56 billion. This figure falls at the high end of the company's previously provided guidance range of $2.45 billion to $2.6 billion. In terms of profitability, the net loss for the second quarter was $626 million, primarily due to substantial interest expenses incurred for expanding infrastructure—this quarter's net interest expense reached $640 million, more than double that of the same period last year. The loss per share was $1.14, still significantly better than the analyst expectation of a loss per share of $1.41. Adjusted earnings before interest, taxes, depreciation, and amortization (EBITDA) reached $1.51 billion, also exceeding the average analyst expectation of $1.43 billion The adjusted operating revenue is within the range of $30 million to $90 million previously provided by management. Despite a year-on-year widening of losses—net loss was $290 million in the same period last year—the loss magnitude was better than market expectations, indicating that the company has maintained a certain level of financial discipline during rapid expansion. In terms of debt scale, as of the end of the quarter, the total debt on the company's balance sheet reached $35 billion, primarily used to pay for the procurement costs of NVIDIA GPUs and other equipment. Interest expenses in the first quarter reached $536 million, exceeding eleven times the adjusted operating revenue, highlighting the financial pressure of a high-leverage expansion model. The most exciting data in CoreWeave's second-quarter financial report is that the backlog of orders reached approximately $104 billion, a year-on-year increase of 246%, continuing to rise from $99.4 billion at the end of the previous quarter. The company also revealed that over $25 billion in net new customer commitments were added in early July, meaning the actual backlog size far exceeds $104 billion. In terms of order conversion, the company expects that about 36% of the remaining performance obligations will convert to revenue within 24 months, implying an annualized revenue pool of approximately $17.8 billion, significantly higher than the midpoint of the revenue guidance of $12.5 billion for fiscal year 2026. Analysts pointed out that the continuously growing backlog indicates that "demand for artificial intelligence remains strong." In terms of customer expansion, this quarter Meta committed an additional $21 billion to CoreWeave; Anthropic signed a multi-year cooperation agreement; and quantitative trading giant Jane Street committed $6 billion. CoreWeave's major clients also include OpenAI and Microsoft. The company also announced a partnership with defense contractor Leidos to provide secure AI cloud services to U.S. federal agencies. CoreWeave expects third-quarter revenue to be between $3.45 billion and $3.6 billion, with a midpoint of $3.525 billion, indicating a year-on-year growth of approximately 158%. Data shows that analysts previously expected sales to be at the lower end of this range. Based on strong demand momentum, CoreWeave has also raised its full-year outlook: the full-year revenue guidance for 2026 has been raised from $12 billion to $13 billion to $12.4 billion to $13.2 billion, while analysts previously expected full-year revenue to be $12.63 billion. The adjusted operating profit expectation has been raised from $900 million to $1.1 billion to $960 million to $1.15 billion. Annual capital expenditures have been raised from $31 billion to $35 billion to $35 billion to $39 billion. As one of the few publicly listed new cloud computing providers, CoreWeave's performance is seen as an important barometer of overall AI computing power demand. Its better-than-expected performance confirms the continued heating up of AI infrastructure investment. As AI infrastructure investment transitions from an "arms race" to "structural growth," CoreWeave, with its deep ties to NVIDIA, over $100 billion in contract backlog, and continuously expanding global data center footprint, is becoming the most representative "new cloud" benchmark in this multi-trillion-dollar AI infrastructure wave. As the company's CEO Mike Intrator stated, the company has reached a turning point where "scale begins to translate into operational leverage"—the key question moving forward is whether the $35 billion in debt can be converted quickly enough into sustainable profits in the context of sustained demand surges ### Related Stocks - [CRWV.US](https://longbridge.com/en/quote/CRWV.US.md) - [NVDA.US](https://longbridge.com/en/quote/NVDA.US.md) - [NVDD.US](https://longbridge.com/en/quote/NVDD.US.md) - [NVDY.US](https://longbridge.com/en/quote/NVDY.US.md) - [NVDX.US](https://longbridge.com/en/quote/NVDX.US.md) - [CORD.US](https://longbridge.com/en/quote/CORD.US.md) - [NVD.US](https://longbridge.com/en/quote/NVD.US.md) - [NVDQ.US](https://longbridge.com/en/quote/NVDQ.US.md) - [CRWU.US](https://longbridge.com/en/quote/CRWU.US.md) - [CWVX.US](https://longbridge.com/en/quote/CWVX.US.md) - [NVDL.US](https://longbridge.com/en/quote/NVDL.US.md) - [NVDU.US](https://longbridge.com/en/quote/NVDU.US.md) - [NVDS.US](https://longbridge.com/en/quote/NVDS.US.md) - [CRWG.US](https://longbridge.com/en/quote/CRWG.US.md) - [07788.HK](https://longbridge.com/en/quote/07788.HK.md) - [CWY.US](https://longbridge.com/en/quote/CWY.US.md) - [NVDB.US](https://longbridge.com/en/quote/NVDB.US.md) - [NVDG.US](https://longbridge.com/en/quote/NVDG.US.md) - [NVDO.US](https://longbridge.com/en/quote/NVDO.US.md) - [NVDW.US](https://longbridge.com/en/quote/NVDW.US.md) - [NVYY.US](https://longbridge.com/en/quote/NVYY.US.md) - [NYYY.US](https://longbridge.com/en/quote/NYYY.US.md) - [DIPS.US](https://longbridge.com/en/quote/DIPS.US.md) - [MAGX.US](https://longbridge.com/en/quote/MAGX.US.md) - [SMH.US](https://longbridge.com/en/quote/SMH.US.md) - [SOXX.US](https://longbridge.com/en/quote/SOXX.US.md) - [SOXL.US](https://longbridge.com/en/quote/SOXL.US.md) - [SOXQ.US](https://longbridge.com/en/quote/SOXQ.US.md) - [XSD.US](https://longbridge.com/en/quote/XSD.US.md) - [PSI.US](https://longbridge.com/en/quote/PSI.US.md) - [FTXL.US](https://longbridge.com/en/quote/FTXL.US.md) - [07388.HK](https://longbridge.com/en/quote/07388.HK.md) - [09388.HK](https://longbridge.com/en/quote/09388.HK.md) - [OpenAI.NA](https://longbridge.com/en/quote/OpenAI.NA.md) - [AVGO.US](https://longbridge.com/en/quote/AVGO.US.md) - [META.US](https://longbridge.com/en/quote/META.US.md) - [MSFT.US](https://longbridge.com/en/quote/MSFT.US.md) - [LDOS.US](https://longbridge.com/en/quote/LDOS.US.md) - [NVD.DE](https://longbridge.com/en/quote/NVD.DE.md) ## Related News & Research - [Nvidia just recruited Wall Street to help fund $500 billion in AI infrastructure. Here’s the catch.](https://longbridge.com/en/news/295590827.md) - [AI weekly: Nvidia does deal with Wall Street, China's AI weather forecasts](https://longbridge.com/en/news/295777975.md) - [Jensen Huang's latest pitch to investors: Nvidia's AI compute is its own asset class](https://longbridge.com/en/news/295562673.md) - [Nvidia’s Best Customers Have a Reason to Stop Buying So Many Nvidia Chips](https://longbridge.com/en/news/295572288.md) - [India's Larsen and Toubro secures AI data centre order worth up to $1.57 billion](https://longbridge.com/en/news/295760762.md)