CoreWeave sees AI-compute demand remaining supply-constrained as it expands into higher-value software and managed services.

Overview

This conference takeaway examines CoreWeave’s effort to capitalize on strong AI-compute demand while broadening its platform from GPU infrastructure into managed inference, software services, Omni, and self-built data centers. Goldman Sachs remains Neutral rated with a $139 12-month target price.

Core views

Goldman Sachs characterizes the AI-compute demand environment as robust, with CoreWeave seeing demand materially exceed available supply and its largest customers consuming substantial infrastructure. Management also identified enterprise demand as an additional growth vector: companies are increasingly treating AI infrastructure as a strategic asset and building dedicated clusters for training and inference, with Caterpillar cited as an example. Although significant new supply has been announced across the ecosystem, the report emphasizes that building and operating AI infrastructure at scale remains difficult, and announced projects must still be executed successfully. Against this backdrop, CoreWeave is increasing control over supply through self-built data centers, with its first facilities expected to begin coming online over the next year. The report says vertical integration is primarily about controlling deployment timelines, operations, and future product innovation rather than simply capturing margins. Owning more of the infrastructure is expected to provide a better environment for iterating on technologies and optimizing performance across the stack. The report’s second major thread is platform depth. Management highlighted managed inference as a way to monetize software and services layered over infrastructure. Customers are seeking a complete AI stack—including infrastructure, networking, storage, observability, orchestration, and operational support—rather than raw compute alone. Goldman Sachs believes these software and service layers can have greater strategic value than GPU access, supporting deeper customer relationships and more durable revenue streams. Goldman Sachs also sees multiple paths to improved economics. Self-built data centers may enhance operational control and accelerate innovation, while Omni allows customers to use CoreWeave’s software stack on infrastructure that CoreWeave does not own. The latter expands the addressable market while using third-party balance sheets. Together with higher software attach rates and managed-services adoption, these initiatives could make growth and profitability less dependent on continually adding GPUs and data-center capacity. Goldman Sachs reiterates its Neutral rating and unchanged 12-month $139 target price, based on 22x CoreWeave’s SNTM non-GAAP EBIT. The target implies 17x 2028 EBIT. The report identifies upside scenarios including AI-demand growth that continues to outpace supply, continued GPU-related technical differentiation, a faster-than-expected path to positive free cash flow, and stronger demand for CoreWeave’s broader software suite. Downside risks include weaker generative-AI compute demand, greater competition from hyperscalers and neoclouds, data-center equipment procurement disruptions, and difficulty raising additional financing at competitive rates.

Analysis framework

The report synthesizes management commentary from the Communacopia + Technology conference, then assesses the supply-demand backdrop for AI compute, CoreWeave’s infrastructure-control strategy, and its software and services monetization paths. It values the company using an EV/EBIT multiple applied to SNTM non-GAAP EBIT and considers longer-term 2028 EBIT valuation.

Methodology notes

• Valuation methods

  EV/EBIT valuation multiple

  Goldman Sachs sets the $139 target price using 22x SNTM non-GAAP EBIT, with the target also implying 17x 2028 EBIT.

• Industry AnalysisSupply-demand framework

  AI-compute supply-demand assessment

  The report evaluates CoreWeave against demand that management says exceeds supply, while noting that announced capacity must still be successfully built and operated.

• Competition & strategyEconomic Moat and Competitive Advantage

  Platform differentiation through integrated infrastructure and software

  The report considers how self-built data centers, managed inference, and Omni could improve control, innovation, customer relationships, and strategic differentiation.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• CoreWeave Inc. (CRWV)

  Primary covered company; positioned to benefit from robust AI-compute demand and software-led platform expansion.

  Strengths

  Demand is described as exceeding available supply; managed inference, software services, Omni, and self-built data centers may deepen differentiation and customer relationships.

  Weaknesses

  Growth and profitability have historically been closely tied to adding GPU and data-center capacity.

  Comparison

  Faces competition from hyperscalers and neoclouds.

  Risks

  Potential demand reduction, competition, equipment procurement disruption, and financing challenges.

Key data

• 12-month price target$139Unchanged; based on 22x SNTM non-GAAP EBIT.
• Closing price$94.94Price as of 9 September 2026 close.
• Implied upside46.4%Table-implied upside to the $139 target price.
• 2028 EBIT valuation multiple17xImplied by the maintained target price.
• Revenue forecast$5,131.1mn in 2025; $12,957.1mn in 2026E; $27,245.3mn in 2027E; $44,185.2mn in 2028EGoldman Sachs forecast table.
• EBIT forecast$667.2mn in 2025; $1,067.5mn in 2026E; $4,191.4mn in 2027E; $8,029.4mn in 2028EGoldman Sachs forecast table.

Impact & implications

The report argues that CoreWeave’s opportunity is increasingly tied to monetizing the higher-value layers of the AI stack, not only supplying GPU capacity. Greater infrastructure control and software-led offerings could broaden its market and create growth and profitability paths that are less dependent on continuous physical-capacity expansion.

Risks

• AI demand could grow faster than expected and remain supply-constrained, creating upside risk.
• Continued technical differentiation through future GPU generations could support upside.
• A faster-than-expected path to positive free cash flow or stronger software-suite demand could support upside.
• Overall generative-AI compute demand could decline.
• Competition from hyperscalers and neoclouds could intensify.
• Data-center equipment procurement could be disrupted.
• Additional financing may be difficult to raise at competitive rates.

What to watch

• Whether AI-compute demand continues to exceed available supply.
• Progress of CoreWeave’s first self-built data centers expected to begin coming online over the next year.
• Adoption of managed inference, software services, and Omni.
• Software attach rates and managed-services adoption.
• The company’s path toward positive free cash flow.
• Competitive developments among hyperscalers and neoclouds.
• Availability of data-center equipment and financing conditions.