Simply put
- CoreWeave’s revenue increased threefold, but its order backlog also grew, and it continued to face losses that have now shrunk from 20% last year to just 2%.
- After-hours trading saw a 10% drop in stocks, as thinner margins and rising losses raised concerns among investors.
- Analysts suggest that current hardware depreciation is making losses appear less severe, but as competitors improve their offerings, CoreWeave’s advantage may diminish.
AI infrastructure provider CoreWeave reported another strong quarter, showing an impressive revenue backlog in its latest earnings announcement.
However, investor reactions were not favorable, as the stock took a hit in after-hours trading, driven by concerns over increasing costs and lower operating margins.
CoreWeave’s revenue hit approximately $1.2 billion, a significant jump from $300 million a year prior, although it faced a net loss of $290 million due to higher interest and infrastructure expenditures. This information was disclosed in a statement released on Tuesday.
Michael Intrator, co-founder and CEO, mentioned the company’s rapid expansion to meet a soaring demand for AI.
The most recent numbers show CoreWeave’s adjusted net losses rising to about $131 million, despite generating an adjusted operating profit of $200 million, reflecting increased operating costs and stock-based compensation. The revenue backlog now stands at around $30 billion, which includes several performance obligations and amounts expected to be recognized in future periods.
Even with positive earnings reports, investor sentiments dipped, leading to CoreWeave’s stock trading down over 10% during post-market hours.
narrows the lead
Some analysts caution that the fragile financial framework underlying CoreWeave’s growth could be obscured by the thin margins and GPU depreciation assumptions.
“The quarterly adjusted operating profit margin was a mere 16%, which doesn’t bode well for a capital-intensive operation,” noted Jeffrey Emmanuel, CEO of Pastel Network, a blockchain infrastructure firm. “They are depreciating all H-100 GPUs linearly, assuming a six-year lifespan.”
Emmanuel added that CoreWeave’s machines currently remain competitive in terms of price and performance, but points out that the gap is narrowing as competitors advance in technology.
Ram Kumar, a contributor to the decentralized AI platform Openledger, mentioned, “CoreWeave initially gained an edge by securing Nvidia’s latest GPUs first, but that advantage is dwindling as rivals catch up.”
He warned that without quicker margin growth and locking in usage, CoreWeave risks slipping from high-growth status to serving infrastructure with low returns, especially amid price competition.
With limited operating margins and significant fixed costs, CoreWeave faces heightened vulnerability compared to various hyperscalers, Kumar further remarked. “Any slowdown in AI training budgets or customer integration could quickly tighten cash flows. Long-term contracts and software loyalty will be crucial for resilience.”
Additionally, there are concerns regarding CoreWeave’s accounting. Emmanuel argued that the depreciation methods applied may not accurately reflect reality, suggesting that their financial position might appear stronger than it is.
CoreWeave has yet to respond to requests for comments on the matter.
