NVIDIA vs. Cisco: Linasbeliunas Analyzes the AI Boom Through a Historical Lens

In a recent LinkedIn post, Linasbeliunas offers a compelling historical analysis comparing NVIDIA’s current market position to Cisco’s during the dot-com bubble, arguing that the two scenarios are fundamentally different. Linasbeliunas suggests that while many are quick to label NVIDIA as a mere repeat of Cisco’s past, a closer look at the underlying financial metrics and market drivers indicates a distinct and potentially more robust situation.

Distinguishing a Bubble from a Boom

Linasbeliunas begins by acknowledging the common comparison between NVIDIA and Cisco in 1999, a period when Cisco’s hardware represented the peak of the dot-com enthusiasm. He highlights the core elements that fueled Cisco’s rise: a real internet, tangible demand, and a compelling narrative. However, as Linasbeliunas points out, the stock valuation significantly outpaced the business’s actual performance.

“At the peak, Cisco briefly reached a ~$550B market cap on roughly ~$19B in annual revenue.”

This dramatic disconnect, Linasbeliunas contends, is where the comparison with NVIDIA begins to falter. He posits that the key differentiator between a speculative bubble and a sustainable boom lies in the relationship between price movement and tangible proof of value. According to Linasbeliunas:

“A bubble is price moving faster than proof. ↳ A boom is proof compounding fast enough to justify price.”

Linasbeliunas elaborates on this distinction by examining NVIDIA’s financial trajectory since 2023. He notes that NVIDIA’s earnings per share (EPS) have largely kept pace with its stock price appreciation, a stark contrast to Cisco’s pre-crash divergence. The post cites significant revenue growth, particularly in the data center segment, and consistently high gross margins as evidence of this compounding proof.

The Funding Difference: Enterprises vs. Startups

A crucial element in Linasbeliunas’s analysis is the source of funding for the current AI capital expenditure (capex) cycle. He contrasts this with the dot-com era, where many startups lacked sustainable business models. In the current AI landscape, Linasbeliunas highlights that investments are primarily driven by established tech giants and major enterprises.

Shifting Capital Flows

As Linasbeliunas observes, the AI capex cycle is being fueled by entities like Microsoft, Google, Amazon, and Meta, as well as sovereign buyers and businesses aiming to enhance productivity through computing power. This, in his view, represents a more grounded and demand-driven investment environment compared to the speculative funding of the past.

“More importantly, the current AI capex cycle is not being funded by fragile dot-com startups with no business model. It is being funded mostly by Microsoft, Google, Amazon, Meta, sovereign buyers, and enterprises trying to turn compute into productivity.”

Potential Risks and Future Outlook

While presenting a case for NVIDIA’s current strength, Linasbeliunas does not dismiss the possibility of a market correction. He outlines several potential triggers, including disappointing hyperscaler return on investment (ROI), a slowdown in capex, advancements in custom chip development, or overly stretched expectations. Linasbeliunas cautions that such events could lead to a significant compression of NVIDIA’s valuation multiples.

However, he differentiates a correction from a bubble bursting. In Linasbeliunas’s analysis, Cisco represented infrastructure built ahead of cash flows, whereas NVIDIA, in his view, is infrastructure being actively pulled by existing and projected cash flows. The ultimate question, as Linasbeliunas frames it, is whether NVIDIA’s earnings growth can continue to outpace market expectations.

Linasbeliunas also references a separate insight from Coatue regarding future AI spending, suggesting further avenues for understanding the evolving AI investment landscape.