Insurance Analysis 71 Comments

NVIDIA: Not Indomitable Anymore

Advertisements

The recent developments surrounding Nvidia Corp.and the artificial intelligence (AI) sector have sparked a notable shift in investor sentiment.Where once investors eagerly pounced on Nvidia stocks during price dips,the narrative has altered dramatically in light of a significant announcement from a Chinese AI startup known as DeepSeek.This announcement,which indicated that DeepSeek could deliver high-performance AI solutions at reduced costs,resulted in a staggering 17% plummet in Nvidia's stock price in just one day.The company's market value sunk by approximately $590 billion,and despite some recovery,the stock remains over 11% down from its January high.

This sudden change in fortunes comes at a curious time.Major clients of Nvidia including Amazon,Alphabet,Meta Platforms,and Microsoft have collectively planned an estimated $300 billion in capital spending for the year ahead.Yet,despite this anticipated investment,the atmosphere among investors has soured,highlighting a deeper unease about the future of AI expenditures.

Only after Nvidia's stock descended more than 21% did buyers start showing interest again,indicating a notable shift in strategy.Historically,such underselling has happened only sporadically,underscoring growing hesitancy about the longevity of the AI investment phenomenon.Pertinently,the features touted by DeepSeek include a claim that it utilizes fewer chips for its AI models,raising concerns over Nvidia's established supremacy.

Gene Munster,a managing partner and co-founder of Deepwater Asset Management,encapsulated the change succinctly,stating,“People are concerned about when the frenzy will end; I believe DeepSeek serves as a wake-up call,reminding us that the culmination of this boom might occur sooner than anticipated.” He elaborated on how the narrative that once depicted AI as an indomitable force is now perceivable as a story ripe for radical shifts.

The adverse mood permeates expectations surrounding Nvidia’s upcoming earnings report set for February 26.Unlike its quarterly updates over the last couple of years that were met with almost universally positive predictions,the current environment contrasts sharply,with a palpable sense of skepticism towards the stock's potential to achieve previous highs.

Morgan Stanley analyst Joseph Moore articulated this prevailing sentiment in a recent report,noting that “the negative stock reaction has taken center stage,” posing significant risks to investors.In an environment where skepticism is on the rise,the crucial question remains: will accelerating revenue assuage these concerns?While Moore expresses a belief that a return to growth is plausible,he acknowledges the debate surrounding it.

Nvidia's growth for the past few quarters has been exponential; however,emerging investor doubts create an uphill battle during what has been a remarkable revenue surge year over year.According to Bloomberg estimates,the company is projected to experience a revenue growth rate of 73%,a stark decline from the previous quarter’s growth of 94%,and notably down from a staggering 265% at the same time last year.

Yet,amidst a backdrop of negativity,some analysts perceive the downturn as an opportunity.Mark Lipacis and his team at Evercore ISI upgraded their tactical ratings before the earnings announcement,positing that DeepSeek’s sell-off presents a buying opportunity.There remains a significant window before the earnings report,potentially allowing time for the stock to recover some lost ground.

The stock's fall might enable investors to reassess Nvidia’s valuation,which is currently set at an anticipated price-to-earnings ratio of about 30 times—considerably lower than its five-year average exceeding 40 times.However,the persistent absence of strong buying interest means the stock remains precariously positioned ahead of the earnings report,particularly if the earnings miss the ambitious benchmarks investors have come to expect.

According to Ivana Delevska,chief investment officer at SPEAR Invest,Nvidia’s potential lackluster performance could result in the stock trading within limited ranges in the first half of the year.Additionally,uncertainty surrounding Nvidia’s Blackwell chips contributes to investor stress.Manufacturing challenges and supply chain constraints have hindered the launch of anticipated new product lines.Despite company assurances of robust demand for Blackwell driving future growth,worries about costs endure.

“There's been a palpable tension among people regarding Blackwell,” Delevska commented.“Although management has consistently asserted that things are going smoothly,there’s a reality that this narrative isn’t being fully embraced by investors—that might explain the lack of rebound in share prices before the earnings release.”

Nvidia’s presence in the AI market is undeniable— it is essentially the linchpin for many players in the sector,providing the essential AI processors to train and run a multitude of models.However,this dominance leads to higher costs and limited availability of computational power required for the training of the next generation of AI models.The question becomes whether other competitors can harness the momentum created by DeepSeek to challenge this status quo.

OpenAI,in particular,has emerged as a pivotal entity within this discussion.Recognized as a frontrunner in the GenAI revolution and a significant Nvidia client,OpenAI has made headlines for its ambition to establish an extensive AI computational infrastructure with a $500 billion initiative dubbed the Stargate Project.If CEO Sam Altman succeeds,Nvidia stands to receive a fraction of this funding.

Reports have surfaced indicating that OpenAI is developing its own AI chips to diminish reliance on Nvidia products.Currently in advanced design stages,this move could allow OpenAI to better control its supply chain for essential hardware.Though production incurs millions in expenses and could take several months,OpenAI is rumored to expedite the development process,aiming to commence mass production by 2026.Importantly,these chips are initially intended for running models rather than training but indicate a strategic pivot aimed at negotiating better terms with Nvidia and other chip suppliers.

The ripples of the DeepSeek announcement have led to a reevaluation of foundational assumptions within the AI model development space.DeepSeek’s assertion that creating its core R1 model costs merely a fraction of similar models from OpenAI and Google has provoked introspection regarding whether massive computational resources are genuinely critical for advanced AI systems.If less computational capability translates to effective performance,the previously irreplaceable Nvidia chips could find themselves at risk of displacement.

Meta,another player intent on breaking free from Nvidia’s grip,is reportedly in discussions to acquire a South Korean AI chip startup,FuriosaAI.This move aims to develop an alternative line of chips,with the deal expected to conclude as soon as this month.Meta has already launched its MTIA (Meta Training and Inference Accelerator) chips tailored for conventional AI workloads,indicating its deeper commitment to reducing reliance on Nvidia as it strategizes for future technology builds.

Microsoft does not lag either; it continues to design its suite of AI chips to bolster performance and support its AI data centers effectively.Amazon is engaged in establishing an entire ecosystem around AI chips that support its expansive cloud infrastructure,rolling out Trainium3 later this year,their latest generation of AI training and execution chips leveraging technology from its 2015 acquisition of Israeli startup Annapurna.

Meanwhile,Google is incorporating AI models into its chip design processes,driven by its DeepMind division to accelerate development.Apple,under the guidance of executive Johny Srouji,is expanding its AI chip initiatives and is reportedly developing its first AI server chip through collaborations with semiconductor giant Broadcom,anticipating commercial production using cutting-edge TSMC technology by 2026.

To date,Nvidia holds a commanding 80% of the global AI chip market.As rival companies pour billions into development efforts,the landscape hints at a probable shake-up.With a myriad of firms vying for a foothold in the AI chip realm,it seems plausible that viable alternatives may soon arise.Coupled with the paradigm shift prompted by DeepSeek’s revelations,Nvidia’s reign may face significant challenges in the near future.

Leave A Comment