Over the past two weeks, the excitement in the world of Artificial Intelligence (AI) has been unprecedentedSince the inception of ChatGPT ignited the AI revolution, this period stands out not only for its rapid developments but also for the sheer uncertainty it has injected into global marketsOn one hand, the approval of a monumental $500 billion AI infrastructure project, championed by Trump and spearheaded by OpenAI, set ambitious new benchmarks for the industryOn the other hand, the emergence of Deepseek has curbed a relentless upward trajectory that had persisted for over two years, leaving investors in a whirlwind of confusion about the essentiality of computational power.

Currently, Nvidia's stock price has once again flirted with the $130 mark, hinting at a recovery of the losses triggered by the Deepseek controversyThis begs the question: do you believe in the ongoing AI computational power crisis? Many of those who have long held faith in AI found themselves swayed by overwhelming market sentiment

Advertisements

Even staunch supporters began to waver, questioning whether such large investments in AI still made sense amid rising skepticism from short sellersAmidst all this uncertainty, several large funds opted to downsize their holdings in a bid to mitigate perceived risks.

One of the heavier conclusions derived from all this is that the advent of Deepseek holds significant implicationsInvestors have been prompted to re-evaluate the critical nature of investments in AI hardware while also exploring more cost-effective technological pathwaysDebates about whether certain stakeholders have already gained the upper hand now seem trivial; what remains undeniably clear is the continued importance of computational powerStocks within the AI chain on Wall Street remain buoyant, and our Chinese companies have started to gain traction, reflecting renewed interest from foreign investorsThe impending push from China towards amplifying their AI investments could further bridge the valuation gap between Chinese and US stocks.

The recent rebound in Hong Kong stocks and Chinese concept stocks can be largely attributed to the impact of Deepseek

Advertisements

For instance, Alibaba's stocks on the US market have bloomed with newfound hope, reminiscent of the narrative surrounding last year's CSP investmentsThis optimism is fueled by the robust Qwen2.5max model performance, Alibaba’s computational reserves, and their comprehensive cloud platform, all of which serve as a scalable template for firms like Xiaomi and Kingsoft Cloud.

In these past few weeks, several hedge funds and long-only investors have begun to replenish their positions in Chinese firms, taking cues from foreign market participants announcing grandiose predictions akin to the "China eats the world" narrative previously seenHowever, the euphoria fueled by such policy shifts appears more sustainable than rapid increases driven by short-term policy spikesIn simpler terms, the market has finally acknowledged that China has viable AI prospects that cannot be dismissed lightly.

For over a year now, the AI landscape has seen a lack of validated applications that can generate profits

Advertisements

Questions abound as to whether investing heavily in AI applications is necessary, and if only large firms can afford to pursue this avenueWith Deepseek's emergence, however, the confidence surrounding 2025 being a breakout year for AI applications has substantially solidified, revealing that small to medium-sized businesses and developers have access to lower-cost pathways.

While the market has indeed acted as an emotional amplifier, exacerbating fluctuations not only in US markets but also across Asia during the Spring Festival period, the direct impact of Deepseek's dramatics was most pronouncedFor instance, during one day, Nvidia plummeted by 17%, Broadcom and TSMC followed suit falling 17% and 13% respectivelyThe mass sell-off triggered even sharper declines within smaller, AI-focused stocks; Credo, which had tripled in price since last year, dropped 30%, and VST, an AI power solution recently bought by none other than investment guru Pelosi, dropped 28% on the same day.

This volatility, while dramatic, was among the most significant single-day declines that Nvidia has faced in two years

• Investment Opportunities in the Nissan-Honda Merger
• Europe's Interest Rate Path May Change
• Fitness Fever Fuels Brand Makeovers
• Nvidia's Short-Term Stability
• The Exiled Korean Cars

Yet, looking back over a more extended timeline indicates that many of these stocks have seen tripling of their values without adequate corrections, and Deepseek's emergence threw the narrative concerning the arms race for computational power into question.

Interestingly, not all stocks behaved uniformlyOn the contrary, tech giants like Apple and Meta held up steadfastly, even recording a 2% spike, underlining the growing certainty surrounding AI applicationsMeta’s innovative smart glasses saw soaring sales last year, with projections indicating over 2.25 million units for 2024. Their success has spurred other tech entities to place substantial bets on AI-integrated eyewear, with companies like Samsung, Xiaomi, Baidu, and even ByteDance getting in on the action.

Apple has taken a different path, embedding AI in a more diverse mannerAlthough their AI initiative seems quieter currently, the potential utility of incorporating advanced AI voice capabilities into the iPhone grows clearer—especially as the performance of models like ChatGPT, Deepseek, and Doubao demonstrate increasingly robust functionality

Although Siri’s current performance lags, its integration could streamline user experience significantly.

One point deserving mention is the pivotal question individuals must consider in their everyday lives: "What tasks could AI assist me with, and how would it make my life more convenient?" Reflecting on personal experiences reveals that AI can indeed facilitate our lives in numerous ways, and AI voice assistants stand out as an accessible entry pointConsequently, the side of edge applications for AI voice recognition will likely unlock a burgeoning market.

How can these efficiencies be achieved? Ultimately, the need for more robust and efficient computational power remains centralThere is little merit in scrutinizing hardware investments while applications are yet to fully take off; such analysis often skews rather emotionalThe emergence of Deepseek merely prompts a restructuring narrative around training, ultimately enhancing the scaling implications for inference narratives—a notion reflected in the renowned Jevons Paradox, suggesting that enhancing resource efficiency can, paradoxically, increase overall consumption.

In the past fortnight, many companies began integrating Deepseek’s API to leverage its capabilities, revealing complaints over server overloads as demand surged

When benchmarking models like 910B against H200 within Deepseek's R1 model, the performance discrepancies are stark, noting significant escalations in processing power that GB200 can provide.

Furthermore, discussions on ASIC inference versus GPU training have raged on in AI circles in the past couple of yearsIf anything, the coalescence of highs for both Nvidia and Broadcom demonstrates that one entity's pivot towards ASIC doesn’t necessarily pose a threat to Nvidia amid soaring demand overall.

Meanwhile, in the US stock market, it’s commonplace for new AI technologies to stimulate simultaneous growth in associated sectorsIn stark contrast, the A-share market seems often characterized by volatile surges frequently oscillating between various sectorsInvestors navigating the murky waters of AI investments in China face increasingly intricate negotiations compared to their US counterparts.

Coming back to Nvidia, the salvation from disaster for the company largely rests in the hands of its CSP clients

Luckily, the Deepseek crisis coincided with the financial reporting cycle for major tech players; otherwise, Nvidia might not have rebounded as swiftlyDespite external analysts reassuring on the ongoing demand for computational power, nothing proves more tangible than concrete Capital Expenditure figures.

As for Q4 capital expenditure specifics, Meta's allocation stood at $14.8 billion—a staggering 88% year-on-year increase—and significantly above market forecastsAmazon also made headlines with a whopping $27.6 billion in capex while Microsoft and Google followed close behind with similarly impressive figuresLooking ahead to 2025, projections from these tech giants reveal an unwavering commitment to bolster their positions in the AI market, with figures soaring as high as a total of $100 billion for AmazonSuch investments illustrate the lucrative prospects of steadfast profit generation from utilizing AI technologies.

With the bulwark of support from four major clients, Nvidia appears insulated from risks in the immediate future, as demand for computational power continues unchecked, spurring a renewed confidence among CEOs regarding reduced inference costs

Given the robust profitability inherent in their principal operations, they seem primed to support hefty AI expenditures—up to billions—while still enjoying growing free cash flow.

Beyond the backing from large clients, the stock plunge day for Nvidia elicited a compelling counter-response from individual investors who have consistently demonstrated faith in AI innovationsFollowing the Deepseek-induced downturn on Wall Street, retail investors integrated a surprising $4.25 billion back into the market, underscoring their resilience in the face of bearish sentiment.

As the immediate computational demand crisis fades, the pressing question remains: how should we view Nvidia moving forward? This merits thorough considerationOn one hand, if China indeed experiences breakthroughs in AI, the repercussions for a sensitive figure like Trump could be significant, especially after his recent targeting of a $500 billion infrastructure initiative

Consequently, one might anticipate heightened restrictions on Nvidia's chip exports, with fears already swirling over the limitations imposed on various versions of the H200.

Nvidia can't afford to dodge these concerns—the projected revenues of $35.08 billion by Q3 2024, with around $5.4 billion sourced from the Chinese market, might be influenced severely if restrictions tightenChanges in this dynamic would unavoidably render valuation impacts.

2. From Wall Street's outlook, while target price revisions can vary, understanding how analysts anticipate Nvidia's growth potential offers useful insightsCitigroup recently lowered its target price from $175 to $163, largely due to the perceived risks linked to restrictions on data centers in ChinaOn the flip side, Morgan Stanley has maintained a "buy" rating while adjusting its target price upward from $152 to $172.

Other financial institutions share similar bullish sentiments; Bank of America projected a "buy" rating with a target price of $190, showcasing faith in Nvidia's growth in emerging markets

Meanwhile, Deutsche Bank's analysts expressed continued confidence with a target price of $220 citing Nvidia’s promises in AI and data center leadership, while Goldman Sachs cited a target of $200 based on anticipated growth from burgeoning AI and cloud computing trends.

In conclusion, Nvidia stands at a pivotal juncture, valued at $130 with a market capitalization of $3.15 trillionWhile these figures don’t imply another doubling of its value in the near term, those committed to investing in AI cannot afford to overlook NvidiaIn the immediate aftermath, Nvidia is likely to remain a vital Beta stock closely tied to the US's expansive AI infrastructure investments.

Investors might, however, find favor with other niche players beyond NvidiaThe $500 billion intergalactic initiative could lead to heightened demand for various hardware components—fiber optics, essential chipsets for GB200 servers, server power supplies, and so forth