Top insights for IT pros
From cybersecurity and big data to cloud computing, IT Brew covers the latest trends shaping business tech in our 4x weekly newsletter, virtual events with industry experts, and digital guides.
As of Feb. 14, chipmaker Nvidia—whose share price has more than tripled in the last year alone—was the third-most valuable company in the US.
CNBC reported that Nvidia’s peak price of $739 per share after close that day translated to a market capitalization of $1.83 trillion, putting it just ahead of Google parent Alphabet, whose market cap was $1.82 trillion. (Nvidia surpassed Amazon, whose market cap was $1.75 trillion the day before.) That left Apple and Microsoft as the only US companies with higher market caps, of $2.8 trillion and nearly $3 trillion, respectively.
On the same date one year earlier, Nvidia stock had closed at just shy of $230 per share.
The hardware manufacturer once primarily sold consumer graphics processing units (GPUs) for gaming and graphics work, but it benefited immensely once GPUs found widespread use as parallel computing units capable of dramatically accelerating workloads in data centers. The company’s recent surge is largely tied to growing interest in AI tools and deep learning methods that rely heavily on parallel processing.
Some of Nvidia’s high-end AI chips sell for more than $40,000 on reseller markets like eBay. Meta alone expects to have purchased 350,000 of Nvidia’s premium H100 chips by the end of 2024, CNBC reported. With analysts at Raymond James estimating Nvidia is selling the H100 for $25,000 to $30,000, that’s at least $9 billion.
Nvidia competitors Intel and AMD remain years behind Nvidia’s cutting-edge technology, according to Yahoo Finance. AMD has countered with an aggressive pricing strategy, selling its comparable MI300X chips for $10,000–$15,000. Intel, which is undergoing a pivot to chip manufacturing, has yet to release its Gaudi3 AI chip, but is also betting on the consumer AI PC market. Yet one of Nvidia’s biggest advantages remains its CUDA software stack, the GPU parallel processing API used in the widest variety of workloads.
Some of the biggest tech firms in the world, including Amazon, Google, Meta, and Microsoft, are trying to break free with in-house AI chips. Nvidia is moving in on their turf as well, building its own cloud offerings and providing sought-after chips to upstarts like CoreWeave.