This week, Grok 3 heightens demand for Nvidia GPUs, Microsoft heats up the quantum market and Palantir falters on defense funding woes

On February 19, 2025, Microsoft unveiled its first-ever quantum chip—codenamed Majorana 1—claiming a breakthrough using a “new state of matter” called topological superconductivity. This innovation, designed to create topological qubits, directly addresses the high error rates that have historically hampered quantum computing’s scalability, with Bloomberg reporting Microsoft’s demonstration as showing “unprecedented stability.”

Meanwhile, NVIDIA stock—hovering just below $140—continued its recovery from January’s “Great DeepSeek Panic” fueled by news of Elon Musk’s xAI deploying 200,000 NVIDIA GPUs for Grok 3 (with plans to scale to over a million). In contrast, Palantir shares experienced a nearly 10% dip on the same day, thanks to growing concerns over potential U.S. defense budget cuts that threatened its AI-driven growth prospects. At the time of writing, its stock was down another 8.5% on Thursday.

Majorana’s quantum leap forward

Microsoft’s Majorana 1 chip announcement sparked excitement in the quantum computing sector. Specialized stocks such as D-Wave Quantum Inc, IonQ Inc and Rigetti Computing saw significant upsides, as Reuters reported. Tapping topological superconductivity, Majorana 1’s topological qubits promise to dramatically reduce error rates—a long-standing obstacle in quantum R&D. Microsoft projects that its new Topological Core architecture (Check out the official announcement) could scale to one million qubits. The company calls the chip the “world’s first topoconductor,” as Verge reported.

That could drive advances in areas ranging from cryptography to pharmaceutical research. For context, last month, NVIDIA CEO Jensen Huang spooked quantum investors by saying that practical quantum computing was likely 15 to 30 years away, with 20 being a base case.

Why it matters: Microsoft’s announcement of enhanced stability and lower error rates thanks to topological qubits address a fundamental engineering hurdle in quantum computing. Should Majorana 1 deliver on its potential, the path to scalable quantum computing could shift from theoretical possibility to tangible reality. Ramifications for data security, drug discovery and complex simulations could be significant.

NVIDIA mostly rebounds from DeepSeek panic

NVIDIA has seen some of the most meteoric stock growth of any company in history in recent years, but its shares lost roughly $500 billion after the launch of DeepSeek R1. Since then, NVIDIA shares have effectively shaken off the “Great DeepSeek Panic.” That was a period of investor apprehension that cheaper GPUs might challenge its dominance in the high-end AI chip market. The rebound is largely attributed to xAI’s massive commitment, with Elon Musk’s venture deploying 200,000 GPUs and planning to scale to a million units (Capacity Media). Adding to this momentum, a substantial order for 10,000 GPUs from South Korea (Hindustan Times) further solidifies NVIDIA’s role in powering large-scale AI initiatives.

Why it matters: Robust demand for advanced GPUs, essential for training and operating increasingly sophisticated large language models and next-generation AI systems, shows no signs of waning. While the trend towards increasingly large models appeared to be shifting given the popularity of “reasoning” approaches and new applications of reinforcement learning to large language models, xAI has indicated that big GPU clusters could enable once fringe players to compete at the top tier. These high-volume orders highlight NVIDIA’s continued dominance in the AI hardware supply chain and suggest a strong near-term outlook, despite recent market jitters.

Palantir shares drop 10% on defense budget headwinds

In contrast to the buoyant trajectories of quantum computing and GPU sectors, Palantir’s stock price plummeted nearly 10% following reports of a potential 8% reduction in overall U.S. defense spending (See: The Washington Post, Bloomberg). With over 40% of its revenue derived from government contracts, Palantir faces heightened uncertainty as federal budgets tighten. Despite the Pentagon’s continued prioritization of AI initiatives, such as a recent $250 million AI research award, the specter of cost-cutting measures casts a shadow over the company’s growth prospects.

While AI hype has fueled Palantir’s stock in the past—helping drive a 410% surge over the previous year—the current 10% plunge does not appear to stem from AI whiplash. Instead, this drop is primarily linked to the newly reported annual 8% defense budget cuts over the next five years. Palantir’s heavy reliance on government contracts (over 40% from defense and intelligence) magnifies the impact of any Pentagon spending cuts. Compounding investors’ worries, CEO Alex Karp has reportedly filed to sell nearly 10 million shares, as Barron’s noted.

Why it matters: Palantir’s significant reliance on government contracts, while lucrative during periods of robust R&D spending, exposes it to vulnerabilities when defense budgets face cuts. While the Pentagon aims to enhance “efficiency” through AI, near-term market anxieties persist regarding Palantir’s capacity to sustain its growth trajectory given DOGE’s slash-and-burn approach to government funding allocations.