Advanced Micro Devices delivered another powerful signal that the artificial intelligence infrastructure race is broadening beyond Nvidia, forecasting second-quarter revenue above Wall Street expectations as surging demand for AI chips fueled explosive growth in its data-center business.

The upbeat guidance sent AMD shares soaring more than 12% in extended trading on Tuesday, extending a rally that has already pushed the stock up roughly 65% this year and cemented its status as one of the market’s strongest AI beneficiaries.

The earnings reinforced a growing investor belief that AMD is emerging as the most credible large-scale challenger to Nvidia in the AI semiconductor market, particularly as hyperscalers, governments, and enterprises race to expand computing infrastructure amid an unprecedented global spending boom tied to generative AI.

AMD forecast second-quarter revenue of about $11.2 billion, plus or minus $300 million, comfortably ahead of analyst estimates of $10.52 billion. The company also projected adjusted gross margins of roughly 56%, topping expectations and signaling improving profitability as higher-value AI products increasingly dominate its sales mix.

The strongest growth came from AMD’s data-center division, which includes both AI GPUs and server CPUs. Revenue in the segment surged 57% year over year to $5.8 billion during the first quarter, beating analyst expectations of $5.64 billion.

Chief Executive Lisa Su said demand for server CPUs is accelerating faster than previously expected as AI adoption moves beyond training large models toward inference, the stage where AI systems process real-world tasks and user requests continuously.

That shift is becoming one of the most important structural changes in the semiconductor industry. For years, investors focused primarily on graphics processing units used to train massive AI models. But inference workloads, which require enormous computing power across cloud platforms, enterprise software systems, search engines, robotics, and AI assistants, are now opening a much broader market for central processing units.

AMD now expects the addressable market for server CPUs to grow more than 35% annually and exceed $120 billion by 2030, sharply above the 18% annual growth estimate it provided only months ago.

The revised outlook underscores how quickly the AI economy is evolving from experimentation into large-scale deployment.

“AMD is levered to insatiable AI compute demand,” said Jake Behan, head of capital markets at Direxion. “The focus now shifts to how efficiently they can convert that into high-margin revenue.”

The company’s rapid rise is also reshaping competitive dynamics across the semiconductor sector. While Nvidia remains dominant in AI accelerators, AMD has gained traction with hyperscalers seeking alternatives amid supply constraints, pricing concerns, and growing fears over dependence on a single vendor.

AMD has aggressively deepened partnerships across the AI ecosystem. Earlier this year, the company agreed to supply up to $60 billion worth of AI chips to Meta Platforms over five years in a deal that also gives Meta the option to acquire as much as 10% of AMD. The company has also secured major agreements with OpenAI as competition intensifies over AI infrastructure dominance.

The agreements highlight how cloud providers and AI developers are increasingly diversifying away from Nvidia-only architectures to reduce costs, improve supply security, and gain negotiating leverage.

Additionally, AMD’s growth is intensifying pressure on Intel, which is attempting a major comeback after years of manufacturing delays and market-share losses. Intel recently issued stronger-than-expected revenue guidance of its own and is rapidly expanding domestic manufacturing capacity as demand for CPUs rebounds in the AI era.

Unlike AMD, which relies heavily on Taiwan Semiconductor Manufacturing Company for production, Intel designs and manufactures chips internally, giving it greater control over capacity during a period of mounting global supply-chain strain.

That manufacturing advantage is becoming increasingly important as the semiconductor industry confronts the tightening availability of advanced components, particularly high-bandwidth memory chips used alongside AI processors in data centers.

The global memory shortage is now emerging as one of the biggest risks to the broader AI expansion cycle. Prices for high-bandwidth memory have surged as hyperscalers scramble to secure supply for next-generation AI systems. Executives across the chip industry warn that the shortages could eventually spill into consumer electronics markets, driving up costs for PCs, gaming devices, and enterprise hardware.

AMD already expects PC shipments to weaken in the second half of the year due to higher memory and component costs. The company also forecast gaming revenue would decline more than 20% in the second half compared with the first six months of the year.

For investors, AMD’s latest results support the idea that the AI spending boom remains far from peaking. Major cloud companies, including Amazon, Microsoft, Alphabet, and Meta, are collectively expected to spend more than $700 billion this year on AI-related infrastructure, including chips, servers, networking equipment, and data centers. That spending surge has triggered one of the most aggressive capital expenditure cycles in technology history, lifting demand across nearly every layer of the semiconductor ecosystem.

AMD’s performance also signals that investors are beginning to reward companies positioned not only in AI training, but also in inference and enterprise deployment, areas expected to generate recurring long-term revenue as AI applications become embedded across industries.

While Nvidia still dominates the high-end AI accelerator market, AMD is increasingly carving out a powerful position as a second major pillar of the AI hardware economy. The next phase of competition may depend less on raw computing performance and more on who can secure enough manufacturing capacity, memory supply, and enterprise partnerships to sustain growth.

Investor Michael Burry has exited his entire position in GameStop and sharply increased bearish wagers against some of the market’s biggest artificial intelligence and semiconductor names, arguing that excessive leverage, speculative valuations, and euphoric investor sentiment are creating conditions similar to past market bubbles.

The investor, whose successful bet against the U.S. housing market collapse was chronicled in the book and film The Big Short, disclosed the moves in a series of posts published Monday on his Substack.

Burry said he abandoned GameStop after concluding that chief executive Ryan Cohen’s proposed $56 billion bid for eBay fundamentally undermined what he had previously described as the company’s “Instant Berkshire” strategy, a vision in which GameStop could evolve into a diversified capital allocator modeled after Berkshire Hathaway.

“I sold my entire GME position,” Burry wrote, describing it as the first full position he has exited since shifting away from running a hedge fund toward publishing investment commentary online late last year.

“Any which way I sliced it, the Instant Berkshire thesis was never compatible with >5x Debt/EBITDA, never ok with interest coverage under 4.0x,” he added.

Burry had previously ranked GameStop among his three largest holdings alongside JD.com and Molina Healthcare, together accounting for more than one-third of his disclosed personal equity portfolio.

His latest comments underscore mounting concern among some investors that GameStop’s attempted transformation under Cohen is veering away from the balance-sheet discipline that helped fuel enthusiasm around the stock following its meme-stock resurgence years ago.

Burry argued that the proposed eBay acquisition could saddle the combined company with dangerously high leverage at a time when borrowing costs remain elevated and consumer-facing retail businesses are already under pressure from slower spending and intensifying competition from giants such as Amazon.

According to Burry’s calculations, the proposed deal would leave the merged entity carrying a net-debt-to-profit ratio of roughly 5.2 times, while annual profits would cover interest expenses by only 2.5 times. He warned that if eBay demanded a higher price closer to $65 billion, leverage could climb to 7.7 times, with profit-to-interest coverage falling toward distressed territory.

“Leverage above 5 times was a knife edge,” Burry wrote in an earlier post. “A level of debt that borders on distressed.”

He dismissed the acquisition strategy as unoriginal and financially dangerous.

“If GameStop wants to do it with billions of interest expense and all manner covenants restricting its movements, it will not be breaking new ground,” he wrote previously. “It will be trotting in well-worn ruts on the road to capitalist Hell.”

The comments represent one of the harshest public critiques yet of Cohen’s attempt to reposition GameStop beyond its struggling core video game retail business. Cohen, who built his reputation through pet products retailer Chewy, has increasingly sought to portray GameStop as a long-term investment vehicle rather than a traditional retailer.

Also, Burry significantly expanded his bearish bets against the artificial intelligence sector, targeting companies that have become major beneficiaries of Wall Street’s AI-driven rally. He disclosed an outright short position in Palantir Technologies ahead of the company’s earnings release, saying his concerns extended beyond valuation alone.

“I am shorting the business model. I am shorting the entire premise upon which the company rests. I am shorting the CEO,” Burry wrote.

Palantir has emerged as one of the biggest winners of the AI boom, surging roughly 800% since the start of 2024 and reaching a market capitalization of about $350 billion. The company has benefited from investor optimism surrounding government contracts, defense applications, and demand for AI-powered analytics platforms.

But Burry’s attack highlights a widening divide between bullish investors betting AI will transform enterprise software and skeptics who believe valuations have detached from underlying fundamentals.

The investor also increased put-option positions against the SOXX, which tracks major chipmakers including Nvidia, Advanced Micro Devices, Broadcom, Micron Technology, and Intel.

“What a rally, a true cherry on top of a decade+ explosive run,” Burry wrote. “I am happy being short every single one of those names at current valuations and at this stage of the cycle.”

He also disclosed fresh bearish positions against the QQQ, which tracks the tech-heavy Nasdaq-100 index, as well as new put options on Nvidia.

“Nvidia remains one of the cheaper ways to short the AI data center bubble because of the continued near-unanimous positive view of Wall Street,” he wrote.

Burry said the options were structured with strike prices well below current trading levels and expirations extending into next spring, signaling expectations of a potentially sharp correction in AI-related equities over the coming months.

Combined, his bearish options positions tied to SOXX, QQQ, Palantir, Nvidia, and Oracle account for nearly 7% of his portfolio, while direct short positions in Palantir and Tesla make up another 2.5%.

The moves come at a time when enthusiasm surrounding AI infrastructure, data centers, and semiconductor demand has propelled technology stocks to record highs, even as some analysts warn that expectations are beginning to resemble prior speculative manias.

Burry’s latest positioning suggests he believes the market’s AI euphoria, much like the housing bubble he famously bet against nearly two decades ago, is entering a more dangerous late-cycle phase where lofty narratives are overwhelming traditional valuation discipline.