AMD Crosses $1 Trillion as OpenAI and Oracle Deals Reshape AI Chip Market
AMD shares topped $620, lifting its valuation past $1 trillion, as an OpenAI deal for up to six gigawatts and Oracle's 50,000-chip MI450 order signal durable gains in AI computing.
By Nathan Brooks
4 min read
Updated

What's News
- AMD shares surpassed $620, pushing market valuation above $1 trillion, up from roughly $127 billion five years ago when the stock traded near $104
- AMD posted Q2 revenue of $11.54 billion, up 50.1% year over year and beating consensus of $11.31 billion, with data center revenue growing more than 100% year over year
- OpenAI agreed to up to six gigawatts of AMD compute and Oracle ordered 50,000 Instinct MI450 accelerators; management expects data center revenue could double by 2027
AMD shares pushed past $620 and carried the company's market valuation above $1 trillion, capping a multi-year turnaround that has transformed the chipmaker from a $127 billion also-ran into one of the world's most valuable hardware designers.
Five years ago, Advanced Micro Devices (NASDAQ: AMD) traded near $104. Today, the stock's re-rating reflects a structural shift in how big tech buys silicon: hyperscalers and AI labs no longer want to depend on a single supplier for high-performance computing. Enterprise customers, wary of allocation bottlenecks and rising capital expenses under NVIDIA's dominance, have deliberately cultivated AMD as a second pillar for advanced accelerators.
Big Tech Hands AMD the Keys
The commercial commitments are the clearest evidence. OpenAI reached an agreement for up to six gigawatts of AMD compute systems — power consumption comparable to millions of homes, and a deep vote of confidence in the underlying architecture. Oracle Corporation (NYSE: ORCL), meanwhile, ordered 50,000 next-generation Instinct MI450 accelerators for Oracle Cloud Infrastructure. These wins underpin management's expectation that data center segment revenue could potentially double by 2027.
The numbers are already showing up in results. AMD posted second-quarter revenue of $11.54 billion, up 50.1% year over year and beating consensus estimates of $11.31 billion. Diluted earnings per share reached $1.66, topping expectations by 4 cents. Data center revenue grew more than 100% year over year, confirming that enterprise buyers are committing serious capital to AMD's hardware ecosystem.
The milestone also signals a change in perception. AMD's Instinct accelerators and EPYC server chips are no longer treated as an experimental backup plan. Tier-one cloud providers are designing multi-year data center roadmaps around them — a shift that raises the ceiling for the entire sector and shows expanding cloud budgets can sustain two profitable market leaders at once.
Turning Chip Shortages Into Cash Flow
Revenue growth only creates lasting shareholder value if it converts into operating earnings. AMD plans a roughly 10% price increase on select high-performance computing chips. Raising prices in the middle of an industry buildout signals genuine pricing power, and the increase should help absorb higher manufacturing costs from semiconductor foundries while expanding operating margins and supporting EBITDA.
That margin expansion matters because AMD's valuation multiples are stretched: a forward price-to-earnings ratio near 96 and a trailing P/E of about 159. Stronger cash generation would compress those multiples over time.
The balance sheet underneath is conservative. AMD carries a debt-to-equity ratio of just 0.03, a current ratio of 2.61 and a quick ratio of 1.91. That liquidity funds complex chip tape-outs and packaging designs without expensive debt, and leaves room to return capital through an active $6 billion share repurchase program.
Pairing EPYC Server Brains With Instinct Brawn
A common Wall Street assumption holds that the AI buildout belongs solely to graphics processors. Real-world data center workloads tell a different story. When Meta Platforms (NASDAQ: META) launched its Muse AI agent, the deployment highlighted the need for balanced computing clusters — agent-driven software models lean heavily on host CPUs to direct traffic, manage inputs and feed data into GPUs.
That dual demand plays directly to AMD's portfolio. While NVIDIA (NASDAQ: NVDA) pushes proprietary processors into the server space, AMD already has an established footprint in high-performance server CPUs with its EPYC family. Pairing EPYC processors with Instinct accelerators gives AMD a complete enterprise computing stack.
Software barriers are falling too. Industry adoption of open-source frameworks like PyTorch and OpenAI Triton means developers are less tied to proprietary standards like NVIDIA's CUDA, and tools such as AMD Inference Microservices make it easier for engineers to move workloads between hardware vendors.
Calculating Value in a $1 Trillion Market
The risks are real. AMD's P/E is elevated, and a beta of 2.48 suggests shares will swing harder than the broader market. Regulatory disclosures show insiders, including CEO Lisa Su, have trimmed portions of their personal holdings into the recent advance.
Wall Street is split. The average 12-month analyst target sits around $565, implying a modest pullback from recent peaks. Bulls see targets as high as $1,250 if multi-gigawatt orders ramp up over coming quarters.
The data suggests the semiconductor market is not winner-take-all. Rising enterprise demand is expanding the total addressable market for both leaders. Investors should watch how upcoming quarterly reports reflect MI450 shipment timelines and data center operating margins — while cautious buyers may wait for pullbacks toward the 50-day moving average near $494 before building positions.
Original: marketbeat.com
More from Nathan Brooks
Show full bio
News editor covering marketplaces and e-commerce at Business Bearings.
242 articles