AMD (AMD) Price on Solana

Ben Reitzes, head of technology research at Melius, named AMD as one of two chipmakers that could reach a $2 trillion market capitalization, CNBC reported. In a client note on September 30, Reitzes said the case for AMD rests on the CPU becoming more important again in agentic AI, as those system designs call for more CPUs relative to GPUs. Micron was the other company named. AMD passed a $1 trillion market value for the first time earlier in September.

Reitzes also set out what AMD would need to show. To add its next trillion, he wrote, the company will likely have to make the case that earnings power of more than $50 per share is achievable. That is well above its current 2030 target of more than $20 in EPS. For Micron, he pointed to memory's growing role in AI efficiency and flagged the cyclical nature of the memory business as the main obstacle.

AMD has agreed to acquire World Labs, the AI startup founded in 2024 by Stanford computer scientist Fei-Fei Li, in an all-stock deal worth $8.2 billion, as reported by TechCrunch and CNBC on September 28. Li, best known for creating the ImageNet dataset, will join AMD as executive vice president and chief scientist. The deal is expected to close before the end of 2026, subject to regulatory approval.

World Labs builds "world models", deep learning systems designed to understand and reason about physical space, which are seen as a key layer for putting generative AI into robots, autonomous vehicles and other machines. Its first product, Marble, targets entertainment and simulated robot-training environments. The purchase fills a gap in AMD's AI stack against Nvidia, which already offers open-weight world models through Cosmos while AMD has so far offered only text and video models. The two companies already had an inference and training partnership, and Li demonstrated Marble at AMD's CES 2026 keynote. Li said the move was about scaling World Labs' research "beyond the lab" and "getting closer to the hardware."

Bank of America reiterated its buy rating on AMD on September 25 and raised its price target to $720 from $620, calling the chipmaker its "top CPU pick." The call appeared in CNBC's roundup of Friday's biggest Wall Street analyst calls.

The bank lifted its calendar 2027 and 2028 earnings estimates for AMD by 2-3%, pointing to "higher CPU ASPs from agentic mix," meaning it expects agentic AI workloads to push average selling prices for AMD's processors higher. The new target values the stock at 30 times estimated 2028 earnings, up from the 27 times multiple the bank used before.

Advanced Micro Devices crossed $1 trillion in market capitalization for the first time on September 22, 2026, as its stock reached a record intraday high of $613.92 — a gain of roughly 9% on the day and about 26% over the preceding five sessions. The milestone makes AMD the fourth U.S. chipmaker to join the trillion-dollar club, after Nvidia, Broadcom, and Micron. The rally reflects investor confidence in AMD's positioning as a full-stack AI infrastructure provider, combining processors with networking equipment and complete systems to compete for data-center and AI workload spend.

The broader semiconductor sector moved higher alongside AMD, with Intel rising 11%, Qualcomm up more than 4%, and the Philadelphia Semiconductor Index hitting a one-month high. Despite the milestone, AMD remains well behind AI chip leader Nvidia, whose market cap exceeds $5 trillion. AMD's path to $1 trillion comes after recovering from a disappointing quarterly revenue forecast earlier in the year, underscoring how quickly AI-infrastructure demand expectations can reshape valuations for chipmakers positioned in that market.

Unverified supply-chain reports, originally from WeChat account ChannelGate and picked up by Wccftech, indicate AMD has notified customers of roughly a 10% price increase across its chip portfolio beginning in Q4 2026. The affected products reportedly span AI accelerator chips, consumer graphics processors, motherboard chipsets, and potentially Ryzen CPUs. AMD has not officially confirmed the reports.

The reported hike tracks a separate announcement from TSMC, AMD's primary foundry partner, which has told customers to expect approximately 10% higher manufacturing quotes as its own production costs rise. AMD shares gained 1.4% in premarket trading on the news, extending a three-session rally, with the stock up roughly 154% year-to-date.

Nvidia's Grace CPU line has quietly become a significant business, generating $5 billion in trailing twelve-month sales through fiscal Q2 2027, with data center CPU revenue projected to more than double from fiscal year 2027 to 2028. The company's next-generation Vera CPUs are now in full production alongside the Vera Rubin GPU family, with Nvidia claiming the platform delivers performance and efficiency that "significantly exceed those of any other data center CPU in key areas."

The threat to AMD is structural rather than merely competitive. Nvidia's commanding position in AI accelerators — roughly 90% share in training and 60-75% in inference workloads — lets it design CPUs explicitly optimized for integration with its own GPUs. Hyperscalers and AI infrastructure buyers building around Nvidia GPUs face a natural pull toward Nvidia CPUs as well, eroding the server CPU market that AMD has spent years growing as a credible alternative to Intel.

AMD and Intel's jointly developed AI Compute Extensions (ACE), a matrix-acceleration architecture targeting AI workloads, reached an early milestone on September 2 when GCC (the GNU Compiler Collection) received an initial implementation of ACE support. The collaborative effort was shaped by the x86 Ecosystem Advisory Group, which outlined the common architecture in an April whitepaper. Standardized instructions could make it easier for developers to write software that runs efficiently across processors from both chipmakers, potentially broadening the addressable market for AI compute hardware.

Analysts caution that compiler support is an infrastructure step, not a product launch — no compatible processors have shipped and GCC 17 is not yet a finished release. Standardization also does not resolve the separate competitive questions of manufacturing execution, delivery timelines, and pricing. Whether the cooperation translates into meaningful revenue for AMD will depend on shipping hardware with convincing benchmarks and actual developer adoption of the shared instruction set.

AMD unveiled the Threadripper Halo Station on September 4, a workstation designed for AI development and inference that can run models exceeding one trillion parameters locally. The system combines Threadripper Pro processors, high-capacity memory, and Instinct AI accelerators to deliver up to 16.4 TB/s of total system memory bandwidth, carving out a new market segment between conventional PCs and data center servers. It competes directly with Nvidia's DGX Station and is seen as a vehicle to strengthen AMD's ROCm software ecosystem, which analysts describe as critical to narrowing Nvidia's longstanding CUDA advantage.

Analysts characterize the Halo Station as a long-term strategic opportunity rather than an immediate revenue catalyst. AMD has not disclosed pricing, but the high-end configuration targets AI researchers and specialized enterprise workloads rather than mainstream buyers. With a launch scheduled for 2027, investors may need to wait before the product registers a meaningful impact on AMD's financial results, leaving near-term upside dependent on its existing data center GPU and server CPU business.

AMD CFO Jean Hu disclosed at a September 2026 conference that the company's server CPU market opportunity could reach $220 billion by 2030, with growth exceeding 50% annually — a significant upward revision from guidance issued just six months earlier that pegged the same market at $120 billion with a 35% CAGR. Hu also cited a total addressable market of $2 trillion by 2030, reflecting management's view that AI-driven infrastructure buildout from hyperscalers is accelerating faster than previously modeled.

The update comes on the heels of AMD's Q2 2026 earnings, which showed revenue up 50% year-over-year to $11.5 billion and adjusted EPS up 246% from the prior year. AMD shares trade at roughly 33x forward earnings, a premium the company's trajectory has so far supported, though the revised TAM figures are management projections and the degree to which AMD captures that opportunity remains subject to execution risk and competition.