Mizuho raises AMD price target to $625 on Helios AI momentum

Mizuho Securities analyst Vijay Rakesh has lifted AMD's price target to $625 from $615, maintaining a Buy rating on the semiconductor maker following its Advancing AI event. The upgrade reflects growing confidence in AMD's AI strategy, particularly the Helios AI platform and the imminent deployment of its Venice server CPUs.

The Helios platform has already secured substantial commitments from leading generative AI companies. Rakesh disclosed that AMD holds agreements with OpenAI, Anthropic, Microsoft, and Meta Platforms for up to 14 gigawatts of MI455 Helios systems—a testament to the platform's appeal among the most capital-intensive AI developers. Beyond near-term revenue, Rakesh expressed bullish sentiment on AMD's long-term business opportunity, projecting that the company's total addressable market may reach $2 trillion（約320兆円） by 2030. That figure includes approximately $1.4 trillion（約220兆円） attributed to AI accelerators, alongside a higher server processor market than previously forecast.

AMD is also broadening its AI portfolio through collaboration. The company announced a partnership with Cerebras, a specialist in AI inference systems, to develop integrated high-throughput inference solutions that combine both firms' capabilities. On the server CPU front, AMD's sixth-generation EPYC Venice CPUs have entered full production. The company anticipates that major cloud providers and hardware vendors will commence deployment in the December quarter, positioning the new architecture to capture incremental demand from enterprise and hyperscale customers upgrading their infrastructure.

Despite recent cooling in semiconductor valuations, AMD stock has climbed 131% over the course of the year, reflecting the market's appetite for AI-infrastructure beneficiaries. Going forward, investors will focus on the pace of Helios system deployments, the speed of Venice CPU rollout into production environments, and the sustained trajectory of AI infrastructure spending across cloud and enterprise sectors.

AMD's Advancing AI event has rekindled investor confidence in the company's path to dominating AI infrastructure deployment. The Helios AI platform, which integrates custom accelerator systems, has already attracted commitments from the four largest generative AI companies—OpenAI, Anthropic, Microsoft, and Meta Platforms—for a combined 14 gigawatts of MI455 systems. This level of customer concentration among leading AI developers signals that AMD's architecture resonates with the companies driving the highest-value AI workloads.

Parallel to the Helios momentum, AMD's sixth-generation EPYC Venice server CPUs have reached full production and are poised for rapid deployment in major cloud and enterprise environments starting in the December quarter. The timing aligns with the accelerating buildout of AI infrastructure globally. Rakesh's projection that AMD's total addressable market may reach $2 trillion（約320兆円） by 2030—with AI accelerators accounting for $1.4 trillion（約220兆円） of that—reflects a substantial expansion of opportunity beyond the company's traditional CPU business. AMD's collaboration with Cerebras to develop high-throughput AI inference systems further diversifies its product surface and deepens its moat in the competitive AI accelerator space.

No comments yet. Be the first to share your thoughts!