Nvidia Expands AI Compute and Edge Partnerships with AWS and MediaTek

Nvidia (NASDAQ: NVDA), a core component of the Dow Jones Industrial Average, secured two material developments this week that reinforce its leadership across AI compute, networking, and edge applications.

AWS to Deploy Additional 2 Million Nvidia GPUs

On August 26, Nvidia and Amazon Web Services announced an expanded strategic partnership. AWS will deploy 2 million more Nvidia GPUs—specifically Blackwell Ultra, Rubin, and Rubin Ultra—across its global infrastructure in 2027–2028. Additionally, the collaboration will deepen across the full AI stack, extending into CPUs (Nvidia’s Vera), networking (Spectrum™), open models, and software. AWS has already begun offering EC2 G7 instances powered by Nvidia’s RTX PRO 4500 Blackwell Server Edition, delivering up to 4.6x AI inference and 2.1x graphics performance compared to prior G6 models.

This move follows AWS’s earlier announcement at GTC 2026 to add over 1 million Nvidia GPUs beginning in 2026—indicating customer demand on AWS for Nvidia-accelerated infrastructure continues to accelerate.

These enhancements underscore Nvidia’s position as a foundational supplier for hyperscale AI infrastructure. The broader integration—from GPU compute to networking and CPU support—positions AWS to deliver more comprehensive, co-engineered AI solutions.

MediaTek Adopts NVLink Fusion with Nvidia Convertible Bond Investment

Also this week, Nvidia announced a $3.5 billion investment in convertible bonds issued by MediaTek. In parallel, MediaTek agreed to adopt Nvidia’s NVLink Fusion platform for custom AI accelerators, local AI systems, and automotive applications.

This expands an existing partnership between the two companies, enabling MediaTek’s custom XPUs to integrate seamlessly into Nvidia’s NVLink-connected, rack-scale AI factories. The effort spans edge-to-cloud AI infrastructure development, reflecting Nvidia’s push into heterogeneous and automotive compute markets.

Significance for Nvidia and Broader AI Infrastructure

Both announcements provide Nvidia with substantial growth drivers in core products and markets that anchor its compute, networking, and edge strategy:

• Cloud hyperscale demand: The AWS deal alone promises to boost revenue from AI datacenter deployments, with millions of additional Nvidia GPUs entering production-grade usage environments over the next two years.
• Edge and automotive push: The MediaTek partnership opens the door to Nvidia’s technologies in automotive and local AI systems, offering NVLink connectivity beyond traditional datacenter applications.
• Stack-level integration: Collaborating across GPUs, CPUs, networking, and software consolidates Nvidia’s role as a vertically integrated vendor for AI infrastructure.

These developments arrived in the same week—the AWS expansion announcement on August 26 and the MediaTek investment around September 1—demonstrating strong momentum in Nvidia’s strategic expansion across data-center, cloud, edge, and automotive domains.

Next Steps and Investor Considerations

Investors should monitor AWS’s rollout of the 2 million GPUs and how these deployments translate into order backlogs and revenue. Similarly, tracking early NVLink Fusion implementations with MediaTek could signal how quickly Nvidia’s edge compute strategy is gaining traction.

Nvidia’s recent live share price of $230.36 (as of September 4, 2026) reflects general market positioning prior to these developments. Any future stock movement should be analyzed in light of tangible revenue or deployment confirmations tied to these agreements.

In summary, these validated deals with AWS and MediaTek represent direct and material advances in Nvidia’s compute, networking, and edge infrastructure strategy—reinforcing its central role in fueling modern AI capabilities.