At CES 2026, AMD CEO Lisa Su detailed how artificial intelligence is influencing hiring and hardware development across the company.
While concerns over automation and job losses have gained attention, Su explained that AMD is continuing to grow its global workforce. The company is placing greater importance on candidates who can apply AI tools across chip design, testing, and manufacturing.
Su also introduced several new products designed to meet expanding demand for AI compute power.
Among the launches were the MI440X processor, the upcoming Instinct MI500 GPU series, and Helios, a rack-scale system with multi-exaflop capacity. She also announced desktop and consumer-grade AI PCs built with Ryzen chips.
Why It Matters: At AMD, hiring is focused on roles that contribute directly to AI workflows, particularly where tools are applied to product development and deployment. This approach mirrors the company’s hardware strategy, which centers on delivering systems built to support real-world AI use.
- Hiring Expands with Emphasis on AI Proficiency: AMD is continuing to grow its workforce, with an emphasis on candidates who can apply AI tools in real-world development environments. Roles are concentrated in areas such as chip design, manufacturing, and quality control. As of late 2024, the company employed around 28,000 people, with expectations for that number to increase as AI-related projects grow.
- AI Tools Are Enhancing Engineering and Production Workflows: Across AMD’s operations, AI is being used to support tasks typically handled by engineers and technicians. In chip development, AI assists with simulations and layout refinement. In manufacturing, it contributes to diagnostics and testing processes. These applications are helping teams manage greater workloads while maintaining precision and reducing delays.
- Global Compute Demand Is Quickly Climbing: Su reported that global AI compute capacity grew from 1 zettaflop in 2022 to over 100 zettaflops by 2025. By 2031, that number is expected to exceed 10 yottaflops. One yottaflop equals one septillion operations per second. These figures are guiding AMD’s new system designs, which aim to handle much higher levels of computing demand than today’s infrastructure can support.
- Helios Introduced as High-Capacity Data Center System: Designed for large-scale AI environments, the Helios rack system will deliver up to 3 exaflops of AI performance per unit. It is scheduled for release later in 2026 and is intended for organizations running intensive workloads, such as AI model training and high-volume inference tasks.
- AI Hardware Launches Extend to Developers and Consumers: In addition to data center products, AMD introduced new devices for everyday users and developers. PCs with Ryzen AI 400 series chips will be available this month. These machines can run AI tasks directly on the device. A more powerful desktop called Ryzen AI Halo is coming in the second quarter of 2026. It’s built for developers who need to build and test AI models on their own machines.
Go Deeper -> AMD’s Lisa Su says AI isn’t replacing people, but is changing who gets hired – CNBC
AMD Promotes New AI Chip, Sees Industry Demand Reaching Yottaflops – Investors Business Daily
Trusted insights for technology leaders
Our readers are CIOs, CTOs, and senior IT executives who rely on The National CIO Review for smart, curated takes on the trends shaping the enterprise, from GenAI to cybersecurity and beyond.
Subscribe to our 4x a week newsletter to keep up with the insights that matter.
