Lisa Su

Who they are

Lisa Su is Chair and CEO of AMD — she joined in January 2012 when the company carried approximately $2.2 billion in debt and its stock sat near $1.61, and has since taken it to a market cap exceeding $550 billion.

Person

Lisa joined AMD in January 2012 when the company was carrying approximately $2.2 billion in debt, had posted four consecutive years of losses, and its stock was trading near $1.61. She came in as SVP and GM of Global Business Units, moved to COO, and was named President and CEO in 2014 — stepping into a company many had written off. Before AMD she was SVP and CTO at Freescale Semiconductor, and before that spent years at IBM where she rose to VP of Semiconductor R&D and served as Technical Assistant to CEO Lou Gerstner — a formative early role that put her inside the strategic thinking of one of the industry's most consequential executives. Earlier still she was a Member of Technical Staff at Texas Instruments. The foundation under all of it is MIT: BS, MS, and PhD in Electrical Engineering, with her doctoral research focused on semiconductors. The through-line is deep technical credibility applied at larger and larger scale — from device physics to running a $34.6 billion revenue company. She speaks at forums where policy and hardware intersect: the U.S. Senate Commerce Committee in 2025 on AI leadership, the AI Action Summit in 2025, Davos in 2025 where she announced the Aleph initiative, and SC24 where she made HPC and AI hardware announcements. Her public voice is focused on AI infrastructure, open ecosystems, domestic semiconductor supply chains, and government AI policy — rarely personal, consistently substantive.

Company

The most live development: AMD announced a strategic partnership with Anthropic in July 2026 — committing up to $5 billion and supplying AI chips including the MI450 series, with deployment starting in the first half of 2027. That follows a February 2026 multi-year deal with Meta to deploy up to 6 gigawatts of AMD Instinct GPUs for AI infrastructure, with shipments starting in the second half of 2026, and a 2025 commitment to supply OpenAI with 6 gigawatts of Instinct GPUs over multiple years. AMD also launched Helios, its first rack-scale AI system, in mid-2026, targeting Nvidia's dominance in AI data center infrastructure — early customers include Microsoft, Meta, OpenAI, and Oracle. AMD reported record full-year 2025 revenue of $34.6 billion, with data center revenue hitting $5.38 billion in Q4 2025 alone, driven by EPYC server processors and Instinct GPUs; data center revenue is projected to reach $31.2 billion in 2026. In May 2026, AMD announced more than $10 billion in Taiwan ecosystem investments to accelerate AI infrastructure and ramp production of next-generation EPYC 'Venice' processors on TSMC 2nm technology.

Market

AMD competes primarily with Nvidia — which dominates AI GPU infrastructure through its CUDA ecosystem — and Intel, with Qualcomm and Samsung as secondary rivals. AMD holds roughly 35% of the global GPU market in 2026 and has positioned itself as a cost-effective alternative to Nvidia in AI data center build-outs, though Nvidia's entrenched software stack remains a structural advantage. The sector is shaped by two compounding pressures: surging AI infrastructure investment driving aggressive supply deals, and U.S. export controls on advanced AI chips — which triggered an $800 million charge for AMD in 2025 and continue to constrain market access in China in 2026.

Network

No direct edge data is available on Lisa Su's immediate network. Her public engagements signal deep relationships with hyperscaler and AI lab leadership — Meta, Microsoft, OpenAI, Oracle, and Anthropic are all named as customers or partners in recent AMD deals. She has also engaged directly with U.S. Senate Commerce Committee members and appeared alongside other semiconductor and AI executives at Davos and the AI Action Summit.

KC McClure· AMD Board of Directors (appointed February 2026)
Jean Hu· Chief Accounting Officer, AMD (appointed 2025)

How they likely show up

Long tenure at AMD (joined 2012, CEO since 2014, still Chair and CEO in 2026) → thinks in multi-year product and strategy cycles, not quarters.
Joined AMD as SVP/GM, moved to COO, then CEO — climbed every rung from within → likely values deep institutional knowledge and operators who can show their work, not just their deck.
PhD-level semiconductor research background, followed by VP of R&D at IBM → will engage at technical depth; surface-level product claims will land poorly.
Regular presence at Senate Commerce Committee hearings, Davos, and AI Action Summit → comfortable operating at the policy-industry interface; treats regulatory dynamics as a strategic variable, not a compliance nuisance.
Oversaw a strategic workforce reduction of about 4% in 2025 while continuing net hiring — deliberate portfolio reallocation from gaming to AI and data centers → signals willingness to make hard structural bets and hold them publicly.
Simultaneous multi-gigawatt supply deals with Meta, OpenAI, and Anthropic signed within a 12-month window → operates at a pace and deal scale that requires high decisiveness and tight alignment across product, supply chain, and finance.

Conversation tips

→ Reference the Anthropic deal specifically — it closed in July 2026 and is the freshest strategic signal; she'll be closer to it than to older announcements.
→ If you want to engage her on technical depth, ask about the MI455X or EPYC Venice on TSMC 2nm — these are the specific product bets she's staked the 2026 roadmap on.
→ Frame any conversation about competition around the software ecosystem gap with Nvidia, not just chips — that's the structural challenge she has to solve and she'll have a considered view.
→ She testified before the U.S. Senate Commerce Committee on AI policy in 2025 — if your conversation touches government or regulation, she has a real position, not just talking points.
→ Don't try to simplify the business to 'chips vs. chips' — she runs a system play (silicon, rack-scale Helios system, software, partnerships) and will distinguish quickly between someone who understands that and someone who doesn't.

Toolbox

Openers

Open on the Anthropic deal — AMD committed up to $5 billion and is supplying MI450 chips with deployment in H1 2027. It's the freshest strategic move and a signal that AMD is now competing for relationships with frontier AI labs, not just cloud hyperscalers.
Bring up Helios — AMD's first rack-scale AI system launched in mid-2026, with Microsoft, Meta, OpenAI, and Oracle as early customers. It's AMD's clearest statement that they're selling systems, not just components, and she drove that transition.
Reference her 2025 Senate Commerce Committee testimony on AI leadership — she went on record about semiconductor policy, domestic supply chains, and export controls. It shows you understand she operates at the policy level, not just the product level.

Discovery questions

The Helios rack system puts AMD in the business of selling AI infrastructure, not just chips — how does that change the sales motion and the relationship with hyperscalers who are also potential competitors?
Export controls cost AMD an $800 million charge in 2025 and continue to constrain China access in 2026 — how are you thinking about the long-term geographic shape of AMD's revenue?
The MI500 series is previewed for 2027 with up to 1000x AI performance increase over MI300X — at what point does raw compute stop being the differentiator against Nvidia, and what does the real battle shift to?

Avoid

Don't treat the AMD turnaround as a past achievement to be celebrated — she's running a present-tense strategic battle against Nvidia's entrenched CUDA ecosystem and will be more engaged by forward-looking specifics than by retrospective praise.

Make it yours