Advanced Micro Devices (AMD) investors have had little to celebrate this month.
The chipmaker’s shares fell 4.86% on June 10 to close at $452.40, stretching the loss over the past five trading days to 12.15%.
The selling was not an AMD problem, as a wave of profit-taking and Middle East tensions hit nearly every major semiconductor stock.
While that selloff played out, AMD CEO Lisa Su stood on a stage in London delivering an interesting message.
She announced a commitment that ties AMD to one of the world’s most ambitious national AI programs for the next five years.
Speaking at London Tech Week on June 8, Su said AMD will invest up to £2 billion, about $2.7 billion, in the United Kingdom over the next five years, according to the company’s press release, also carried by the Globe and Mail.
More AI Stocks:
The money funds advanced computing, scientific research, and workforce training, with AMD’s Instinct GPUs and EPYC processors at the center of each project.
AMD CEO Lisa Su announced a £2 billion, five-year UK investment at London Tech Week on June 8.
NurPhoto &sol Getty Images
The bet centers on sovereign AI, which means a country building and controlling its own AI computing infrastructure instead of renting it from foreign cloud providers.
Governments now treat that capacity as a strategic resource.
The UK government has made the buildout an official policy through its AI Opportunities Action Plan, and AMD’s investment plugs directly into that program.
Lisa’s pledge in London is also indicative of a clear AMD pattern.
Just five days earlier, AMD announced a quantum computing research collaboration with Oxford Quantum Circuits and JPMorganChase in the UK, a sign that the company is methodically stacking British partnerships.
AMD was not the only one to make a financial commitment.
On June 8, Cloud provider Nebius pledged about £1.7 billion to expand UK AI capacity built on Nvidia hardware, according to the company’s newsroom, so competition for Britain’s AI budget is already fierce.
The timing was rough.
Semiconductor stocks shed a combined $1 trillion in market value on June 5 after Broadcom’s guidance disappointed investors, and renewed conflict between Israel and Iran knocked the Nasdaq lower again this week.
AMD’s UK announcement briefly broke that pattern.
Related: AMD CEO Lisa Su drops rare message on AI careers
AMD shares rose in premarket trading on June 8, Stocktwits noted, as the Nasdaq staged a small comeback.
The stock was up 1.94% at $461.18 in the premarket session of Thursday, June 8, a sign that buyers are stepping back in.
Even after the pullback, AMD has crushed the broader market:
The £2 billion commitment is not, on its own, sufficient to significantly improve AMD’s near-term income statement.
The company generated $10.25 billion in revenue in the first quarter of 2026, up 37.85% from a year earlier, so the UK plan averages out to a small fraction of one quarter’s sales per year.
For shareholders to feel the payoff, the Cambridge systems need to come online and prove AMD’s chips can run national-scale science.
The Oriole project must show that photonic networking can cut the energy cost of inference, the operational phase where trained AI models answer real requests.
Most importantly, those research developments need to convert into commercial orders across Europe.
In the short run, AMD stock will keep trading on the chip selloff, Middle East headlines, and the ramp of its MI450 GPUs rather than on a five-year government program.
However, the £2 billion commitment still matters.
Governments are becoming a new class of AI customers, and AMD has positioned itself as the leading chip supplier for one of the largest national AI programs outside the United States, reducing its dependence on Big Tech buyers.
The risks are also real.
Nvidia hardware is competing for the same UK budgets through deals like the Nebius buildout, and the broader market just erased 12% of AMD’s value in a week.
But for patient investors, Su’s London move widens where AMD’s next decade of growth can come from.
Related: Anthropic sounds the alarm on AI danger


