When Lip-Bu Tan took over Intel’s leadership team in March 2025, the company was in dire need of a turnaround.
The Silicon Valley-based semiconductor pioneer has struggled for years to maintain its market-leading position as a chipmaker, losing ground to Nvidia, AMD and Qualcomm. Meanwhile, Taiwan Semiconductor Manufacturing Company (TSMC) produces 90% of the world’s most advanced chips, leaving little room for Intel’s own foundries to compete.
The issues plaguing Intel were not lost on Tan when he made his first public comments during the company’s quarterly earnings call. “There are areas that need improvement, but there are no quick fixes,” Tan said in April 2025.
Mr. Tan, 66, is a Malaysian-born executive who spent more than a decade at the helm of Cadence Design Systems, but has kept a low profile since taking over at Intel. So when Mr. Tan gave a public keynote speech last week at Computex, an annual trade show that has grown in popularity due to the AI boom in Taipei, a packed crowd turned out to hear him speak.
The crowd was probably there to learn more about Intel’s core product, the central processing unit (CPU), the chip that acts as the brain of the computer. These chips have been powering laptops and servers for decades.
But now CPUs are seen as essential to the AI race and are receiving new attention, giving Intel new opportunities for rebirth.
“CPU revival could save the company,” Dan Nested, vice president of TriOrient, an Asia-based private investment firm, told CNN. “The bulk of[Intel’s]business is CPUs.”
Intel said in a statement to CNN that Tan is focused on execution and putting customers back at the center of everything it does. “With a strengthened balance sheet, a refreshed leadership team, and a renewed focus on engineering, Intel is well-positioned to seize the AI-driven growth opportunities before us,” the company said.
Agent AI could be a lucky boost for Intel’s turnaround thanks to a new generation of AI that can go beyond simply answering questions and complete tasks on behalf of humans.
Graphics processing units (GPUs), the type of chip that Nvidia specializes in, are ideal for training AI models. But demand for CPUs is surging because they help with inference, the process of making trained AI models work. Inference is key to powering the new wave of AI agents being developed and deployed by companies like OpenAI, Anthropic, and Google.
OpenClaw independently interacts with apps, web browsers, and smart home appliances based on commands sent by users through messaging apps like WhatsApp. It is currently one of the most popular AI agents.
“The CPU is now the conductor and the GPU is the orchestra,” Nvidia CEO Jensen Huang said in a keynote speech in Taipei this week ahead of Computex. Nvidia announced a big push into CPUs during the conference, introducing new chips for laptops and desktops and revealing that its Vera data center CPUs are now running at full capacity.
Tan told reporters on the sidelines of Computex that over the past month he has received calls from many CEOs asking for more CPUs.
“This is a big opportunity for us,” Mr Tan said. “We were really excited.”
Analysts say that if Intel can ramp up production quickly and ensure manufacturing quality, demand for CPUs will increase and sales could increase in the coming years.
“We are very focused on supply and strive to delight our customers,” Mr Tan added.
Tan’s strategy at Intel appears to be a direct continuation of his strategy at Cadence. At Cadence, he streamlined operations, invested heavily in engineering, expanded the company through strategic acquisitions, and built deep customer relationships.
Mr. Tan laid off about 34% of Intel’s workforce and suspended plans to expand manufacturing facilities in Germany and Poland. He also flattened the company’s bureaucracy, brought in new talent, formed new partnerships, sold non-core business assets and returned the company’s focus to its roots.
“At our core, Intel is an engineering company,” Tan said in his Computex keynote. “And that’s what I decided from day one…I have all the engineering reports.”
Tan has promoted engineers, cut middle management positions and given Intel more direct oversight of its chip and AI businesses. In exchange for liquidity, he sold control of an Intel subsidiary and hired senior executives from Qualcomm and Arm to lead the data center and AI business units. Mr. Tan also introduced investments from Nvidia and Softbank, and secured strategic support and partnerships from influential companies in the AI industry.
“Tan is trying to bring in more outside resources to Intel to not only adjust the chip manufacturing process but also give Intel the time and space to bring in more customers and increase production,” Chia-Yang Yao, an analyst at Taiwanese market research firm Digitimes Research, told CNN.
Meanwhile, Intel is receiving aid from the Trump administration.
In August, the U.S. government invested $8.9 billion in Intel stock, representing approximately 10% of Intel, to support the company’s continued efforts to expand its research and manufacturing facilities in the United States while ensuring access to a domestic advanced chip supply chain for national security purposes. Since then, Intel (INTC) stock has increased nearly 300%.
The agreement was part of an effort to boost U.S. semiconductor manufacturing, a key priority of President Trump’s second term, and solidify the country’s position as a leader in the global semiconductor manufacturing industry.
But analysts say Intel still faces major challenges, particularly in its chip-making foundry business, which is struggling to add customers and improve production quality.
Still, it was a much-needed breakthrough for the company, which has been in decline for years.
“(Tan) was able to successfully stop the bleeding and was transferred from the intensive care unit to the general ward,” Yao said.
CNN’s Clare Duffy and Lisa Eadicicco contributed to this report.
