If you have asked Airbnb’s help bot to sort out a booking, part of the answer may have come from a Chinese model. Airbnb Inc. (ABNB) CEO Brian Chesky told Bloomberg in October 2025 that his company relies heavily on Qwen. The model, built by Alibaba Group Holding Ltd. (BABA), is “fast and cheap,” he said, as Fortune reported.
Intel Corp. (INTC) CEO Lip-Bu Tan just explained why choices like that are spreading. At a panel hosted by lender DBS in Singapore on Monday, October 12, 2026, he called Chinese models like DeepSeek, Kimi and GLM “good enough.” They cost less per token and are being adopted quickly, he said, according to Reuters.
Tan also said quantum computing could become mainstream within three to five years. That is a strikingly short window for a technology many investors still file under someday.
Read together, the remarks sound less like a CEO praising rivals and more like one describing his market. Cheaper models spread AI work across more hardware, and much of it lands on processors Intel sells.
Also read: Elon Musk just handed Intel a new problem Wall Street can’t ignore
Cheap tokens create more work for Intel’s chips
Lower prices rarely shrink demand for computing. When each answer from a model costs less, companies build AI into more products, and total usage climbs. Tan made a similar point in July, saying AI is driving “unprecedented demand for compute.”
Intel’s latest results show that pull. Data center and AI revenue jumped 59% to $6.3 billion in the second quarter, according to Intel’s earnings release. The release highlighted Xeon systems built for inference and agentic workloads, the everyday running of AI rather than training it.
AI agents push that further. Each agent coordinates tasks, moves data and calls tools, much of it on CPUs rather than GPUs. Low-cost open models make agents cheaper to deploy, which multiplies that background work.
China is Intel’s customer, and Washington DC its owner
The China angle cuts both ways. China generated about $12.7 billion of Intel’s $52.9 billion in 2025 revenue, based on billing location, according to its annual report. That is nearly a quarter of sales, tied to the same market whose models Tan praised.
Server processors also largely sit outside U.S. export curbs on advanced AI chips. Chinese buyers have been signing longer purchase agreements for them as prices climb. Cheap domestic models running on available server chips create a setup that plays to Intel’s strengths.
However, the politics are delicate. The U.S. government took a 9.9% stake in Intel for $8.9 billion in August 2025, according to Intel. Two weeks earlier, President Donald Trump called Tan “highly conflicted” over his China ties, Al Jazeera reported.
Public praise for Chinese AI from that CEO will not go unnoticed in Washington. Investors should treat it as a political risk worth tracking, not a footnote.
Tan’s quantum clock runs far faster than Nvidia’s
The quantum remark deserves more attention than it got. In January 2025, Nvidia Corp. (NVDA) CEO Jensen Huang said that very useful quantum computers were likely 15 to 30 years away, CNBC reported. Quantum stocks plunged, and Huang later softened his view.
Intel’s own program offers few public clues. Its latest disclosed quantum chip, Tunnel Falls, dates to 2023, according to Network World. Still, its venture arm, Intel Capital, joined a $178 million funding round for QuantWare, a quantum processor startup, in May 2026.
That makes Tan’s short timeline look more like a bet on software than on Intel’s hardware. He tied quantum’s progress to software that uses infrastructure more efficiently, the same logic behind his support for open models.
More Intel:
- Top analyst sets outrageous Intel stock target ahead of earnings
- Intel stock jumps as SK Hynix weighs U.S. memory-chip deal
- Intel stock surges 9% as surprising turnaround takes shape
Wall Street sees the demand but splits on value
Analysts lean positive but cautious. Intel carries a consensus Buy rating from 49 analysts with an average price target of $118.05, according to StockAnalysis. Still, 32 of those analysts rate the stock Hold.
As of Monday, the shares traded roughly 29% below their 52-week high of $142.35, per StockAnalysis, after climbing from a low of $32.89. That gap suggests investors want proof that hot demand turns into durable profit.
The latest calls show caution:
- Hold, $115 target: Susquehanna’s Christopher Rolland, Monday, October 12, 2026.
- Buy rating maintained: Bank of America’s Vivek Arya, Monday, October 12, 2026.
- Hold, $114 target: Mizuho’s Vijay Rakesh, Tuesday, October 6, 2026.
- Hold, $110 target: Bernstein’s Stacy Rasgon, Monday, September 21, 2026.
The bull case is that cheaper models carry AI into more apps, often through models users have never heard of. Each new deployment adds demand for processors Intel is racing to supply. The bear case is that Intel’s China exposure becomes a political liability while rivals chase the same CPU demand.
AI’s next phase may be decided less by who builds the smartest model than by who runs millions of adequate ones cheaply. Tan is betting Intel sells the hardware for that phase.