Intel CEO Says DeepSeek, Kimi, GLM Are 'Good Enough' – Reportedly Backs Open Source For Efficiency Gains

• Tan’s support for open-source AI echoes Nvidia CEO Jensen Huang’s statements that accessible AI models can broaden adoption rather than necessarily reduce demand for computing hardware.
• Intel’s data-center business could benefit if lower inference costs encourage more companies to deploy AI applications, including agentic AI systems.
• Intel reported second-quarter revenue of $16.13 billion, up 25% year over year, while its Data Center and AI segment grew 59%.

Intel (INTC) CEO Lip-Bu Tan reportedly said on Monday that Chinese frontier AI models cost less per token and that adoption has been “very fast.” 

According to Reuters, he called the performance of models like DeepSeek, Kimi and GLM “good enough.” He added that “I’m a strong believer of open source; that’s the best way to drive efficiency,” and said markets will decide which models suit which use cases.

INTC stock fell as much as 1.6% in pre-market trade, on track to hit lows last seen in mid-September. On Stocktwits, retail sentiment around the shares remained in ‘bearish’ territory over the past day. 

Intel CEO Sees Opportunity In Lower-Cost AI Models

Tan’s view puts him in the same cohort as Nvidia’s (NVDA) CEO Jensen Huang, who has repeatedly defended Chinese open-weight models. Huang said in July that models from DeepSeek, Alibaba (BABA) and Moonshot AI were “excellent” and should be used, adding that open models can broaden AI adoption rather than reduce demand for Nvidia chips.

For Intel, wider adoption of lower-cost AI models could create opportunities beyond the models themselves. Cheaper inference could result in wider deployment and, in turn, increase demand for the CPUs, servers and data-center infrastructure needed to run those workloads.

Tan has positioned Intel’s data-center business as a beneficiary of agentic AI, where software agents perform multi-step tasks and require substantial computing capacity.

In the second-quarter (Q2), Intel’s revenue rose 25% to $16.13 billion, with data center and AI up 59%. It guided third quarter (Q3) revenue to be around $15.8 billion to $16.8 billion and $0.38 in non-GAAP earnings per share (EPS).

Retail Traders Watch The $100 Level

INTC stock has more than doubled this year on demand for data center CPUs, which Tan has tied to agentic AI. However, the rally now faces a tougher backdrop. The iShares Semiconductor ETF (SOXX) fell about 5% last week, its worst weekly performance in a month. The drop came after reports that OpenAI’s annualized revenue was around $50 billion at the end of September.

Despite Intel’s strong gains this year, some retail traders on Stocktwits are watching for a further pullback, with most keeping an eye on the $100 level. 

Overall, U.S. semiconductor stocks were mixed in early morning trade on Monday, after a tough week for the sector, as concerns over OpenAI’s revenue outlook and intensifying competition in the memory-chip market weighed on investor sentiment.

The stocks were under pressure as memory-chip competition intensified. China’s ChangXin Memory Technologies (CXMT) reported progress in advanced DRAM, raising concerns for Micron (MU), SK Hynix (SKHY) and other chip makers. 

Read also: MAG 7 Voices: How Nvidia, Microsoft, Alphabet, Meta Reshaped AI This Week

For updates and corrections, email newsroom[at]stocktwits[dot]com.