Lip-Bu Tan, the legendary semiconductor executive and investor who transformed Cadence Design Systems from a struggling firm with stock below $3 to over $320 per share, is now strategically positioning himself in the AI chip revolution. Tan, whose influence spans across major chip industry players including Amazon’s Annapurna Labs, SoftBank, and Intel, is backing alternatives to NVIDIA’s GPU dominance through his investment firm Walden International, which manages over $1.5 billion in assets.
Tan’s investment strategy focuses on addressing NVIDIA’s primary weakness: power consumption. He personally invested in SambaNova Systems in 2018, a company claiming to deliver equivalent performance at one-tenth the power consumption of traditional GPUs. He’s also backing Rivos, a startup founded in 2021 that offers alternative computing architectures promising dramatic gains in power efficiency, speed, and cost-effectiveness.
Despite his close personal friendship with both AMD CEO Lisa Su and NVIDIA CEO Jensen Huang, Tan recognizes the market opportunity for NVIDIA alternatives. He recalls arguing with Huang years ago when the NVIDIA chief predicted his company would become a full-platform software and systems company rather than just a semiconductor firm. Tan now admits Huang was right, acknowledging that NVIDIA’s CUDA software and 70% profit margins make it resemble a software company more than a traditional chip manufacturer.
According to Tan, the opportunity for NVIDIA competitors lies in capturing 10-15% of different workloads where alternative solutions offer better performance. “Everybody is looking for an alternative. They are still going big time with Nvidia, but they are looking for 10%, or 15% of the different workloads that can use a better solution,” he explained. Beyond startups, Tan sees growth potential in established players like AMD, Broadcom, Marvell, and Micron Technology as compute demands expand.
Tan’s investment philosophy centers on proximity to data. “AI is already a 60-year-old technology. But really, the big difference is the data, the massive, massive data that is starting to become available,” he noted. He emphasizes that success in AI requires being close to data sources. Currently, Tan is heavily investing in healthcare AI, particularly in identifying biomarkers and medical drug discovery, where he sees the most significant opportunity for applied AI impact.
Key Quotes
I said, ‘No, you’re semiconductor company.’ And he said, ‘No, you’re completely wrong. I’m a software and system company.’
Tan recounts an argument with NVIDIA CEO Jensen Huang from years ago about NVIDIA’s future direction. This quote matters because it shows how Huang’s vision of NVIDIA as a software-first company proved prescient, with CUDA software now being a key competitive moat that gives NVIDIA 70% profit margins.
We can deliver the same performance at one-tenth of the power.
Tan describes the value proposition of SambaNova Systems, one of his portfolio companies. This quote highlights the critical weakness in GPU-based AI computing that creates opportunities for alternative architectures focused on power efficiency.
Everybody is looking for an alternative. They are still going big time with Nvidia, but they are looking for 10%, or 15% of the different workloads that can use a better solution.
Tan explains the market opportunity for NVIDIA competitors. This reveals that the path forward isn’t about displacing NVIDIA entirely, but capturing specific workloads where alternative solutions offer advantages in cost, power, or performance.
AI is already a 60-year-old technology. But really, the big difference is the data, the massive, massive data that is starting to become available. Whatever business you want to be in, you have to be close to the data.
Tan articulates his investment philosophy for AI opportunities. This quote matters because it shifts focus from the technology itself to data access as the determining factor for AI success, providing strategic guidance for businesses evaluating AI investments.
Our Take
Tan’s perspective is particularly valuable because he bridges the semiconductor hardware world with AI application opportunities. His willingness to challenge NVIDIA despite personal friendships with industry leaders demonstrates intellectual honesty about market dynamics. The focus on power efficiency is prescient—as AI scales, energy consumption becomes not just an operational cost but an existential constraint for data centers. His 10-15% market share thesis for alternatives is realistic rather than hyperbolic, suggesting a mature understanding that NVIDIA’s ecosystem advantages won’t disappear overnight. Most intriguingly, his emphasis on healthcare AI and data proximity suggests the next wave of AI value creation will come from domain-specific applications rather than general-purpose infrastructure. Tan’s investment strategy essentially bets on specialization and efficiency over raw performance, which could define the next phase of AI infrastructure evolution.
Why This Matters
This story is significant because it reveals how one of the semiconductor industry’s most respected figures is betting on the future of AI computing infrastructure. Tan’s track record at Cadence and his influence across major tech companies lends credibility to his investment thesis that NVIDIA alternatives will capture meaningful market share. His focus on power efficiency addresses one of the most critical challenges in AI infrastructure as data centers struggle with energy consumption and costs.
The broader implications suggest the AI chip market is maturing beyond NVIDIA’s near-monopoly, with opportunities emerging for specialized architectures optimized for specific workloads. Tan’s emphasis on data proximity as the key to AI success provides a strategic framework for businesses considering AI investments. His heavy focus on healthcare AI signals where experienced investors see the most transformative potential. For the industry, this validates the growing concern about diversification away from single-vendor dependence and highlights the importance of energy-efficient computing solutions as AI scales globally.
Recommended Reading
For those interested in learning more about artificial intelligence, machine learning, and effective AI communication, here are some excellent resources:
Recommended Reading
Related Stories
- Biden hails $20B investment by computer chip maker in Arizona plant
- Jensen Huang: TSMC Helped Fix Design Flaw with Nvidia’s Blackwell AI Chip
- Encharge AI Secures $21.7M Series A Funding to Revolutionize AI Chip Efficiency
- Pitch Deck: TensorWave raises $10M to build safer AI compute chips for Nvidia and AMD