Kazakhstani investment bank Investbanq has released an analytical review on the impact of the new artificial intelligence model DeepSeek R1 on the semiconductor market. The bank concluded that the reduction in AI computing costs is more likely to accelerate technology adoption than reduce chip demand.
What is Known About DeepSeek
DeepSeek R1 is a new AI model that claims a breakthrough reduction in training costs to $5.6 million. This is significantly less than what was previously spent on similar projects. The emergence of such a model raised questions among investors: will cheaper AI lead to a drop in demand for expensive semiconductors?
In its analytical review, Investbanq notes that key questions remain open: the total development cost of R1 including previous research, reliance on existing open-source models, and DeepSeek’s proprietary optimization methods.
Historical Examples
Investbanq analysts cite historical examples where increased efficiency led to demand growth rather than reduction. This is known as Jevons paradox. Server virtualization in the 2000s increased demand for processors and memory. The proliferation of ARM in mobile devices led to the mass adoption of IoT. Cloud migration sparked a surge in demand for high-performance hardware.
The bank also highlights the trend of transitioning from general-purpose GPUs to specialized ASICs. Cloud providers are increasingly choosing custom chips for AI training and inference due to cost and energy efficiency advantages.
Why It Matters
For Kazakhstan, the emergence of such analytics from a local investment bank demonstrates the development of expertise in high-tech investments. The main signal here is that Kazakhstani financial institutions are beginning to analyze global technological trends more deeply, which could influence investment decisions in the region.
Investbanq remains optimistic about semiconductor manufacturing companies, including NVIDIA, Broadcom, and Marvell. According to the bank, these companies will benefit from the growing demand for high-performance AI chips.
What’s Next
The development of analytical competencies in technological investments could become a competitive advantage for Kazakhstani banks when working with institutional clients. This is especially relevant amid growing interest in tech assets within Central Asian investors’ portfolios.
Investbanq predicts that as AI models become more complex, the need for advanced semiconductors will only grow. As an example, the bank cites NVIDIA’s World Foundational Models for physical AI applications.