ALMATY – A global shift in artificial intelligence (AI) from model training to real-world deployment is reshaping the semiconductor industry and accelerating demand for faster, more efficient computing – a trend that is also putting countries like Kazakhstan on the radar of international tech companies.
Jason Qu, president of Turing Evolution, and The Astana Times news reporter Ayana Birbayeva. Photo credit: The Astana Times. “A few years ago, the focus was on training models. Now it’s about inference and reasoning. Customers are extremely sensitive to speed, latency and cost,” said Jason Qu, president of Turing Evolution, a semiconductor company, in an interview with The Astana Times on the sidelines of GITEX AI Central Asia & Caucasus 2026.
According to Qu, this transition is driving demand for more efficient computing architectures, in which performance is increasingly measured not just by raw power but by cost-efficiency metrics such as processing speed per watt and per dollar. “In the future, AI systems will interact with each other agent to agent, generating more data and requiring more processing.
This will increase the need for computing power,” he said. Against this backdrop, Kazakhstan is emerging as a potential regional hub for AI infrastructure and deployment. Qu said a combination of geography, policy direction and economic fundamentals shapes the country’s strategic positioning.
“Kazakhstan plays a very important role in Central Asia. It is a hub, and the government has been consistently emphasizing AI development. This is a very fast-moving direction,” he said. According to the expert, in addition to policy focus, Kazakhstan’s relatively low-cost energy resources, both fossil-based and renewable, are becoming a key advantage for data centers and AI infrastructure.
“Energy is directly linked to performance and cost in AI. Countries with affordable and stable energy have a strong competitive edge,” Qu said. According to Qu, Kazakhstan’s approach to AI development is structured, with investments in data centers, digital infrastructure and support for open platforms.
“You have a very clear strategy, building data center capacity, supporting open models and APIs for startups. This is a very strong move,” he said. Such policies, he added, are critical to attracting global technology companies, which are looking not only for markets but also for ecosystems.