Kneron’s CEO explains how their purpose-built NPUs are revolutionizing edge AI by offering efficient, customizable, and privacy-focused AI processing directly on devices, outperforming traditional CPUs and GPUs. He highlights Kneron’s technological leadership, global market presence, and strategic partnerships, positioning the company for growth and a future IPO following the launch of a major customer project.
The CEO of Kneron discusses the evolution of AI hardware, tracing the shift from CPUs to GPUs and now to NPUs (Neural Processing Units). He likens this progression to the transition from VHS to DVD to NBC, emphasizing that while NPUs are physically small, they deliver significant computational power tailored for AI applications. Unlike CPUs, which were designed for logical operations, and GPUs, which were originally for graphics and gaming, NPUs are purpose-built for efficient AI processing, making them more suitable for modern edge AI needs.
Kneron positions itself as a pioneer in edge AI, having introduced its own NPU architecture a decade ago. The CEO highlights that their technology is versatile, capable of being deployed in both edge devices like smartphones and wearables, as well as in data centers. He notes that their architecture avoids the heavy memory requirements of traditional GPUs, making it more efficient and practical for a wide range of applications, including those with strict power and size constraints.
The CEO explains that Kneron’s edge AI solutions enable businesses to build their own AI applications directly on devices, reducing reliance on cloud-based services like OpenAI. This approach offers advantages in privacy, latency, and customization, allowing companies to create their own AI models and services tailored to their specific needs. He draws a parallel to how advances in chip technology have made today’s smartphones more powerful than supercomputers from decades ago, suggesting that Kneron’s NPUs are similarly transformative for AI at the edge.
Geographically, Kneron’s main markets are Korea, Japan, the US, and recently the Middle East, with a new site in Saudi Arabia. The company also serves customers in China, particularly in consumer electronics and home appliances, as well as in sensitive sectors like defense. The CEO acknowledges China’s aggressive push for AI self-sufficiency but maintains that Kneron remains a technological leader, with core innovations and patents that set it apart from competitors.
Regarding industry partnerships and future plans, the CEO mentions collaborations with major players like Foxconn and Qualcomm, and notes that some large companies even use Kneron’s technology in their products. He sees Kneron as a key player in the next era of edge AI and is focused on building a strong customer base and industry influence before pursuing an IPO. The company is awaiting the launch of a major customer project in Q2, which the CEO believes will further solidify Kneron’s position as a leader in the field and prepare it for a meaningful public offering.