As artificial intelligence (AI) continues to integrate across various industries, the demand for computing power is skyrocketing. Chu Junhao, a member of the Chinese Academy of Sciences and a semiconductor specialist, shares his vision on how to meet these growing needs.
During a recent interview with CGTN, Chu emphasized that sectors like medical, materials science, information technology, energy, and health are set to benefit immensely from AI advancements. However, this progress hinges on the availability of substantial computing resources.
Chu, also a popular content creator on Bilibili, advocates for enhancing and expanding computational infrastructure to keep pace with AI's demands. \"It's essential to strategically plan the distribution of computing power, both regionally and internally,\" he stated. He also highlighted the critical role of advancing semiconductor hardware technology in boosting computing capabilities.
Innovations such as compute-in-memory (CIM) technology are paving the way for on-chip processing. This approach minimizes the need to transmit large quantities of unprocessed data, significantly reducing storage requirements. \"By organizing data into smaller, pre-processed units, we can streamline subsequent processing stages and save on computing power,\" Chu explained.
Amid the deceleration of Moore's Law, which predicted the doubling of processing power every two years, Chu points to the necessity for new materials and innovative computing models. Technologies like spintronics, quantum chips, and photonic computing are emerging as pivotal solutions for future challenges in AI and semiconductor development.
Chu also underscores the importance of embedding human-like intelligence into hardware. This integration aims to create more efficient AI systems that emulate human thought processes, potentially reducing energy consumption and enhancing performance. \"Embedding the scientific principles of human thought into computers is a crucial aspect of AI's future,\" he noted, acknowledging that this endeavor requires extensive research and a gradual, step-by-step approach.
As the tech industry navigates these advancements, Chu Junhao's insights highlight the intricate balance between human intellect and cutting-edge hardware innovations in shaping the future of AI and computing power.
Reference(s):
Tech Talk: How to cope with the increasing need of computing power?
cgtn.com