Intel Joins Forces with Naver and KAIST to Build an AI Research Center
To challenge NVIDIA’s overwhelming dominance in the semiconductor market essential for creating artificial intelligence, a new collaboration has emerged. American tech giant Intel, South Korea’s internet titan Naver, and the prestigious Korea Advanced Institute of Science and Technology (KAIST) have decided to establish a joint AI research center.
At NVIDIA’s developer conference, the unveiling of their next-generation AI semiconductor, Blackwell, was marked by a demonstration where CEO Jensen Huang highlighted robots that were trained in-house. These robots represent the culmination of NVIDIA’s technologies – an amalgamation of computer graphics, physics, and AI – described metaphorically as “NVIDIA’s soul”.
NVIDIA’s confidence stems not only from its dominant market share in AI accelerators but also from the significance of its AI development programming language, CUDA. CUDA has been indispensable for creating AI models like ChatGPT so far.
Meanwhile, tech behemoths such as Meta, Google, and OpenAI are ramping up efforts to challenge NVIDIA’s solo run. Intel showed early interest in Korean enterprises, with its CEO Pat Gelsinger recognizing Intel’s partnership with Naver as a key player in Asia’s rapidly growing cloud services sector.
Today, Intel, alongside KAIST and Naver, confirms the establishment of the AI research center, set to commence its research activities from July. With ambitions to create a unique ecosystem, the center plans to combine Intel’s AI accelerator ‘Gaudi’ with KAIST’s software research and Naver’s AI model services.
Intel selected its Korean partners, considering their superior AI technology, development capabilities, and the potential to create an independent ecosystem amidst the suitable size of Korea’s AI market.
Key Questions and Answers:
Why is NVIDIA considered dominant in the AI field?
– NVIDIA’s dominance in the AI field is attributable to its advanced GPU technology, which is widely used for deep learning applications. Their GPUs, coupled with the CUDA programming language, have become the de facto standard for AI and machine learning research and implementation.
What is the importance of CUDA in AI?
– CUDA (Compute Unified Device Architecture) is a parallel computing platform and API model created by NVIDIA. It allows software developers to use a CUDA-enabled graphics processing unit (GPU) for general purpose processing – an approach known as GPGPU (General-Purpose computing on Graphics Processing Units). CUDA is crucial for AI because it enables dramatic increases in computing performance by harnessing the power of GPUs.
What are the objectives of the new AI research center formed by Intel, Naver, and KAIST?
– The AI research center aims to advance AI technologies and create a unique ecosystem that integrates Intel’s AI accelerators like ‘Gaudi’, KAIST’s software research capabilities, and Naver’s AI model services. The collaboration seeks to foster innovation and possibly reduce the dependency on NVIDIA’s technology.
How does this strategic alliance intend to counter NVIDIA’s dominance?
– The alliance could counter NVIDIA’s dominance by developing competitive AI accelerators and fostering a robust software ecosystem that rivals CUDA and NVIDIA’s hardware. The partnership, leveraging the respective strengths of Intel, Naver, and KAIST, may produce innovative AI solutions and nurture talent that can advance the field independent of NVIDIA’s offerings.
Key Challenges or Controversies:
– Technology Competition: Developing AI technology that can genuinely compete with NVIDIA’s offerings is a significant challenge. NVIDIA has a well-established ecosystem and a strong reputation in the AI industry.
– Ecosystem Creation: Building a new ecosystem that is as comprehensive and efficient as NVIDIA’s will be difficult. The adoption rate by developers and industries that are already accustomed to NVIDIA’s technology could pose a barrier.
– Market Penetration: Convincing the market to switch or adopt a new platform when there is already a dominant player like NVIDIA is a major hurdle.
Advantages and Disadvantages:
– Advantages:
– Diversification of technology and reduction of industry reliance on a single company, NVIDIA, can prevent monopolistic scenarios.
– Promotion of innovation and competition in the AI semiconductor market could lead to more advanced and cost-effective solutions.
– Collaboration brings together different areas of expertise and could lead to a more integrated and possibly superior approach to AI research and development.
– Disadvantages:
– R&D and production costs for developing a new AI ecosystem are high and it may take a considerable amount of time before competitive products are brought to market.
– Market inertia may inhibit the adoption of new technologies, especially when there is a well-established incumbent.
– Intellectual property rights and regulations could hamper the seamless collaboration among multinational companies and institutions.