Intel Corporation (INTC.US) plans to launch a new AI chip next year to once again make a push into the market.

Intel Corporation (INTC.US) announced on Tuesday that it plans to launch a new artificial intelligence chip for data centers next year. Intel Corporation's Chief Technology Officer, Sachin Katti, stated that this new chip (graphics processor) will be optimized for energy efficiency and support various purposes, such as running artificial intelligence applications or conducting inference. This means that Intel Corporation is once again making a move to try to penetrate the artificial intelligence chip market. This new chip, named Crescent Island, is the latest attempt by the struggling American chip manufacturer to capitalize on the boom in artificial intelligence spending, which has brought in billions of dollars in revenue for AMD (AMD.US) and NVIDIA Corporation (NVDA.US). The company's plans lag behind competitors, representing a significant challenge for Intel Corporation's executives and engineers as they seek to capture market share in artificial intelligence chips and systems. Intel Corporation's CEO, Chen Liwu, has vowed to restart the company's stagnant artificial intelligence projects. Previously, the company had suspended projects like the Gaudi series chips and Falcon Shores processors. Crescent Island will be equipped with a form of memory that is slower than the high-bandwidth memory (HBM) used in AMD and NVIDIA Corporation's data center artificial intelligence chips. This chip will be based on the design used by Intel Corporation for its consumer-grade graphics processors. Intel Corporation did not disclose what manufacturing process Crescent Island will adopt. Since the release of OpenAI's ChatGPT in November 2022, startups and large cloud operators have been rushing to purchase GPUs that help run artificial intelligence workloads on data center servers. The surge in demand has led to shortages and price hikes for chip designs or chips suitable for artificial intelligence applications. Katti stated that the company will release new data center artificial intelligence chips each year, aligning with the annual pace set by AMD, NVIDIA Corporation, and several cloud computing companies that manufacture their own chips. NVIDIA Corporation dominates the market in building large artificial intelligence models, such as those used in ChatGPT. Chen Liwu stated that the company plans to focus on designing chips that can run these artificial intelligence models behind the scenes to ensure that artificial intelligence software runs smoothly. Katti said, "We are no longer trying to design for all workloads; our focus is increasingly on inference." Katti mentioned that Intel Corporation has adopted an open and modular strategy, where customers can mix and match chips from different vendors. Last month, NVIDIA Corporation announced it would invest $5 billion in Intel Corporation, acquiring about 4% of its shares and becoming one of its largest shareholders. As part of the collaboration, they will work together to develop future personal computer and data center chips. Katti stated that this deal is part of Intel Corporation's efforts to ensure that Intel Corporation's central processing units (CPUs) are installed in every artificial intelligence system sold.