Welcome Here  Shenzhen Mingjiada Electronics Co., Ltd.

At AMD's annual event focused on introducing data centre chips and other hardware, AMD CEO Lisa Su unveiled a new AI chip claimed to outperform NVIDIA's competitors in terms of speed, and reiterated her previous forecast that the AI processor market is expected to exceed $500 billion by 2028.

At the Advancing AI 2025 conference held on Thursday, 12 June, in the Eastern Time Zone, Lisa Su projected that the AI processor market size will exceed 500 billion dollars by 2028. From when AMD launched its new products in October last year until last month, Lisa Su had consistently stated that the AI chip market size would reach 500 billion dollars by 2028.

Su Zifeng painted a more optimistic market outlook at the conference, indicating that AMD is increasingly bullish on the potential of the AI chip sector. Previously, she stated that the AI accelerator market would grow by approximately 60% annually over the next three years. On Thursday, she said that the AI accelerator sector is growing at a compound annual growth rate (CAGR) exceeding 60% and predicted that the market for inference chips ‘will grow even faster.’

The logic behind Lisa Su's forecast is straightforward: global tech giants are investing hundreds of billions of dollars in new infrastructure, and supply-demand imbalances have driven the price of a single chip to tens of thousands of dollars. During the earnings call in May, Lisa Su noted that despite tightening restrictions on AI chip exports to China, AMD still expects its AI chip business to achieve strong double-digit growth.

Media reports noted that AMD has recently completed a series of small-scale acquisitions to bolster its chip design and AI software teams. On Thursday, Lisa Su revealed AMD's increased focus on AI. She said that over the past year, AMD has acquired 25 companies related to its AI initiatives.

Lisa Su noted that over the past year, AMD has established 25 partnerships to enhance its capabilities in the AI field, including a collaboration with OpenAI. She also highlighted that xAI, the AI startup founded by Elon Musk, uses AMD's MI300 AI chips.

Commentators noted that one of the core objectives of AMD's acquisitions this year is to address its biggest weakness—the software ecosystem. AMD's ROCm software platform has consistently lagged behind NVIDIA's CUDA in competition, with CUDA viewed by industry insiders as a key defensive barrier for NVIDIA to maintain its market dominance.

At the Advancing AI conference, AMD unveiled its new MI350 series GPUs, specifically designed for AI workloads: the MI350X and MI355X. The company claims the new accelerators offer three times the performance of the previous-generation MI300X, targeting NVIDIA's Blackwell architecture product line. The technical specifications of the two new products demonstrate AMD's ambition to challenge NVIDIA.