Nvidia (NASDAQ: NVDA) holds a dominant position in the data center GPU market with an estimated 92% market share as of the end of last year, significantly ahead of Advanced Micro Devices (NASDAQ: AMD), which has 4%. This landscape may change with AMD’s introduction of its MI350 series of GPUs, emphasizing a 4x AI compute improvement and up to a 35x boost in inferencing performance, manufactured on a 3-nanometer process node.
AMD is projected to launch its MI400 series in 2024, aiming for 432 GB of high-bandwidth memory (HBM) and over 19.6 terabytes per second in memory bandwidth. This advancement could enable AMD to capture a 10% share of the AI GPU market, potentially generating $50 billion by 2028, contributing to a total projected revenue of $71 billion by that year.
In contrast, Nvidia’s sales growth may slow, with its shares trading at a high price-to-sales ratio of 26. If AMD aligns with the tech sector’s average and captures its target market share, its market capitalization could reach $575 billion, indicating a potential 150% gain from current levels.
