AMD held the "ADVANCINGAI" AMDAIPC Innovation Summit in Beijing today. AMD Chairman and CEO Lisa Su attended the event and delivered a speech: "In 2024, the application of artificial intelligence has just exploded. Artificial intelligence is the most transformative technology in the past 50 years. It is reshaping the industry and our lives in ways we cannot imagine today."
"AMD's chips are powering many AI applications." Su Zifeng gave an example: "We have increased the speed of genome analysis by 100 times, helping doctors quickly diagnose life-threatening diseases; we also power some artificial intelligence supply chains and optimize logistics in real time to help companies run smoothly."
Su Zifeng said that some of the applications she is most interested in are medical breakthroughs, scientific research, building smarter and more efficient factories, and truly improving the productivity of every enterprise and every industry. "One of the most important things that people are not fully aware of is that there is no one-size-fits-all standard in artificial intelligence computing, and only AMD has a complete portfolio of high-performance CPU and GPU products to solve a wide range of artificial intelligence workloads." Su Zifeng said.
It said that hardware alone for artificial intelligence computing is not enough, there must also be great software. Over the past many years, AMD has been working very hard and hopes to cooperate with the open source community to achieve these advances. (Wen Meng)
Su Zifeng said, "We all know that hardware alone is not enough, software must also be available. We have been making great efforts in this regard over the past many years." One thing that is extremely important to AMD is that it hopes to work with the open source community to promote technological progress.
"ROCm is our open source software platform. We have added a lot of new features and functions to make it easier for customers to deploy our technology in their software environments. Today, we see that open source can really accelerate the pace of artificial intelligence innovation. We are committed to creating an open and developer-friendly artificial intelligence ecosystem." Su Zifeng said.
Su Zifeng shared that our current technical achievements support all industry standard frameworks, such as PyTorch, TensorFlow, ONNX, etc. We also cooperate with many leading model suppliers, such as HuggingFace, ModelScope, etc., to ensure that AMD hardware can widely support various models.