Recently, AMD held a grand AI PC Innovation Summit in China, which gave a panoramic introduction to AMD's many technological progress, plans and advantages in the AI era, and demonstrated the widespread application of AMD AI in various industries on a large scale.
AMD Chairman and CEO Dr. Lisa Su attended the summit and delivered a keynote speech on AMD's artificial intelligence strategy and global trends. She also accepted an exclusive interview with Kuai Technology after the meeting.
Advertisement
Talking about this visit to China, Dr. Lisa Su said that AI is a field that is changing and progressing rapidly and is full of new development trends, so I hope to come to China and hold an event specifically for the Chinese ecosystem.
[AI PC development trend: hardware and ecology are both critical]
Regarding the currently popular AI PC concept, some people think that it is a marketing concept, and the topic of AI is also a bit overheated. Dr. Lisa Su bluntly said that she does not think AI PC is a marketing concept.
Advertisement
She pointed out,In fact, we are all looking for “killer applications” that can drive PC upgrades and enter a new cycle. She truly believes that AI will be the ability to trigger growth in the PC market.
From the interview with Dr. Lisa Su, the author realized that the concept of AI is still very popular at the moment, because everyone has seen that AI technology is so powerful and hopes to get started using it as soon as possible. Everyone is trying and learning, and AMD is also internally We are using AI to help make better chips in the future and help improve all aspects of the company's operations.
Regarding what exactly is an AI PC, Dr. Lisa Su proposed,From AMD's perspective, AI PC, or AMD's own Ryzen AI, should actually be a combination of CPU, GPU and NPU, especially considering that some of the models can run on any of CPU, GPU or NPU. .
For AI PCs, AMD hopes to optimize the TOPS (processor computing power units) of AI computing power, so it usually gives a TOPS number. The Hawk Point processor can provide a total computing power performance of up to 39 TOPS, including all engine.
As time goes by, you will see that each generation of NPU will become more and more powerful, and the reason why NPU is so important is that AMD wants to use AI heavily on the PC side, but it also wants to get good battery life, so NPU can be efficient TOPS per watt running AI operations is a very important metric.
This has to mention the importance of ecology, because the development and popularization of any new products and new technologies are inseparable from the rapid prototyping and improvement of ecology, especially AI PC.
Dr. Lisa Su said,The traditional PC ecosystem has been running for many years and is quite mature. The AI PC ecosystem has just started. The number of AI PC applications and users is increasing rapidly every month. She believes that many different types of applications will appear and will also be used in the future. All walks of life see the same situationregardless of the financial industry, productivity applications, content creation or games, many new and different use cases will emerge in the ecosystem.
The author believes that currently, AI PC has some key applications in improving collaboration experience and enhancing local productivity experience. Most local training and local reasoning can be implemented on the PC side, compared to always calling on the cloud. , it will definitely bring a better experience to users.
AMD has set a goal to have more than 150 ISV software partners using Ryzen AI technology by the end of this year. Dr. Lisa Su added: “I believe this goal will actually be exceeded because there are so many new companies emerging with new applications in China and abroad.”
Regarding PC development trends, especially the driving role of AI PCs, Dr. Lisa Su believes that the demand for PCs during the epidemic is very high. After the epidemic, the PC market demand has declined somewhat, but she believes that the emergence of AI PCs will trigger the growth of the PC market. It is expected that We will see some growth in 2024.
in,For the AI PC category, we should see continued growth in the next few years, because the fundamental capabilities of AI PC will prompt consumers to replace their PCs.
Of course, AI is still in its early stages. 2024 is a very important year.
Although most AI PCs are positioned at the higher end, Dr. Lisa Su believes thatOver time, expect to see AI PCs appearing in every price segment, there will be many developments in hardware in the next few years. As for software, we have also seen many applications implemented. In the next few months, you will see more applications emerging.
[AMD AI Advantages: The True Trinity]
If AI and AI PC want to develop, the most fundamental and fundamental thing is strong hardware and sufficient computing power, so as to provide the most appropriate support for various AI models and AI applications.
In this regard, AMD is undoubtedly the most advantageous. Compared with its competitors, AMD has strong and balanced CPUs, GPUs, and NPUs, covering both the data center and consumer markets.
Dr. Lisa Su also pointed out,Depending on the specific use case, some may be CPU-heavy, some may be GPU-heavy, and some may be NPU-heavy, so having all three types of processors in AMD's portfolio is a huge advantage.
For example, for laptops, I believe that as time goes by, the importance of NPU will be further highlighted, and each generation will become more and more powerful, because laptops have to consider power and battery limitations to run AI, while on desktops Computers, workstations and clouds can have higher power space and ultimately reach different optimization points. This is the core competitiveness of AMD's product portfolio.
The Ryzen 8000G series brings a truly comprehensive AI engine to desktops for the first time. When combined with notebooks, AMD regards it as a continuum of various systems, allowing AI capabilities to fully penetrate AMD's product portfolio.
Regarding AMD's core competitiveness, Dr. Lisa Su believes that the most important thing is to use the right engine to run the right application. AI not only exists in the cloud, which is important, but also exists on the edge and on the client, so in terms of AMD's product portfolio, having adapted engines for all of these areas is AMD's fundamental advantage.
Going forward, AMD expects there will be more interaction between the cloud and the client. If the engine can be provided on all types of devices, AMD will be able to bring a more optimized experience to users.
In terms of hardware, Dr. Lisa Su is very confident in AMD's ecosystem because AMD has invested heavily in ROCm ecological capabilities and has made great progress.
The author's feeling is that AI developers actually do not often write on lower-level software, and prefer to use higher-level software to write, such as frameworks such as PyTorch and TensorFlow. Therefore, AMD has invested a lot of energy in these parts to ensure that development Readers get very, very good performance when using AMD's platform.
In the author's opinion, the era of AI PC has just begun, with heroes vying for the throne and chaos everywhere.
AMD undoubtedly stood at the forefront of the storm for the first time, especially with its three computing engines of CPU, GPU, and NPU covering all fields and all scenarios, coupled with the strong support of many software and hardware industry partners, it has firmly grasped its core competitiveness.
Through this innovation summit and the conversation with Dr. Lisa Su, we saw AMD and the entire industry’s investment in AI, especially AI PC, which is undoubtedly exciting and even more exciting.
There will be times when the wind blows and the waves break, and the clouds and sails are hung directly to help the sea. I look forward to AMD, and the entire AI PC industry, bringing more excitement!
[End of this article]If you need to reprint, please be sure to indicate the source: Kuai Technology
Editor in charge: Shang Shangwen Q