Gearbest Tech
[Gearbest Technology]Nowadays, top notebook brands are vigorously embracing artificial intelligence technology. In just one year, the best laptops on the market have been equipped with a new generation of neural processing units (NPUs) and processors, which are designed to seamlessly integrate AI technology into our daily lives. The problem is that the AI revolution has been going on for several years, but the actual effect is not that significant.
Now, the PC industry has ushered in Qualcomm’s Snapdragon X Elite chip, which is integrated into the new Copilot laptop. At the same time, AMD launched the Ryzen AI 300 series chips. Intel’s Lunar Lake processor will also join the competition soon.
But they don’t seem to be designed for the future of AI, but are focused on meeting current needs – and often they can’t have both fish and bear’s paw.
Is the AI processor useful?
In the field of chip design, a key factor that is often overlooked is space. If you often visit some hardware forums or websites where enthusiasts gather, you will find how important space is for chip design.
However, for ordinary users, this may not be the issue you need to consider. Companies like AMD and Intel have the ability to make high-end chips, but this time the company has not done so. One of the challenges of chip design is how to tuck as many features as possible in a limited space.
This is very important because adding any hardware to the chip is not free, it means squeezing space from elsewhere. I’m not specifically targeting AMD or Ryzen AI 300 CPUs that perform poorly, but they are not bad in fact. AMD, Intel and Qualcomm always have to make various choices during the design process to put everything needed into the chip. This is not something that can be done simply by adding a little cache. Each change affects many other parameters, all of which need to be balanced.
This also shows that adding NPUs to the chip is not something designers can do easily, and there is no need to make other compromises. At present, these NPUs are almost useless. Even those applications that support AI acceleration tend to use the computing power of integrated GPUs. If there is an independent GPU, it will far exceed the performance of NPUs. NPUs also have some specific uses, but for most users, NPUs are mainly used to provide better background blur capabilities.
The Ryzen AI 300 is an example, and Intel’s Lunar Lake chips will face similar problems. Both AMD and Intel are working to get certified by Microsoft Copilot+ PC, which means that NPUs that meet certain Microsoft power standards must be included. Prior to Copilot+, AMD and Intel had integrated AI coprocessors into their chips, but now with new and higher requirements, these coprocessors are not really of much use.
Without Copilot+’s push, it’s hard to say whether AMD and Intel would design their processors in different ways. But now it looks like we’re seeing a piece of silicon that has barely worked on the Ryzen AI 300 and the future Lunar Lake. This is reminiscent of Intel’s push toward Meteor Lake, which is almost outdated.
AI functions that stay verbally
As AMD and Intel promised, they will eventually join the Copilot+ ranks. Currently, only Qualcomm’s Snapdragon X Elite chips have won Microsoft’s recognition, but AMD has stated that their chips will support Copilot+ functionality by the end of the year. The problem is that there are actually no Copilot+ features available at the moment.
Recall has been the most popular feature since Microsoft announced Copilot+, but no one has actually used it except the media. This feature was postponed by Microsoft, only available in the Windows Insider version, and was delayed indefinitely when the Copilot+ PC was ready to go. AMD and Intel may join Copilot+ by the end of this year, but if Copilot+’s features are not released for a long time, none of this makes much sense.
We can see that Microsoft’s impact on the PC industry is beginning to emerge. There are already a series of new chips from Qualcomm and AMD on the market, and Intel will soon join, all of which are equipped with an AI processor, but they actually don’t play much role. It looks like a hasty move, a bit like what we see in Bing Chat. I can’t help but wonder if Microsoft is really as devoted to the platform as they say.
But then again, Copilot+ PC focuses on not the AI function itself, but longer battery life.
It is predicted that in the next few years, 500 million AI-powered laptops will be sold, and by 2027, such computers will account for more than half of all PC shipments. Obviously, that’s why Microsoft and the PC industry as a whole are trying so hard to promote AI.
There is a chicken or an egg
Even so, it is still important to figure out the status quo. The United States has encountered a typical “Chicken first or egg first” problem in the field of AI PC. Even if the launch of Copilot+ and Recall is delayed, this problem still exists. Intel, AMD and Qualcomm are working to lay the foundation for possible future AI applications, hoping that these applications will be so seamlessly integrated with the way the industry uses PCs that they won’t even realize they are using NPUs. Apple has been doing this for years, and Apple Intelligence is like a natural evolution of this trend.
However, this is not the state of the current situation in the US PC industry, so if you plan to invest in AI PCs, you must be prepared to become an early taster. Unless you look for it, not many applications can make the most of your NPU; even in applications with native AI capabilities, they tend to run on your GPU. Not to mention the changing goals we see in Copilot+ and the first wave of NPUs at AMD and Intel.
Of course, as time goes by, the global PC industry will eventually reach that stage – there is so much money now invested in the field of artificial intelligence that it is destined to become an important part of personal computers. However, it is unknown whether it will become what we expect.
magicCubeFunc.write_ad(“dingcai_top_0”);
Login with Google