Looking back on the past, you will find that the hardware iteration, design innovation, and experience innovation of PC products have never stopped. In the process of architecture and process upgrading, the new hardware continues to break through the shackles of the old hardware and bring users a stronger performance experience, while the diversified forms and materials make the design of PC equipment continue to innovate. The innovation of screen resolution, refresh rate, and panel technology has improved the visual experience, the maturity of touch technology has improved the control experience, and the hardware iteration has brought about the improvement of performance experience.
So you'll find that innovation in these areas of PC products has never been a real bottleneck.
However, innovation at the functional and application level has not left many traces in the past history. Therefore, once innovative functions and applications such as super terminals appear, it will cause quite a lot of waves. But in addition, the core applications of PC products seem to always stay at the conventional level of office, games, entertainment, etc.
And when we gave the PC a prefix called "AI", we suddenly found that this is the real innovation at the functional and application level that the PC industry has been longing for for a long time. Oh no! Maybe it's more accurate to use innovation.
PC devices have decent computing power, and this computing power is diverse. CPUs and GPUs perform their own duties and depend on each other. Powerful memory, SSD make it look "doable", but unfortunately it's not too smart. Traditional PCs are often defined as "tools" that do different jobs, but may not be much different in nature from screwdrivers, pliers, and shovels. Humans take the initiative to control, give instructions, and then the PC solves the corresponding problem. Even though the hardware is constantly being upgraded, the form is changing, and the experience is changing with each passing day, it is essentially impossible to get rid of the limitations of functions and applications.
AI PCs, on the other hand, have become smarter thanks to AI capabilities and application innovations. Let's take an example that might make it easier to explain.
At the CES2024 pre-show event, Adam Munder, co-founder and general manager of Omnibridge, an Intel partner, demonstrated how AI technology can be used to enable deaf people to communicate smoothly with able-bodied people without the use of translation. As you can see in the image below, the deaf-mute person gestures to sign language in front of the camera of a Core Ultra AI PC, and then the software quickly translates their sign language into English, which allows the deaf-mute to communicate with the able-bodied person quickly and accurately on a daily basis. At this point, you will find that the meaning of the PC device has changed, it is no longer just a cold tool to be manipulated, but "an interpreter who can understand sign language", it has become a bridge between the deaf and the able-bodied.
Not only that, Intel CEO Pat Gelsinger believes that "as we try to find new and more complex models, from simple text models to 3D animation, physical world modeling, etc., the demand for AI computing power will continue to increase." As a technologist, I think this is one of the most exciting moments in the history of computing. We will rapidly advance technological innovation to solve the real problems of humanity. ”
In this point of view, we can see that Pat Gelsinger has a strong technical background, which is not harmful, because from Intel's point of view, diversified AI application innovation will eventually return to the need for technical power. If the technology can't keep up, the functions and applications will be blocked. For example, tablet computers came out as early as 1989, but the immaturity of touch technology and hardware technology has led to a fault in the development history of such devices.
Therefore, paying attention to the technological changes and computing power improvement in the development process of AI PCs can naturally effectively avoid repeating the mistakes of the past.
At present, the AI PC is still in its infancy, and even though the Intel Core Ultra platform has laid a solid hardware foundation for it through CPU, GPU, NPU, and various AI acceleration engines, the maturity of the application is the real maturity. Otherwise, it will be like VR, and the tide will rush and then quickly fall silent.
But that's not all, AI use cases are also creating new and extraordinary economic value. People are accelerating productivity gains through AI use cases. It is precisely because of such huge economic value in AI use cases that the world is actively exploring them.
In Pat Kissinger's eyes, the cycle of exploring the value of AI is like the birth of e-commerce and new ways to reach customers after we enter the Internet, which will last for a decade or even decades. Across the field, humans will discover the economic value of AI use case development, overcome the illusion of AI models, and become more confident in the work they do.
This is just like the early days of Intel's Wi-Fi technology, which didn't change much in the first 2 to 3 years. But with the advent of the Intel Centrino platform, this laptop with direct Wi-Fi functionality was heavily marketed, and suddenly, every coffee shop needed Wi-Fi, every hotel needed Wi-Fi, and every office needed Wi-Fi. This is known as the "Centrino Moment" in Intel's eyes, and it changes the form factor, specifications, and application scenarios of computers.
Today, 20 years later, PCs with the prefix "AI" will also usher in new computer forms and specifications driven by innovative function applications.
As Pat Gelsinger puts it, "We're adding a new neural network visual computing module to the processor for AI workloads, and I believe this will change the user interface." For example, on a computer, users can speak instead of typing. Suddenly, everything that requires access to the cloud can be done locally. ”
Intel is the first to launch AI PC chips, and is confident that it will work together to stay ahead of the curve with quantity, quality, and ecology. More than 100 independent software vendors (ISVs) are already building their applications on top of Intel Core Ultra processors. The result is new visual effects, communication capabilities, audio use cases, text transcription, and more, and the ability to generate solutions on local computers with the support of local computing power. When the scale and experience of these diversified applications reach a singularity, the world will officially enter the era of big AI PCs. Pat Kissinger also believes that this is an important moment for PC. If the personal computer is a "Darwinian" device, the AI PC is one of those "Cambrian" moments.
However, with the rapid development of AI PCs, the related security and privacy cannot be ignored.
In Pat Kissinger's view, "Technology is a neutral platform, and there is no good or bad platform in itself, but how we shape technology and how we use it." We can shape technology into a force for good that continues to help us solve problems that we haven't solved before. I've always believed that we can shape technology, including AI, into an incredible force for good. ”
Nowadays, AI has begun to penetrate into all walks of life, such as radiologists using AI to assist in radiological medicine, financial personnel using spreadsheets to record and process accounts, painters beginning to use AI painting to find creative inspiration, and large language models such as Chatgpt have made visible changes to the creative efficiency of writers like us. Technology is a powerful and wonderful force that helps us work more efficiently, and we can also make it better. Technology is also a complex, and Intel is committed to being a trusted technology provider. In the AI space, Intel is pushing for AI technology to be open, accessible, explainable, and accountable. This is the key to whether AI can be used well and whether it can benefit society.
After solving the "responsible" problem, AI not only needs the spark of AI PC to achieve the momentum of prairie fire, but also needs to give AI breadth and strength through edge, data center and other fields.
In the future, more AI use cases will be applied to edge devices. In this regard, Pat Kissinger gave three reasons:
First, economic cost: running AI on local devices is cheaper than running on the cloud, and there is no need to rent cloud servers outside the device.
Second, at the physical level, it is faster to run locally than to transfer data to the cloud and back again.
Third, privacy and security.
That's part of what "AI is everywhere" means, and Intel will push for it to enable a wide range of native devices to run AI. For example, instead of using hundreds of languages on my local device, I just need the language I can use, and then download it in the cloud if the language I need is not available on my local device.
The edge brings breadth, and the data center brings strength.
With data centers, AI inference workloads are taken on and computed more efficiently. For example, the 5th Gen Intel Xeon Scalable processors, which have AI accelerators built into each core, bring a leap in AI and overall performance, as well as a lower total cost of ownership, and are the industry's leading mainstream data center processors with built-in AI accelerators. This allows business users to use Intel processors in their own data centers to complete higher computing power tasks without installing special servers in the data center or transferring data to the cloud.
The year 2024 is of great significance for the AI PC, because it is the starting point of its journey to the sea of stars. And when AI is combined with the PC, as mentioned earlier, you will find that this is a very rare and truly valuable PC function and application innovation. It is not the so-called innovation that is more gimmick than practical in the past, it will enrich the label of PC, it makes PC not only tool attributes, it gives PC wisdom, it will change all the ways of using PC in the past, which is a simple hardware upgrade, form change, experience optimization can not give magic.
(8510865)