Six advantages of Intel AI PC application: comprehensive coverage from ecology to experience

Recently, the 2024 new Intel commercial client AI PC product launch conference was officially held. In addition to releasing Intel's Core Ultra vPro processor platform and sharing Intel's progress in ecological layout, software and hardware cooperation in the commercial AI PC field, Intel also demonstrated in detail the current six major advantages of Intel in the commercial AI PC field, from ecological coverage to actual user experience, It comprehensively demonstrates Intel's leadership in the commercial AI PC industry.

First of all, in terms of ecological coverage, Intel's commercial AI PC solutions have been extended to every detail of AI PC applications in just over three months since the beginning of the year. For example, LLM localized language interaction class applications based on the big language model, Stable Diffusion cultural map, graphic map applications, Controlnet image complex processing applications, AI image viewing and speaking applications, and third-party partners Office AI plug-in application and legal assistant application based on big language model cover all aspects of current AI PC mainstream application scenarios.

Secondly, from the perspective of experience, AI PC based on Intel Core Ultra processor platform can provide more excellent AI computing power than competitive products.

Intel Core Ultra processor has quite good AI computing power. Taking Core Ultra 7 155H as an example, and referring to UL Procyon float16 AI computing power test, its NPU score is 283 points, and the sharp GPU score is 365 points. If there is no specific concept of these scores, we can actually refer to the desktop Core i7-14700K released by Intel last October. Its float16 score is 160, and the computing power of Core Ultra 7 155H AI is nearly twice as high; In addition, you can also refer to the Sharp Dragon 7 8845H, which also has AI computing ability. Its Radeon 780M float16 computing ability score is 241 points.

Therefore, thanks to the excellent AI computing capability of the Core Ultra platform, Intel AI PC has a good experience of localized AI applications and has good efficiency in large model applications. So next, let's take a look at how Intel AI PC's six dominant applications can bring users experience.

·LLM language interaction speed is faster

AI Chatbot applications based on the big language model are the most convenient and commonly used AI applications available to mass users. When using ChatGPT, or domestic Wenxin Yiyan and other web applications, it is not the hardware computing power of the personal PC, but the network quality. Fast internet speed, fast enunciation, and good interactive experience.

If the localization deployment is carried out on AI PC, such as using a container such as LLM or LM Studio, it will be affected by the AI computing power of the local hardware. At this point, the Intel Core Ultra platform with better AI computing power can bring users a faster interactive experience. Intel also made an intuitive demonstration at the press conference. Compared with competing products, AI PC text generation on the Core Ultra platform is faster and can quickly respond to and answer user questions. Therefore, it can help users significantly save time on related applications and improve efficiency.

·Stable Diffusion

Stable Diffusion can be said to be the most powerful open source AIGC container at present. Intel also launched the version that supports Intel OpenVINO acceleration as early as last year. It is the most convenient and extensible open source container for cultural and graphic AIGC applications.

As all friends who know about the application of Wenshengdu know, this task load is mainly completed by GPU. Intel Core Ultra platform integrates the outstanding performance of the sharp core display. As we have seen from the test, its AI computing power is about 17.4% ahead of its competitors. In addition, OpenVINO will provide acceleration in the actual application process (can be regarded as Intel's CUDA), Therefore, it only takes about 6 seconds for the Core Ultra platform to generate a 1024 × 1024 high-definition image with the help of Stable Diffusion, while it takes 2.4 seconds for competing platforms to generate 512 × 512 images, which is equivalent to generating 4 times more image area, but the time is only 2 times higher. It can be seen that Intel AI PC, with the help of the Core Ultra platform, has also achieved excellent efficiency in localization of cultural maps.

·SD Controlnet image complex processing speed is faster

In addition to creative AI applications such as cultural maps and graphic maps, AI can also help us with complex Picture processing Work, such as generating pictures that meet specific requirements. At this time, it is necessary to use the Stable Diffusion ControlNet model to perform such tasks.

In brief, ControlNet is a model that controls the image diffusion model by using an additional input image adjustment model. Hmm, it's a bit awkward. You can generally understand it as using AI large model to control parameters such as edges, sketches, and poses to better control image generation.

If this application is carried out locally, it also needs to make use of the GPU with stronger AI computing power and large memory. AI PCs on Intel Core Ultra platform are currently generally configured with 32GB of memory capacity as standard, and the AI computing power of Ruixuan Core has also been seen before. Therefore, the efficiency advantages of AI PCs on Core Ultra platform are easier to be highlighted in this kind of relatively loaded AI applications.

·"Talking from the picture" makes AI PC more intelligent

I wonder if you have seen the Xiaomi Auto Press Conference recently? Lei Jun asks the AI intelligent assistant what the car ahead is? Then a picture appears on the screen, and AI quickly identifies the model of the car and gives the correct answer. This is actually one of the most traditional applications of AI vision. From the era of machine learning and deep learning years ago, to enable AI to understand pictures and give answers to disease analysis pictures is one of the important directions of AI research and development.

Now, human beings have already crossed the border in the era of machine learning, deep learning, and big models drive The AI under is smarter. It can quickly perceive the content in the image, and even help users write an article through the image, rather than just telling users what is in the image, which is a manifestation of AI's more intelligent today.

Intel also made relevant demonstrations at the press conference. AI quickly identified the image content, analyzed the content, gave the answer based on user needs, and even wrote a copy according to the image, which made AI PC really smarter, thus bringing about improvement in efficiency and experience. Imagine that when we see a beautiful picture, but can only say "lying slot" because of poor words, at this time we send the picture to AI to help us quickly write a copy and send it to the circle of friends, or as a copy of our own video production. This is the magic experience AI has brought us today.

In addition, Intel has also joined forces with ecological partners such as Baichuan Intelligent and Zhipu AI to launch intelligent assistant applications for AI PCs, which reduces the threshold for public users to use AI and provides them with convenient AI application experience.

Intel AIGC Assistant Supporting Baihuan2-7B Large Model

Quickly analyze and summarize the core content of PPT through the smart AI PC assistant

·Kutools, making Office applications more intelligent

In the process of promoting the development of AI PC, Intel has brought many AI innovative applications to the ground by relying on its strong ecological power and cooperating with many third-party manufacturers. As mentioned below, Kutools, a third-party plug-in that can help make Office applications more intelligent, has the advantage of being able to be deployed locally, which means that you can use it without networking at all, and then assist users in Office work with the help of a large language model.

For example, in Word In the application, Kutools can help users to quickly write a copy, sort out and summarize the core points of the document; Another example is in Excel In the application, Kutools can help users directly use many hidden and useful functions of Excel, thus greatly improving users' office efficiency.

·Legal assistants improve the efficiency of lawyers

Of course, the magic of AI is not only reflected in the creative application scenarios, but has penetrated into various industries. At the press conference, the Intel AI PC solution exhibition area can be described as a variety of solutions, including slow motion AI framing applied to sports broadcast; AI applications on industrial robots; There are AI related solutions for digital people, medical care, education and other industries. Although many of them are still in the initial stage, the development of AI PC cannot skip this initial process.

Among many applications, the Rubik's Cube intelligent legal assistant shown below can be said to be the best partner of lawyers, judges and other users. According to the introduction of the R&D team, this legal assistant can now perform legal related tasks such as writing legal documents, case analysis, and providing legal advice. Moreover, the legal articles are quoted very accurately, which can greatly improve the work efficiency of relevant users.

·One More Thing

The launch of Intel Core Ultra platform is one of the marks of the first year of AI PC. It adopts a separate heterogeneous hybrid architecture design, and through various instruction sets, it enables the CPU, GPU and the first NPU computing module to have good AI computing power, so that it can deal with different types of AI applications.

For example, CPU, although not good at heavy AI computing, can support AI functions such as face recognition and noise reduction; Another example is NPU, which has low power consumption and is suitable for background and continuously running AI functions. For example, camera background virtualization, eyeball correction, wonderful moment capture, etc., such AI applications on NPU are both power saving and efficient; The GPU has a strong floating point computing capability, especially after the Core Ultra upgraded the Sharp Core Display, the integrated graphics card also has efficient AI computing power, which is why the current integrated display notebook It can also implement AI applications with heavy load, such as cultural map, graphic map, and super-resolution image expansion.

In addition to these natural advantages, don't forget that one of the principles of Intel when designing the Core Ultra platform is to provide better energy efficiency. And what we call One more thing comes from this. Thanks to the better performance of energy efficiency ratio, the Core Ultra platform is now being applied to Windows handheld computers. For example, MSI CLAW, OneXPlayer ranger X1, etc. are all Windows handheld devices based on Intel's Core Ultra platform. They can provide excellent game performance, lower power consumption and longer endurance, bringing more choices to the handheld market.

·Conclusion

Intel AI PC has six advantages. It can provide users with more efficient use experience in AI Chatbot, cultural maps, graphic maps, controllable image generation, complex image processing, speaking by looking at pictures, Office extension plug-ins and legal assistants brought by third-party partners and other industrial applications. In fact, the reason why this is so can be seen from the performance test part at the beginning that Intel Core Ultra is the cornerstone of AI PC's powerful AI experience. It provides better GPU and NPU AI computing power than competitors, and is more skillful in all types of AI applications.

Since Intel first put forward the concept of "making AI everywhere" in September last year and first defined PC as AI, the construction of AI PC software and hardware ecosystem has achieved good results only half a year ago. Intel AI PC can be said to have taken a solid first step, whether it comes from Intel's first party software and hardware support or third-party solutions. In the future, with the maturity of applications and the upgrading of hardware, AI PC will continue to become the core force to change people's production and lifestyle.

This article is an original article. If it is reproduced, please indicate the source: Six advantages of Intel AI PC application: comprehensive coverage from ecology to experience https://nb.zol.com.cn/863/8631870.html