Since the advent of smartphones, the word smart has become more and more popular. In particular, ChatGPT, an intelligent artificial text conversational application, has become so popular that it has even sparked the topic of intelligent humans defeating humans around the world**. However, these intelligent AI applications are all based on industry cloud applications. Including Wensheng graph applications such as Midjourney, these require high-performance discrete graphics cards or even dedicated AI computing power cards, and require GPU support with ultra-high parallel computing power to achieve, so ordinary consumers can only experience by purchasing corresponding cloud services, which involves personal privacy and trade secrets are difficult to be effectively guaranteed. As a result, AI PCs for individuals are gradually emerging.
An AI PC is a personal PC with its own AI accelerated computing. This greatly reduces the entry threshold for AI computing for ordinary consumers, and now only one laptop is needed, which can realize various AI operations on a local computer, and at the same time has a thinner body, longer battery life, lower power consumption and heat generation, stimulating consumers to use and popularize AI PCs, and further enriching and expanding the AI ecosystem.
As 2024 approaches, the two hardware giants of AI PCs, Intel and AMD, have begun to increase their efforts. The year ahead is a pivotal year for AI on PC platforms, and even Qualcomm is jumping on the bandwagon. This is a good thing for us consumers, the popularity of AI PCs will greatly improve our office productivity, of course, provided that we choose the right brand platform. In this regard, I think Intel is a worthy leader. Its excellence lies not only in its hardware, but also in its ability to combine software and hardware.
In fact, as early as 2018, Intel began to lay out AI PCs, and Intel put forward plans to introduce AI technology into PCs that year, and at the same time brought the "AI on PC Developer Program" PC Artificial Intelligence Developer Program. And we all know that software and hardware must complement each other to perform perfectly, and both are indispensable. Take the family bucket software owned by Adobe, the head of graphic design software, as an example, the Intel platform obviously has more advantages, and the experience of multiple scenarios such as plane processing and export is better and more efficient, which is the typical result of the complementarity of software and hardware.
Therefore, in Intel's forward-looking layout, in addition to hardware, software is also a very critical component. At the level of software optimization and deployment, Intel and its own investment in OpenVINO continue to make efforts, so that developers can quickly optimize various AI tools into programs that can run more efficiently on the x86 platform, easily realize powerful AI computing localization, greatly popularize the efficiency of AI development and use on the PC platform, and reduce the impact of AI on PCs The demanding requirements of hardware performance make X86 computers independent of AI intelligent cloud computing, and can perform high-performance AI computing on a thin and light laptop without the need for a network.
In the field of consumer hardware, in fact, Intel began to introduce AI acceleration engines as early as the 10th generation Core processors, which further provided an efficient, stable, and reliable development and experimental environment for AI developers in many fields. After four years of accumulation, in the middle of this month, Intel will launch the Core Ultra platform with the first integrated NPU. At the same time, it also integrates a new Arc core GPU that doubles its performance again, which will also bring more efficient performance to AI acceleration. In the future, all laptops equipped with Intel Core Ultra will have powerful AI acceleration capabilities.
For developers' hardware, Intel provides stronger environmental support for AI development. At the same time, there are powerful hardware such as the powerful Pontevecchio GPU and the data center-level GPUflex series, and Intel has provided diversified hardware-level solutions for AI acceleration.
At the software level, Intel has launched a developer cloud platform, which can be used to test and build high-performance applications such as AI, which are guaranteed by Intel hardware, which is stable and more efficient. At the same time, Intel has also piled up a new distribution of OpenVino 20231. A tool suite, which can help developers add more fun and easy-to-use AL functions.
At present, 96EU's Iris XE core graphics are accelerated by OpenVino, which can easily support running FP16 precision models on StableDifusion. In addition, tools such as OneAPL and Webassemblv can help developers achieve efficient operation of large language models, image generation and other algorithms on PC and web. This allows developers to get an efficient AI acceleration experience whether they are in the cloud or on-premises.
There is also this year's popular ChatGPT AIGC, some of which are Intel EVO certified 13th generation Core products, which can run local AI graphics and large language models smoothly, which shows its strength in the industry.
Nowadays, a large number of graphic image applications and creative applications are basically based on the effects and efficiency achieved and achieved by AI, and AI can even help complete creative work. However, there are still big differences in the AI experience on different hardware platforms. For example, the efficiency is different, the effect achieved is different, and even the accuracy of the final effect is different. And this all depends on the depth of cooperation between hardware manufacturers and software application manufacturers. This. In terms of the ability to deepen cooperation, Intel is indeed the strongest. It provides strong AI acceleration capabilities for all laptops based on this platform, and has become the core medium to promote the full implementation of AIGC.