Text: Yang Jianyong.
Since OpenAI released ChatGPT at the end of last year, it has promoted artificial intelligence into the era of large models. As a global GPU leader, generative AI has driven a surge in market demand for NVIDIA AI, and NVIDIA GPUs in the market are hard to find, which in turn has driven revenue to double and become the world's largest chip manufacturer.
Under this wave of generative AI, Nvidia's performance in the capital market is also quite amazing, reaching 245 times to become a member of the trillion-dollar club with a market capitalization. The current market capitalization is as high as 12 trillion dollars (about 8..)6 trillion yuan), second only to Apple, Microsoft, Google and Amazon in terms of market capitalization.
In the new wave of artificial intelligence triggered by the large model, Intel also performed well in the capital market, with a market value approaching the $200 billion mark, and the current market value is $194.6 billion, with a range of up to 80% during the year.
It's just that in the capital market, Nvidia is obviously more favored by investors, and the difference in market value between the two chip giants is $1 trillion. Mainly benefiting from the demand for artificial intelligence and cloud computing, companies in various industries have used Nvidia's products for AI computing, resulting in record revenue for Nvidia's data centers and its stock price climbing.
In the third quarter of fiscal 2024, Nvidia's data center revenue reached $14.5 billion, an increase of 279% from the same period last year. total revenue was US$18.1 billion, a year-on-year increase of 206%; Net profit 92$4.3 billion, up from $6.3 billion in the year-ago quarter$800 million, an increase of 1,259%.
For comparison, Intel's revenue fell 8% to $14.2 billion in the third quarter of this year, and its net profit was 2$9.7 billion, down 71% year-on-year. Among them, the revenue of the data center and artificial intelligence segment was 4.8 billion yuan, a year-on-year decrease of 10%. It is worth mentioning that personal computers, as a core business segment, fell 3% year-on-year to $7.9 billion in the quarter.
Overall, in the past few years, Intel's revenue has been sluggish, and the PC computer market has deteriorated, which has had a huge impact on performance. The data center and artificial intelligence markets also face strong competitive pressure from NVIDIA and AMD. In particular, global technology giants rely on NVIDIA to develop AI technology, including Microsoft, Google and other global technology giants rely on it to train AI, driving a surge in market demand for NVIDIA AI.
Nvidia is undoubtedly one of the biggest winners of the ChatGPT wave, providing GPUs, CPUs, networking, AI Foundry services, and AI Enterprise software solutions that are all growth engines running at full speed. In the generative AI outlet, NVIDIA's revenue has not only doubled, but also surpassed Intel's revenue scale and become the world's largest chip manufacturer in terms of revenue, showing strong competitiveness in the artificial intelligence market.
Generative AI is making significant progress, and Nvidia founder and CEO Jensen Huang said the strong growth shows that industries are undergoing a platform transformation from general-purpose computing to accelerated computing and generative AI. In addition, according to Gartner's survey, in the era of large models, enterprises will accelerate the upgrade to intelligence. Of these, 98% of leaders believe generative AI will be able to transform the entire operation of the organization. With massive data proliferation and the continued development of scalable computing power, generative AI is on the stage.
In order to improve competition with NVIDIA and AMD in the AI market, Intel has also launched a series of outstanding AI product portfolios, aiming to help users create ubiquitous AI solutions in various fields from data centers, clouds, networks, to edge and PCs.
Intel's CEO claims that Nvidia's CUDA dominance won't last forever. At Intel's "Make AI Everywhere" event, Intel's mission is to make AI ubiquitous through well-built platforms, secure solutions, and support for open ecosystems. Whether it's Intel Core Ultra, which is ushering in the era of AI PCs, or the 5th Gen Intel Xeon Scalable processors built for the enterprise, we have expanded our AI portfolio.
Among them, the 5th Gen Intel Xeon Scalable processors are improving performance per watt for critical workloads such as artificial intelligence, scientific computing, networking, storage, databases, security, and more. According to the report, a major leap forward in performance and energy efficiency has been introduced, with an average performance improvement of 21% over the same thermal design power range and up to 36% performance per watt over a range of workloads compared to the previous generation.
In addition, the new 5th Gen Intel Xeon Scalable processors are capable of fine-tuning large language models with up to 20 billion parameters and improving their inference performance by 42% with latency of less than 100 milliseconds. At this stage, Intel Xeon Scalable processors are also the only CPUs with MLPERF training and inference benchmark results that continue to improve performance.
Sandra Rivera, executive vice president and general manager of Intel's Data Center and AI Group, said that the 5th Gen Intel Xeon Scalable processors for AI acceleration are designed to deliver higher performance to customers deploying AI in cloud, network, and edge environments. Thanks to our long-standing relationships with our customers, partners, and developer ecosystem, we have delivered 5th Gen Intel Xeon Scalable processors on a proven basis that will enable customers and partners to quickly adopt and scale at a lower total cost of ownership.
Finally, artificial intelligence has been developed and widely used. Now, driven by ChatGPT, it will further promote the development of artificial intelligence, and generative AI is reshaping all walks of life and creating huge business value. In addition, a recent IDC report pointed out that in the long run, the innovation and iteration of AI technology has driven the further implementation of application scenarios, and the hot spots represented by AIGC, digital human, multimodality, AI large models, and intelligent decision-making have brought more imagination and possibilities to the market.
In the face of the generative AI wave, Intel's Gaudi3 accelerator, which will be available next year as scheduled, was also showcased, and said that the performance will be better than NVIDIA's H100. So, under the artificial intelligence boom, it remains to be seen whether it can seize the Nvidia market and whether it can restructure its business.
Yang Jianyong is a Forbes China contributor who provides insights into cutting-edge technologies such as the Internet of Things, cloud services, artificial intelligence, and smart homes.