Intel and AMD soar together! Will the "AI latecomers" start an upward trend in stock prices?

According to the Zhītōng Finance APP, as of the close of the U.S. stock market on Monday, the stock price performance of Intel (INTC.US), which has fallen significantly behind chip giants like NVIDIA (NVDA.US) this year, has surged by over 6%. This increase in Intel's stock price is rare for this year. On the other hand, AMD (AMD.US), which has seen a significant decline in its stock price since March, has rebounded significantly for several days in a row, with a nearly 4% increase on Monday. Apple has reclaimed the title of "the world's highest market value listed company". Previously, the well-known Wall Street investment firm Melius Research released a report stating that the stock prices of "latecomers" in the AI field, such as AMD (AMD.US), Apple (AAPL.US), and Intel, may begin to catch up with absolute leaders in the artificial intelligence field like NVIDIA in the second half of this year.

Compared to popular chip giants like NVIDIA (NVDA.US), Micron Technology (MU.US), and Broadcom (AVGO.US), which have seen increases of over 60% this year, the stock prices of the three major giants Intel, AMD, and Apple appear much weaker. Especially in the data center AI chip market where chip foundry business is slow and profitable, Intel's market share is far behind NVIDIA and AMD. Before the sharp rise on Monday, Intel's stock price had long been hovering at the bottom, with a decline of up to 30% since the beginning of this year.

The term "latecomers" here does not mean that they are lagging behind in AI technology development, but rather that their market share in a specific field is far behind the industry leaders. For example, the global AI chip leader NVIDIA holds close to 90% market share in the data center AI chip field, while AMD's share is far below NVIDIA's. However, NVIDIA's dominant position undoubtedly faces a huge threat from AMD.

"We believe that in the second half of the U.S. stock market this year, for some chip, AI-related hardware, and even software products, these stocks that have lagged behind in price increases may 'catch up' with the leaders benefiting from the AI boom," said analyst Ben Reitzes from Melius. Reitzes added that AMD, Intel, Apple, and even traditional tech giant IBM may benefit from this trend of "catching up".

Melius analyst Reitzes pointed out that the "winners of AI stocks" in the first half of 2023 experienced a period of stagnation in the second half of 2023, while the "weak group" in the first half of the year saw significant gains. The analyst cited the VanEck Semiconductor ETF (SMH.US), which saw a high increase of 50% in the first half of 2023 but only rose by 15% in the second half of the year. They also mentioned that companies like Intel and Dell performed well in the second half of last year. Therefore, Melius suggests that a similar pattern may emerge in the second half of 2024 Melius pointed out in the research report that AMD and Intel are expected to benefit fully from the AI PC wave. The two x86 architecture chip giants will integrate CPU+NPU+GPU processors for new AI PC terminal devices running Microsoft Windows operating system. Meanwhile, Microsoft is set to introduce the "Recall" feature in its "Copilot+ PC," which Melius believes could be the upgrade reason closest to a killer application.

Melius also mentioned that Apple is poised for a "AI-enhanced" driven iPhone 16 super upgrade cycle. Apple unveiled iOS 18 with breakthrough on-device artificial intelligence technology called Apple Intelligence at WWDC in June, and announced a collaboration with OpenAI for on-device ChatGPT to enhance the AI experience on iPhones.

Analyst Jordan Klein from Mizuho Securities pointed out that there is a "clear short squeeze trade in progress for chip companies like Intel." "Unless you are betting on a significant decline in third-quarter performance, or there are competitors launching new products based on cutting-edge nodes for AI PCs and data centers, it is difficult to see a gradual negative catalyst leading to a downturn," Klein said.

Intel's stock price surges significantly amid "darkest moment"

Intel's stock price jumped over 6% at Monday's close, marking the largest single-day gain since March 4. However, the stock has already fallen by about 33% year-to-date.

Intel recently disclosed financial information about its chip outsourcing business, including a decline in sales and a significant increase in losses. According to financial data, Intel's chip outsourcing business is expected to have sales of around $18.9 billion in 2023, down from $27.5 billion the previous year. Intel also expects the overall operating losses of its chip outsourcing factories to peak in 2024, leading investors to sell off Intel shares.

Its competitor NVIDIA dominates around 90% of the profitable data center AI chip market with its unbreakable AI GPU ecosystem built on the CUDA software-hardware platform, while another competitor AMD holds close to 10% market share with performance-optimized CDNA architecture and high-performance AI GPU based on 3D chiplet design. In contrast, Intel's Gaudi AI accelerator seems unprofitable, leading to a sharp decline in its stock price.

According to Melius, AI PC could be a significant catalyst for Intel's continued stock price increase. Intel has been deeply rooted in the x86 architecture CPU field on the PC side for decades, already possessing a strong software-hardware and supply chain collaboration system as well as a large loyal user base in the PC domain Intel's newly launched Core Ultra processor integrates AI-specific Neural Processing Unit (NPU) and Arc GPU into the CPU. The NPU is dedicated to accelerating AI inference tasks, making this integrated CPU+NPU+GPU central processor the company's "most efficient processor," marking the official arrival of the AI PC era. Intel's Lunar Lake processor for laptops is scheduled to be launched in the second half of 2024, featuring a "brand-new low-power architecture and significant IPC improvements," with AI data processing performance of the GPU and NPU modules three times higher than Meteor Lake.

Research firm Canalys' latest forecast data shows that 2024 is hailed as the "first year of AI PC," with global AI PC shipments expected to reach 51 million units, accounting for 19% of total personal computer (PC) shipments. However, this is just the beginning of market transformation. It is projected that by 2028, AI PC shipments will reach 208 million units, with the potential to account for over 70% of total PC shipments. The compound annual growth rate (CAGR) from 2024 to 2028 is expected to reach an astonishing 42%.

AMD's Full Efforts in AI Data Centers and AI PCs

AMD, the strongest competitor of NVIDIA in the fields of PC discrete graphics cards and data center server AI chips, is sparing no effort to rapidly launch new server AI chips and edge AI chips for AI PCs, attempting to weaken NVIDIA's absolute dominance with a market share as high as 90% in the lucrative server AI chip market, while aiming to gain a first-mover advantage over chip peers in the edge AI market.

It is understood that the upgraded version of AMD's AI chip for AI data center servers, the MI325X, will start selling in the fourth quarter. AMD CEO Lisa Su mentioned in the opening keynote speech at the Computex conference in Taiwan that the successor of the MI300X will have larger memory and faster data throughput.

The more advanced MI350 series from AMD will be launched in 2025, followed by the MI400 series a year later. AMD's approximately annual release cycle aligns with NVIDIA CEO Jensen Huang's plan for annual AI chip product releases announced the night before at a speech in Taipei In terms of performance, the newly launched MI325X by AMD, based on TSMC's 3nm manufacturing process, will continue AMD's powerful CDNA3 architecture. It will also adopt the fourth-generation HBM storage system - HBM3E, similar to Nvidia's H200, with a significantly increased memory capacity of 288GB and bandwidth boosted to 6TB/s. The overall performance will be further enhanced, while other benchmark specifications and compatibility will remain largely consistent with the MI300X, making it convenient for AMD customers to upgrade. Su Zifeng pointed out that the AI performance improvement of MI325X is the largest in AMD's history, with over 1.3 times the improvement compared to the competing Nvidia H200.

It is reported that the American cloud computing giant Microsoft has provided AMD's MI300X AI accelerator to customers on its cloud service platform Azure. Although AMD is a major GPU manufacturer globally, its development and expansion in the field of AI chips for data center server have always lagged behind Nvidia.

However, as major cloud computing service providers begin to seek alternatives to Nvidia's expensive and scarce AI chips such as H100/H200, and with AMD making some progress in the AI chip field by providing better software and hardware support, the AMD MI300X has now become a popular basic hardware in the AI field. Scott Guthrie, Executive Vice President of Microsoft's Cloud Computing and Artificial Intelligence Business, described AMD MI300X as "the most cost-effective AI GPU in the current Azure OpenAI cloud service product".

For consumers, AMD has announced its third-generation Ryzen AI processor, named "Strix Point", scheduled to be launched in July. They are customized for laptops embedded with large AI models, combining the RDNA 3.5 GPU with the XDNA 2 NPU architecture on the basis of the AMD Zen5 CPU architecture, fully integrating CPU+GPU+NPU for accelerating artificial intelligence tasks.

In terms of the core NPU performance in AIPC, the new Ryzen AI 300 series processors have surpassed Intel and Qualcomm, becoming the most powerful NPU today. Qualcomm Snapdragon X Elite NPU has a computing power of 45TOPS, Intel's upcoming next-generation Core Ultra Lunar Lake NPU also has a computing power of 45TOPS, while the Ryzen AI 300 series reaches up to 50TOPS A more powerful NPU, combined with AMD's high-performance CPU and GPU, can deploy higher-level parameters of edge AI large models in multiple application scenarios.

"AI Smartphone" as the Core Catalyst for Apple's Stock Price

As of the close of the U.S. stock market on Monday, Apple's stock price continued to hit a historic high, rising more than 0.65% to $227.82, with a total market value approaching $3.5 trillion, surpassing Microsoft to return to the top spot in the U.S. stock market. Previously, industry insiders indicated that Apple has raised the stocking target for the iPhone 16 series to around 90 million units this year. During the 618 e-commerce promotion period, Apple's sales have significantly increased, coupled with the confidence in the replacement drive brought by Apple Intelligence, which may jointly boost Apple's sales expectations for the iPhone 16 series.

On June 10, Apple unveiled a series of artificial intelligence features at its global developer conference. This new artificial intelligence technology called "Apple Intelligence" includes AI one-click summary notifications, automatic transcription of voice memos, and Apple's self-customized tools for generating custom emojis.

However, Apple's exclusive AI chatbot technology is not as advanced in usage scenarios as its competitors, prompting it to seek partners, such as ChatGPT launched by OpenAI. The company also believes that customers may want to switch between AI chatbots based on different large models according to their needs, just like they switch between Google search engine and Microsoft's Bing search engine.

Under the Apple Intelligence technology, AI chatbots embedded with edge AI large models such as ChatGPT will be fully integrated with Apple's Siri voice assistant. This move will allow Apple iOS system users to control all functions of a single application (App) with just their voice, and to some extent achieve comprehensive linkage and interaction of applications, innovating the smartphone experience in a combined form of "edge AI + cloud AI". Siri's positioning may no longer be a clumsy formalized voice assistant, but rather a dedicated "all-knowing AI companion" similar to the one in the movie "HER" that understands user preferences and operating habits extremely well.

Insiders also stated that Apple is actively discussing cooperation with Anthropic, the strongest competitor of OpenAI in artificial intelligence startups, and is considering ultimately making the company's AI chatbot a core choice. It is understood that Apple Intelligence will be rolled out to Apple customers later this year as an important part of the iPhone, iPad, and Apple Mac operating systems In 2024, it is also the "AI smartphone" era. In order to run AI large models on smartphones, smartphone manufacturers are actively researching and integrating dedicated AI accelerators. These proprietary AI chips are designed for efficient processing of AI tasks, aiming to improve computational efficiency and energy consumption ratio. Smartphone manufacturers like Apple are also striving to combine cloud computing and edge computing, completing certain complex AI processing tasks in the cloud, while handling real-time or sensitive tasks locally. They are also exploring techniques such as model pruning, quantization, and knowledge distillation to reduce the actual scale and computational requirements of large models, making them suitable for running on resource-constrained application devices at the terminal.

Recently, the well-known research institution Canalys released a research report, predicting that by 2024, 16% of global smartphone shipments will be AI smartphones. By 2028, the institution expects this proportion to grow exponentially to 54%. The compound annual growth rate (CAGR) of the AI smartphone market size from 2023 to 2028 is projected to reach 63%, mainly driven by strong consumer demand for enhanced features such as AI virtual assistants and on-device automation processing. This transformation is expected to first appear in high-end smartphone models, then gradually adopted by mid-range smartphones, reflecting the trend of on-device generative AI as a more ubiquitous advanced technology penetrating the overall smartphone market