[Apple Shockwave](39) NVIDIA, Did They Buy ARM for This?

NVIDIA Begins ARM-Based CPU Design Amid Lunar New Year Spread
Synergy Expected from Combining GPUs and CPUs Powered by Generative AI
Card to Block Rise of ARM-Based Apple Computers
Market Cap Gap with Apple Expected to Narrow if 3Q Earnings Strong

Editor's Note[Apple Shockwave] is a content series that examines the upheaval caused by Apple entering the semiconductor market. You might wonder why Apple is involved in semiconductors. Apple is no longer just a company that makes smartphones and computers. After long efforts starting from its founder, the late Steve Jobs, Apple has designed world-class semiconductors used in mobile devices. If Intel was the leader in the PC era, Apple has become the top predator in the mobile era semiconductor ecosystem. Amid the global semiconductor supply chain crisis and large-scale investments in semiconductor production facilities, we will carefully examine the upheaval and prospects in the semiconductor market brought about by Apple Silicon to broaden our readers' insights. Apple Shockwave will visit readers every weekend. After more than 40 installments, it will be published as a book.

NVIDIA, which is enjoying its prime era with semiconductors for generative artificial intelligence (AI), has revealed its hidden "claw." It is preparing for a full-scale entry into the central processing unit (CPU) field, which was practically considered abandoned. Reports that NVIDIA is preparing a PC CPU based on ARM architecture rather than Intel's x86 are significant news. This could lead to a phenomenon similar to when Intel, which once dominated the CPU market, bundled its own graphics chips (GPUs). Even Apple is within NVIDIA's reach. A showdown between the "Green Team," representing NVIDIA's AI camp, and Apple can be anticipated.

NVIDIA's move is a counterattack amid Apple, Microsoft, and others strengthening AI semiconductors for 'on-device' AI centered on terminals. Although Apple is also preparing generative AI recently, its basic direction is on-device AI. Apple's A17 Pro chip in the iPhone 15 Pro features a neural processing unit (NPU) responsible for AI computation, with performance roughly twice that of the A16. Intel from the Windows camp has also joined the on-device AI camp by integrating an NPU into its latest Meteor Lake CPU. Enhancing AI functions requires boosting the performance of GPUs or NPUs rather than CPUs. On-device AI processes AI computations directly on the user's device. This contrasts with generative AI, which is a cloud service concept based on NVIDIA chips.

Although generative AI led by OpenAI has rapidly risen, on-device AI is advantageous for users to directly utilize AI functions. On-device AI can enhance various smartphone functions used in daily life. Besides Apple, Qualcomm and Samsung Electronics' smartphone application processors (APs) are also equipped with on-device AI and are further enhancing features to compete with generative AI.

In response to the AP camp's offensive, NVIDIA has proposed a new strategy: a counterattack on the enemy's main force, the CPU. This appears to be a choice targeting the weaknesses of the CPU camp, whose development has slowed. Despite Apple A17 Pro being the world's first to adopt a 3-nanometer process, the CPU's performance improvement has not been remarkable, indicating that the CPU's development speed has reached its limit. This gives NVIDIA ample room to catch up.

CPUs have been a sore spot for NVIDIA. After acquiring PortalPlayer, the core chip design company Apple first used in the MP3 player iPod, NVIDIA introduced the Tegra chip. Apple Shockwave mentioned the 'quadrangle relationship' involving PortalPlayer, Apple, Samsung, and NVIDIA in the third installment. NVIDIA, strong in graphics processing, had only hoped to target the AP market through this. Although Tegra was used in Tesla's Model S electric car and Nintendo Switch game console, its influence in the semiconductor market was minimal. Compared to the record growth of NVIDIA's GPU sector driven by Bitcoin mining and generative AI, Tegra's status was even more modest.

However, it has recently been confirmed that NVIDIA has not given up on CPUs. Some foreign media have reported that NVIDIA is designing ARM-based CPUs for Windows PCs. The focus on PCs rather than mobile devices is noteworthy.

Recently, the PC market, especially the laptop PC market, has seen remarkable growth of Mac computers based on Apple Silicon. The combination of Apple Silicon and Mac has shown outstanding growth, coinciding with the surge in PC replacement demand after the COVID-19 pandemic. Compared to Intel chips, Apple computers attracted consumers due to longer battery life and superior performance. This was a threatening factor for Microsoft (MS), the maker of Windows OS, where ARM-based PCs were virtually absent. For MS, which once pushed Apple to the brink of bankruptcy after Apple opened the personal PC era with Apple II, this change is unwelcome.

To support ARM-based PCs, MS needs chips. Qualcomm recently unveiled the 'Snapdragon X Elite,' but this alone is insufficient to compete with Apple. NVIDIA's introduction of ARM-based Windows-supporting CPUs aligns with MS's direction to expand ARM Windows PC adoption. NVIDIA CPUs are expected to appear in the second half of 2024 or during 2025. NVIDIA already owns Grace, a high-performance ARM-based CPU for servers.

This is a 'win-win' situation for both NVIDIA and MS. Even AMD from the x86 camp is reportedly designing ARM-based CPUs.

NVIDIA Grace CPU

NVIDIA's attempt to embrace ARM also increases the possibility of ARM CPU development. NVIDIA attempted one of the most dramatic mergers and acquisitions in the semiconductor industry. In 2020, NVIDIA announced plans to acquire ARM, which SoftBank of Japan had put up for sale, shaking up the semiconductor industry. The acquisition deal was about $40 billion, the largest semiconductor M&A deal at the time. Had it succeeded, it would have been a historic mega-deal. This remains a painful point for NVIDIA. Following the failed ARM acquisition, NVIDIA's market capitalization surpassed $1 trillion due to the rapid rise of the generative AI market. If it had a new growth engine like CPUs, its corporate value could have soared even higher.

Currently, NVIDIA's corporate value is the undisputed number one in the semiconductor field, surpassing Intel, AMD, and TSMC. In the second quarter, its stock price surged beyond market expectations, and investment banks are again predicting another 'earnings surprise.' BOA Merrill Lynch expects NVIDIA's earnings per share to increase by 475% year-over-year to $3.34, recommending a buy with a target price of $650. NVIDIA's stock price rising 22% in November also reflects strong market expectations for expanded earnings.

NVIDIA's third-quarter earnings will be announced on November 21. The market is paying close attention to how much the valuation gap between Apple and NVIDIA will narrow following this earnings announcement.