If you frequently follow the semiconductor or PC market, you should be aware that Intel's current situation is not ideal. Stability issues with the 13th and 14th generation Core processors, which began brewing earlier this year, only started to ease in September. The market reception for the Core Ultra 200S series processors released in October was also less than satisfactory, marking one of Intel's toughest moments in history.

As one of the world's most important semiconductor manufacturers and the only one still producing semiconductors in the United States, Intel's importance is self-evident. To save Intel from crisis, the United States has been discussing various solutions, including but not limited to: financial subsidies, tax reductions, and policy inclinations, but most of them are difficult to implement in practice.

For example, the United States previously passed a bill that would provide Intel with billions of dollars in subsidies and up to $100 billion in tax relief if it built new semiconductor factories in the United States. However, Intel's CEO stated that they had already invested over $30 billion in constructing new factories but had not received any subsidies.

In other words, the false promises of the US government are putting Intel in an even more difficult situation. Therefore, to further save Intel, the relevant departments have proposed a "shocking" proposal: merging Intel's chip design department with AMD's relevant department.

Image source: X

When I first saw this news, my immediate reaction was to check the date in the lower right corner to confirm it wasn't April 1st. If Intel merges with AMD, it could be one of the rarest "big events" in the entire history of technology. Intel and AMD are virtually synonymous with PC processors, with their combined market share exceeding 90%. If considering the global installed base of PC devices, their combined share could exceed 99%.

Moreover, AMD has always been Intel's biggest competitor since its inception. The two companies fought fiercely in the 20th century, with patent lawsuits rarely ceasing until they eventually signed a patent cross-license agreement, recognizing each other's right to use the x86 architecture, thus ending the war that affected the entire semiconductor market.

However, this does not mean that AMD and Intel have reconciled. The competition between them in the CPU field has never ceased. It is not easy to get these two companies to shake hands and make peace. Nevertheless, their proposal is obviously not baseless.

Intel and AMD have shaken hands and made peace

Although the merger of Intel's design department with AMD is still far from certain, it does not prevent us from discussing the feasibility of this plan from the perspectives of the market and technology. Firstly, from a market perspective, the current relationship between Intel and AMD is far less tense than it was a decade ago because they now have a common enemy—ARM.

Although ARM processors are not a new concept, they were previously concentrated in the mobile and portable device market, commonly known as smartphones. ARM can also be found in various small electronic devices such as MP3s and MP4s, as well as in devices with low performance requirements. ARM processors have a market share exceeding 90% in the mobile market, which may sound familiar.

It can be said that Arm's (the company that owns the ARM architecture patents) position in the mobile market is similar to the combined entity of Intel and AMD. However, unlike the latter, Arm does not produce specific chips but only provides architectures and chip solutions, resulting in lower profits and revenues. Additionally, since there are numerous actual ARM chip manufacturers, it does not trigger antitrust concerns.

Although Arm is not very interested in the PC market, major semiconductor manufacturers are. Apple's M-series processors and Qualcomm's Snapdragon X-series processors are both PC processors based on the ARM architecture. Although they still cannot match the peak performance of x86 processors, they have impressive energy efficiency ratios, and their performance is sufficient for smooth operation of most applications.

Image source: Lei Technology

Although Apple has already demonstrated the potential of ARM in the PC market with its M-series processors, this has only been limited to Mac devices and has not affected the general Windows PC market. Qualcomm, on the other hand, is directly targeting the Windows PC market.

It can be said that the only obstacle facing ARM PCs is the "system ecosystem." Windows' support for ARM processors is still only adequate. However, this issue is gradually being resolved through cooperation between Qualcomm and Microsoft. Additionally, Qualcomm's cooperation with major PC manufacturers is already on the agenda, putting significant pressure on AMD and Intel.

At Lenovo Tech World recently, we rarely saw Intel and AMD together, announcing the joint establishment of an x86 ecosystem advisory group to jointly promote the development of the x86 ecosystem and address future challenges.

Image source: AMD

Although the two companies are still competitors in the market, they are on the same side when facing the ecosystem competition. After all, no matter how fiercely the two companies compete, users will still stay within the x86 ecosystem. As long as the next-generation processor outperforms the other, the situation can easily be reversed.

However, if users switch to the ARM ecosystem, it will be much more difficult to get them back. In the face of ecosystem competition, cooperation between AMD and Intel is not unthinkable but has already made substantial progress.

Is the merger between Intel and AMD unlikely?

In my opinion, there is indeed a basis for their merger. Firstly, both are maintainers of the x86 ecosystem, and their chip businesses have a high degree of overlap. If they can put aside their competition and cooperate, they can better respond to the attacks of the ARM ecosystem. Secondly, they have cooperated in chip design before. The Kaby Lake-G architecture processor released in 2018 was the first product of their cooperation.

The Kaby Lake-G processor combines Intel's 8th-generation Core CPU and AMD's Radeon RX Vegas M GPU, introducing the following five processors:

Intel Core i7-8809G

Intel Core i7-8709G

Intel Core i7-8706G

Intel Core i7-8705G

Intel Core i5-8305G

Since they are only supplied to a few mainstream laptop manufacturers such as Dell and Lenovo and used in some high-end models, most people may not be aware of these five processors. However, based on actual test results, the performance of Kaby Lake-G processors is not poor and can even be considered the most powerful processors in the mobile PC market at that time.

It is evident that if Intel and AMD can truly put aside their prejudices and cooperate, there is great potential in the field of CPU design. However, Kaby Lake-G was their last hardware collaboration, and no similar CPUs have appeared since. This does not surprise me, as AMD's Ryzen processors subsequently launched a counterattack against Intel and even surpassed it in shipments for a time.

Image source: AMD

Against this backdrop, cooperation between Intel and AMD is naturally unsustainable. Today, six years later, AMD's Ryzen processors are very mature and have iterated to Zen5, performing on par with or even exceeding Intel's. AMD's previous shortcomings have been addressed, making it unnecessary for them to partner with Intel.

From Intel's perspective, merging with AMD is not a good idea either. As the core of Intel, the design department, merging with AMD would mean losing control over the future processor roadmap. Even if the core team can be retained, Intel would completely lose hope of returning to its peak.

Or unless AMD is willing to bring its design team to form a separate department with Intel's design team, jointly serving both companies (similar to Arm's model). However, based on AMD's current development in the CPU market, this plan is unlikely to be approved.

Moreover, the combined share of AMD and Intel in the x86 ecosystem is nearly 100%. The merger of these two companies would also involve antitrust and other issues, making it difficult to obtain approval from governments such as China, the United Kingdom, and France. In my opinion, putting this plan on the table is only to demonstrate the relevant parties' determination to save Intel. There are more options beyond this plan.

For example, splitting the fabs to enable Intel to operate more lightly. However, this plan conflicts with the US desire to achieve semiconductor localization, making it difficult to implement. As the "eldest son of the United States," Intel will obviously not withdraw from the market due to a temporary setback. However, how it will return and whether it can regain its peak position are questions worth discussing.

Perhaps the upcoming wave of AI popularization will become an opportunity for Intel to return to its peak.

Source: Lei Technology