Three years ago, Intel was at its peak, with a market value more than double its current level, and its CEO Pat Gelsinger was on the lookout for acquisition targets. But now, the tables have turned, and Intel itself has become the target of acquisition rumors.

According to CCTV Finance, citing a September 20 report in The Wall Street Journal, Qualcomm has made an acquisition offer to Intel in recent days and has contacted Intel regarding the acquisition. Separately, Reuters reported that Qualcomm CEO Cristiano Amon has participated in negotiations to acquire Intel, exploring various options for the deal. However, the talks between Qualcomm and Intel are still in the early stages, and Qualcomm has not yet made a formal offer to Intel.

Both Intel and Qualcomm declined to comment on the above rumors.

Regardless of the veracity of the rumors, Intel's predicament is undeniable. Today, Intel's market value is only 4% of NVIDIA's and half of Qualcomm's and AMD's. Since its share price plummeted in August, Intel has been shedding jobs, laying off tens of thousands of employees, and delaying factory construction plans in Germany and Poland, where Germany hasn't even been able to distribute the promised €10 billion in subsidies yet.

So how did Intel, once the global chip giant, end up with rumors of a potential sale?

01. The first wrong choice: insisting on going it alone and not partnering with TSMC

As the saying goes, Rome wasn't built in a day, and neither was Intel's decline. Without the cumulative effect of a series of misjudgments, Intel might have avoided many of the setbacks in its destiny. One of the most critical misjudgments was its unwavering commitment to the IDM (Integrated Device Manufacturer) model, which involves handling all aspects of chip design, manufacturing, packaging, testing, and sales in-house.

This model once gave Intel a significant advantage: optimizing resource allocation across all stages, significantly improving production efficiency, and effectively reducing transaction costs, resulting in a high gross margin of over 60% that set it apart in the industry. However, when Morris Chang, the pioneer of chip foundry services, approached Intel with a cooperation offer, it was politely declined.

But the semiconductor industry is notoriously capital-intensive and technology-driven. Once technology and capital reach a certain level, the IDM model can become a hindrance to progress, and Intel's failure to deliver on 10nm technology marked a turning point in its decline.

On the other hand, TSMC's clear positioning as a downstream manufacturing specialist has proven that the more specialized and focused a company is in a particular area of the semiconductor industry, the faster it can grow.

Moreover, Intel's competitors have chosen to partner with TSMC. NVIDIA has risen rapidly to become a leading supplier of graphics processing chips, leveraging TSMC's specialized foundry capabilities. Qualcomm has eased the financial burden of its massive R&D investments by outsourcing manufacturing to TSMC, propelling it to the forefront of the industry. Even AMD, once considered a minor competitor to Intel, has focused on design under TSMC's support, gradually eroding Intel's market share in chip design.

02. The second wrong choice: betting on chip foundry services in a transformational move

Seeing TSMC's success in chip foundry services, Intel also considered entering this field. For example, Intel's current CEO Gelsinger advocates the IDM 2.0 strategy, hoping to revitalize manufacturing capabilities and restore the company's former glory.

However, reality often falls short of idealistic aspirations. In terms of process technology, TSMC leads Intel. To catch up, Intel has recruited talent, built factories, and invested in advancing chip process technology, aiming to surpass TSMC and Samsung. But despite the significant investments, tangible results have yet to materialize.

According to Reuters, Intel has abandoned its previously announced 20A node plan or, at least, it won't appear in any desktop processors. Instead, Intel has announced that it will shift resources from developing the 20A node to the smaller 18A node. The 20A node was set to underpin Intel's upcoming Arrow Lake CPUs. The company has stated that it is currently using "external partners" to manufacture Arrow Lake chips, which are likely to be chip manufacturer TSMC.

Currently, Intel's foundry services are deeply in the red. The company's second-quarter financial report revealed that the Intel Foundry Services segment incurred a loss of $2.8 billion, with an operating margin of -65.5%. Despite this, Intel continues to expand its global wafer fab capacity, investing in Ireland, Malaysia, and four U.S. states. However, due to significant losses, Intel has announced a two-year suspension of projects in Poland and Germany.

03. The third wrong choice: focusing on CPU for AI deployments

Not only did Intel's misguided bet on foundry services fail to facilitate its transformation, but it also saddled the company with a heavy burden and caused it to miss out on the generative AI trend.

As Intel's costs increased, generative AI began to take off. This shift redirected demand away from Intel's central processing units (CPUs) and towards NVIDIA's graphics processing units (GPUs), which are better suited for designing and deploying the most complex AI systems. As tech companies scrambled to acquire scarce NVIDIA AI chips, many of Intel's processors sat idle on shelves.

Gelsinger was forced to cut costs to maintain the company's profitability efforts. Starting in 2022, Intel laid off thousands of employees and cut dividends last year. But it wasn't enough. Last month, Gelsinger announced that Intel would lay off an additional 15,000 employees next year, cut costs by $10 billion, and eliminate dividends.

To be frank, Intel has made efforts in AI, but unfortunately, it once again chose the wrong direction.

Chronologically, in 2015, Intel acquired Havok, a company specializing in 3D physics engine technology. This laid the foundation for Intel's graphics processing and game AI capabilities, particularly in providing enhanced AI experiences for real-time rendering and physics simulations.

In 2016, Intel acquired Nervana, an AI chip startup focused on accelerating deep learning computations. Nervana's technology was integrated into Intel's AI hardware accelerator product line, significantly boosting AI training and inference capabilities. That same year, Intel also acquired Movidius, a company that provided computer vision chips for AI inference and computer vision applications in low-power devices.

Unfortunately, in the era of AI, the protagonist is no longer the CPU but rather processors that can significantly enhance computing power, such as GPUs and NPUs. In contrast, all of Intel's acquisitions have centered around the CPU, attempting to build a powerful computing ecosystem around its x86 CPUs. However, the CPU is no longer the main player in the AI era.

Some analysts believe that Gelsinger's costly recovery strategy has further exacerbated the problem. He failed to foresee the explosive growth of AI, which shifted market demand away from CPUs towards GPUs developed by rival NVIDIA.

04. Conclusion: Choosing the wrong direction leads to greater misfortune through harder efforts

In summary, we can see that Intel made several crucial mistakes at pivotal moments that have shaped its destiny. In the era of global specialization, Intel chose to vertically integrate its upstream and downstream operations. In the AI era, Intel chose to stick with its CPUs.

It is undeniable that for a long time, Intel has consistently provided better and better processors to the world, directly contributing to the popularity of personal computers. However, Intel is no longer the dominant force it once was. As the wheel of time turns, regardless of past glories, being abandoned can happen without warning.

For technology companies, fate can sometimes be extreme. Choosing the right direction can lead to immense success, while choosing the wrong one can result in increasing misfortune despite harder efforts.

As the saying goes, 'When the times are favorable, even the heavens and earth seem to lend a helping hand; when fortune fades, even heroes are not free.'

- End -