ARM intends to acquire AMD? Warren East, the chief executive of processor licensing provider ARM, said that this is not a legend.

ARM's first quarter financial report hit a record high. East said that at the analyst seminar on the first quarter results, someone asked him and the chief financial officer Tim Score about AMD, he simply answered that for a long time ARM had always wanted to acquire AMD, ARM's shareholders also had this expectation. Because AMD is currently rethinking its strategic choices, it is apparent that it has contributed to the "opportunity increase."

In the past 20 years, AMD has always insisted on the x86 processor architecture, but now AMD may abandon the architecture, or at least use ARM processors to strengthen the architecture - this sounds like a very bold idea, but obviously makes sense. Of course, such things cannot happen overnight. Since AMD's x86 processor is ingrained in some applications, it may take several years to get rid of it, but the basic reasons for giving it up seem compelling.

This move also benefits ARM against Intel and strengthens its position in the IBM/Globalfoundries Common Platform Architecture camp.

Since AMD has provided the x86 core as its main business, it has been in the shadow of Intel. This is partly because Intel is the definer of the architecture, and AMD is a follower trying to provide a code-compatible processor at a more favorable price. Intel, as a leader in the financial market, can always use price and manufacturing advantages to maintain pressure on AMD.

Indeed, due to the inability to continue to develop advanced manufacturing processes and to spend billions of dollars while developing microprocessors to build fabs using these advanced technologies, AMD was forced to shake off its manufacturing operations two years ago. This led to the birth of Globalfoundries, which is currently one of AMD's foundry partners.

When Intel's followers are still based on ARM's leader?

However, the transition from simple multicore to many-core processing is obviously a fundamental change for companies like AMD. Software has become more important. To a certain extent, software has redefined the competition model. Intel's use of many-core processor chips, as well as other manufacturers, may not be as much known as others. So now is a good time for AMD to turn to a new battlefield.

Although AMD is focused exclusively on the x86 architecture, its main appeal is to make its IC performance close to Intel's products. With the development of multi-core, keeping pace with Intel is becoming more and more difficult, and for chip design companies such as AMD, which are only involved in the PC market, it will be extremely costly to make it difficult to afford.

AMD's tracking of Intel’s only loot is a mountain of spending and will always follow Intel. Intel itself may not always be able to find the right direction.

However, if AMD joins the broader ARM ecosystem and obtains several million dollars and a few percent of the patent fee for each chip, it will have a ready-made architecture and thus will have time to focus on system-level performance. With the full version of Windows available for ARM processors, this may allow AMD not only to fight in the PC market, but also to meet the broader consumer platform needs of many OEMs.

This will allow AMD to jump out of the lucrative PC market, enter a broader world, and be able to work on everything from mobile phones and tablet PCs to PCs. AMD will face competition from Samsung and other giants, but AMD must be willing to create value by producing something.

Now, choosing whether Windows runs on x86 or Windows on ARM, it becomes a matter of choosing which situation is more power efficient and which customer prefers.

Microsoft and ARM executives will attend a forthcoming Fusion development conference and deliver a keynote speech, which may reveal some clues. The conference will be held in Bellevue, Washington, USA, June 13-16, and includes more than 50 technical seminars involving multimedia, user interface, commercial and high-performance computing and security. Bellevue is the backyard of Microsoft headquarters.

This meeting may be an opportunity to announce the expansion of the Fusion scope to include authorized kernels such as Cortex-A series kernels or Mali graphics x86 cores.

If the board makes a strategic decision to align with ARM, it may also be in line with the fact that AMD's CEO has left. Dirk Meyer resigned from AMD’s chief executive in January 2011 when reports suggested that he was forced to resign because the board was dissatisfied with AMD’s lack of chips for tablets and mobile markets.

The current problem may be that in the fast-growing PC industry, AMD is in no position to design suitable chips for these different areas and performance requirements. But this is the reason to get rid of x86 and let ARM bear some difficult work. And after missing the Tablet PC boom, the ARM license will provide AMD the fastest way to market chips and recover lost ground.

AMD's licensing activities are, of course, fully in line with ARM's business needs with Intel. Indeed, this will make ARM replace AMD, become Intel's chief rival, oppress the world's largest chip maker and make it difficult.

When the x86 architecture is not worth having a second supplier, what impact does this have on the value of the first silicon vendor?

ARM is unlikely to acquire AMD

One thing that may confuse the situation is that if it is part of its strategic considerations, AMD's board of directors prefers to sell the company to a company that is well-funded and keen to acquire off-the-shelf product lines, IC engineering teams, and customers.

So is ARM likely to bid for AMD? I doubt it.

Even if ARM now accumulates funds at an alarming rate and may use borrowing to buy AMD, the move to a chip design company is itself in conflict with the business model that Warren East has inherited and nurtured.

In the microcontroller field of the chip market, ARM had the impulse to acquire other companies. In January 2004, ARM announced plans to acquire Triscend. The latter is engaged in the development of configurable system-on-chip devices and customizable microcontrollers. By February, ARM's acquisition attempt failed, and it is believed that the advice of ARM's authorized companies played a role. In March 2004, Xilinx, an FPGA company, announced that it will acquire Triscend.

In the same way, a heart-thinking idea through the acquisition of AMD to compete with Intel may trigger strong opposition from authorized ARM vendors. ARM's AMD will become their competitors. Why spend money on AMD when AMD is ready to pay to ARM to join the club?

For these reasons, AMD, as an independent chip design company dedicated to multi-core ICs, makes using the ARM architecture seem the most cost-effective and the most likely outcome. This, in turn, may prompt the world's largest semiconductor company to reconsider some strategic options.