AMD: The acquisition of Xilinx cannot be completed on schedule and will be postponed!

Latest update time：2021-12-31

Reads：

Source: The content comes from Semiconductor Industry Observer (ID: icbank), thank you.

AMD and Xilinx, Inc. today provided an update on the progress of global regulatory reviews of AMD’s proposed acquisition of Xilinx.

The parties said: "We continue to make good progress on the regulatory approvals required to complete the transaction. While we previously expected that we would obtain all approvals by the end of 2021, we have not yet completed that process and therefore we now expect the transaction to close in the first quarter of 2022. Our dialogue with regulators continues to be productive and we expect to obtain all necessary approvals."

“We continue making good progress on the required regulatory approvals to close our transaction. While we had previously expected that we would secure all approvals by the end of 2021, we have not yet completed the process and we now expect the transaction to close in the first quarter of 2022. Our conversations with regulators continue to progress productively, and we expect to secure all required approvals.”

There are no other changes to the previously announced terms or plans regarding the transaction, and the two companies continue to expect the proposed combination to create the industry's high-performance and adaptive computing leader.

There are no additional changes to the previously announced terms or plans regarding the transaction and the companies continue to look forward to the proposed combination creating the industry's high-performance and adaptive computing leader.

Is the data center chip landscape expected to be rewritten?

If data centers continue to evolve according to the needs of five years ago, Intel will definitely continue to be the king of this market. Because of their Xeon CPUs and their friendly software ecosystem, Intel can sit firmly on the Diaoyutai. But after AI quickly became popular, GPUs and even FPGAs have become the new upstarts in the data center market. This can be seen from the revenue performance of Nvidia and Xilinx, the leaders in these two fields.

First, let's look at Nvidia. Thanks to the application of GPU accelerator cards in data centers, according to their second quarter financial report released in August this year, the company's data center revenue was $1.75 billion, more than double the $655 million revenue in the same period last year. Most importantly, in this quarter, Nvidia's data center business surpassed the company's gaming business, which has always been famous for the company. This is enough to see the importance of the GPU giant to the data center market.

The same revenue performance also appeared in Xilinx. According to the latest financial report released by the company, as of the quarter ended September 2020, the company's revenue increased by 5% month-on-month, while data center revenue increased by 30% year-on-year. This further enhances its position in Xilinx.

In addition, AMD has also cooperated with TSMC to leverage its process and architecture advantages to create a new generation of products, gradually approaching Intel.

Although the above three companies have achieved some success, they did not stop there and then started the road of "joining hands to fight the enemy". And Intel is their "imaginary enemy".

From the above we can see that AMD, Nvidia and Xilinx each have their own strengths. AMD has leading CPUs and GPUs that can solve data and acceleration problems in data centers. Nvidia's GPUs are the first choice for acceleration cards. Xilinx also hopes to grab a share of the data center market through new ACAPs and FPGAs.

However, judging from the current data center trends, they not only need to deal with computing and acceleration issues, but communication is also a factor they need to focus on. Especially after the current data center expansion has become the mainstream of the industry, communication has become more important, which is why the three major manufacturers are targeting these markets to expand their armaments.

According to industry experts, Intel completed the acquisition of Altera a few years ago, and Nvidia also completed the acquisition of Mellanox this year, and set its sights on Arm. The GPU giant aims to use Mellanox's DPU to solve communication problems, and also hopes that Arm's CPU can become the choice for data center computing.

On the other hand, AMD, which owns CPU and GPU, acquired Xilinx and also has its own combination of "computing + acceleration + communication". Moreover, Xilinx not only has FPGA, but also has powerful ACAP. This acquisition makes AMD even more powerful.

On the other hand, Intel, the leader in data centers, has CPUs, GPUs, and FPGAs, but in fact, according to industry experts, Intel's GPUs are not as good as before, and Altera's FPGAs are also inferior to Xilinx. This means that in the data center market, they are still the same as before, with no outstanding features except CPUs. On the contrary, from the surface, AMD may be the one with the most even strength.

In other words, after this year's acquisitions, the data center monopoly is expected to change. However, we should also realize that whether it is Nvidia's acquisition of Arm or AMD's acquisition of Xilinx, there is a possibility that the acquisition will not be allowed. So all this is still full of variables.

Will the FPGA market usher in new changes?

In recent years, the FPGA market has been dominated by two oligopolies. According to MRFR 2017 statistics, the global FPGA market is dominated by Altera (acquired by Intel in 2015) and Xilinx (acquired by AMD in 2020), and these two giants monopolize about 71% of the global market share. But today, these two giants have been acquired one after another, and the entire FPGA industry has ushered in new opportunities.

Before discussing the new variables in the entire FPGA market, we must clearly realize that even if Altera and Xilinx are acquired, it will not prevent them from continuing to be industry leaders. However, from the transaction of Intel's acquisition of Altera, we can see that because the acquirer acquired the FPGA giant only because of the synergy between some of their products and their product lines, they did not have enough energy to involve in other markets. This is why some people say that Altera has been getting worse and worse after being acquired by Intel.

Referring to the above ideas, Xilinx may face the same situation after being acquired, which will bring new opportunities to the FPGA market.

At present, in addition to Lattice from the United States, there are many independent FPGA manufacturers in China, including Tongchuang, Anlu, Fudan Micro, Jingwei Qili, Gaoyun and Zhiduojing. In addition, there are also companies such as Achronix, Flexlogic and Meta that make eFPGA.

After the acquisition by Xilinx, it may be the biggest benefit for Achronix. Because looking at the current FPGA manufacturers, no one except them can launch FPGAs with strong enough performance to serve the market in the short term. Of course, the domestic situation will be different. Under the dual influence of internal and external factors, domestic manufacturers will have to invest their experience in FPGAs of various capacity ranges. According to the author, there are already teams in China planning 7nm FPGAs, and Xilinx's selling will undoubtedly push them further, because according to past experience, their acquisition will not only give up part of the market, but also a group of employees will leave. These people will inevitably become the "variables" that cannot be ignored in the current FPGA market.

Appendix: AMD's thoughts on acquiring Xilinx

*Disclaimer: This article is originally written by the author. The content of the article is the author's personal opinion. Semiconductor Industry Observer reprints it only to convey a different point of view. It does not mean that Semiconductor Industry Observer agrees or supports this point of view. If you have any objections, please contact Semiconductor Industry Observer.

Today is the 2905th content shared by "Semiconductor Industry Observer" for you, welcome to follow.

Latest articles about

■SiC giant, rebirth: how to predict the future?

■Apple chips may hit Qualcomm hard

■Chip cost per car: soaring to $1,000

■TSMC 2nm, important information

■Huang Renxun's latest views

■The risks of this type of chips that are promising have increased significantly!

■NPU, how to see it?

■Storage giants are abandoning DDR 4

■Intel, why?

■Nvidia will definitely be disrupted