Two of the markets AMD has been absent from are FPGAs and Artificial Intelligence. The purchase of Xilinx is intended for AMD to develop a significant presence in both markets, as well as to improve the performance of its CPUs and GPUs.
Shareholders approve the transaction
One of the processes that must go through any purchase of a public company, therefore listed on the stock exchange, is the approval of the shareholders of both companies, especially those who hold shares in the acquired company. And although AMD’s purchase of Xilinx was announced a few months ago, those who own Xilinx shares and are off the board had yet to give the green light.
AMD and Xilinx announced earlier this year the purchase of the former by the latter, which will help AMD expand its portfolio of technologies by adding those of Xilinx to its own. It is therefore the largest purchase made by AMD in its history and crucial for the future of the company. AMD already made another similar purchase in the 2000s to buy ATI Technologies, which is in charge of AMD’s GPU design.
The purchase of Xilinx by AMD does not cause the same controversy and the same consequences as those of ARM by NVIDIA and is no different from the purchase that Intel made a few years ago with Altera. So, unless there is a last minute negative surprise, the buying process should be completed without opposition from regulators.
AMD’s future lies with Xilinx
One of the reasons why one business buys from another is because it hasn’t developed the processes and skills to create certain products and services. Who are the key to the long term strategy of the company. In the case of the hardware market in which AMD operates, all of the strategy and movement of parts is done with a five-year forecast on the roadmap. This time corresponds to the design and deployment on the market of a new technology.
AMD and Intel know that with the increasing cost of wafers per zone, new techniques must be researched to achieve maximum performance in maximum space. The solution? The use of integrated FPGAs that can be programmed to perform certain hardware functions required at specific times. So in the same way that they integrated Radeon GPUs into their CPUs with their AMD Fusion project derived from the company’s current APUs and SoCs.
We should therefore expect a similar process, ending with the integration of Xilinx FPGAs into future AMD designs, both continuing the Xilinx business and integrating its technology into future AMD SoCs, CPUs and GPUs.