AMD Acquiring Xilinx In Bold, $35B Semiconductor Mega-Deal

I talked to AMD CEO Lisa Su last night and got the download on what could be a historic, $35B semiconductor deal to acquire Xilinx. AMD had a monster Q3 and guide and has been on a growth tear for years, executing brilliantly on differentiated processor architectures for the PC, datacenter, and game console markets. It is hard to believe the company's meteoric rise, and this acquisition is just icing on the cake. I want to provide the essential news and my opinion on it.

The news per the AMD press release and deck

The following are the deal highlights:

• Price: $35 billion enterprise value, all stock
• Ownership: 74% AMD, 26% Xilinx
• Accretion: Immediately accretive to AMD margins, EPS, and free cash flow generation before synergies.
• Proforma P&L: $11.6B revenue, $5.9B (51%) GM, $2.4B (21%) opinc, $1.8B operating cash flow (non-GAAP)
• Cost synergies: $300 million within 18 months of closing the transaction 
• Combined TAM: $110B
• CAGR: approximately 20%
• Deal closure: expected close end of 2021

The following are management and governance basics:

• CEO: Dr. Lisa Su
• CFO: Devinder Kumar
• President: Victor Peng, responsible for Xilinx business and "strategic growth initiatives."
• Board: Two Xilinx directors join the AMD board of directors

Analyst opinion

AMD and Xilinx have hugely different product portfolios and are focused on different markets with a few exceptions. That is a good thing when it comes to acquisitions and I believe the two companies are complementary.

AMD is focused on high-performance CPUs and GPUs for PCs and datacenter servers and SoCs for game consoles and notebooks. Xilinx is focused on high-performance FPGAs, SoCs for datacenter (including SmartNICs), communications, automotive, industrial, aerospace, and defense markets.

Short term, I can see go-to-market (GTM) synergies. So, imagine one company is strong in automotive and communications, but the other isn't but is strong in CPU and GPU, and that account and engineering team have a whole lot more to offer the customer. While it doesn't sound sexy, it is to the customer.

The two product and market similarities are that the companies are both focused on the higher-performance, higher power draw markets and target the datacenter. This becomes important when you look at technology. 

While both the companies are currently focused on different products and markets, both are on very synergistic, advanced technologies. Think of TSMC leading-edge, chiplets, die stacking, interconnects, and HBM, where I expect the combined to do more with the same, versus a significant cost-cutting exercise. AMD did talk about $300M in synergies. Still, I see those in overlapping operating expenses (finance, human resources, and operations) and potentially COGs for bleeding-edge foundry and high-performance memory costs.

AMD CEO Lisa Su explained to me, "we're coming at it from the point of strength. Both companies have very strong power portfolios today. As we bring the companies together, we'll be able to take advantage of a lot of technology synergy. There's actually a tremendous amount of technology synergy so the products are different, but the baseline technology that we use is very similar."

While the short-term focus appears to be GTM synergies, I expect that the two companies will get much closer to sharing IP for unique SOCs. It may make sense for Xilinx's long-term CPU portions of its advanced, adaptable SoCs to use more AMD's processor and GPU IP. The reverse is true when it comes SERDES. I can see some more interesting datacenter and automotive SoC and ACAPs as a result. Finally, while both companies have a lot of AI opportunity, neither separately have taken advantage of it and would expect to see synergies here.

Su had some interesting things to say about what looks like a red hot SmartNIC market: "We love their (Xilinx) SmartNIC technology. We think that's very strategic for us in the datacenter. I think we'd become more important to the largest cloud providers. As you know, we put our (AMD) CPUs next to their SmartNICs".

AMD doesn't make many acquisitions, and just in the past few years, has it had the resources to do it. Therefore, it doesn't necessarily have the integration reflex like other semiconductor rollups that thrive on it and need it. I'm not concerned with it as much as I was last week as the post-acquisition organization has Xilinx CEO Victor running the business as a distinct business, which means focus. As for the ATI acquisition? I believe it was too slowly integrated, but this was 14 years ago, with almost no AMD management left who can even speak to it. Don't forget, AMD's core business declined not because of the ATI acquisition and a lack of focus, but rather the weakness in its Bulldozer architecture and a fab that couldn't produce it at the right frequency. Whatever integration takes place needs to be done quickly, unlike ATI.

Wrapping up

AMD's intent to acquire Xilinx is a bold move that I think makes sense and caps off an incredible run by AMD. I believe AMD will continue to organically grow with or without this acquisition, as hopefully evidenced by its monster Q3, its upward guide, and roadmap. I do think AMD has a brighter, long-term future with Xilinx as it creates a larger entity that is more diversified across different markets and products while leveraging similar technologies. While I was more excited by its long-term possibilities, I am now equally impressed by its day one accretion. In the near future, I will be doing more analysis on its shorter-term GTM leverage.  

Disclosure: Moor Insights & Strategy, like all research and analyst firms, provides or has provided paid research, analysis, advising, or consulting to many high-tech companies in the industry, including 8x8, Advanced Micro Devices, Amazon, Applied Micro, ARM, Aruba Networks, AT&T, AWS, A-10 Strategies, Bitfusion, Blaize, Calix, Cisco Systems, Clear Software, Cloudera, Clumio, Cognitive Systems, CompuCom, Dell, Dell EMC, Dell Technologies, Diablo Technologies, Digital Optics, Dreamchain, Echelon, Ericsson, Extreme Networks, Flex, Foxconn, Frame, Fujitsu, Gen Z Consortium, Glue Networks, GlobalFoundries, Google (Nest-Revolve), Google Cloud, HP Inc., Hewlett Packard Enterprise, Honeywell, Huawei Technologies, IBM, Ion VR, Inseego, Intel, Interdigital, Jabil Circuit, Konica Minolta, Lattice Semiconductor, Lenovo, Linux Foundation, MapBox, Mavenir, Marseille Inc, Mayfair Equity, Meraki (Cisco), Mesophere, Microsoft, Mojo Networks, National Instruments, NetApp, Nightwatch, NOKIA (Alcatel-Lucent), Nortek, Novumind, NVIDIA, ON Semiconductor, ONUG, OpenStack Foundation, Oracle, Poly, Panasas, Peraso, Pexip, Pixelworks, Plume Design, Portworx, Pure Storage, Qualcomm, Rackspace, Rambus, Rayvolt E-Bikes, Red Hat, Residio, Samsung Electronics, SAP, SAS, Scale Computing, Schneider Electric, Silver Peak, SONY, Springpath, Spirent, Splunk, Sprint, Stratus Technologies, Symantec, Synaptics, Syniverse, Synopsys, Tanium, TE Connectivity, TensTorrent, Tobii Technology, Twitter, Unity Technologies, UiPath, Verizon Communications, Vidyo, VMware, Wave Computing, Wellsmith, Xilinx, Zebra, Zededa, and Zoho which may be cited in this article.