AMD reportedly in talks to acquire Xilinx for $30 billion

[midian182]

Rumor mill: AMD is reportedly in the advanced stages of buying field-programmable gate array (FPGA)-maker Xilinx for more than $30 billion. Should the talks be successful, a deal could be reached as early as next week.

The Wall Street Journal notes in its report that discussions between the two companies regarding a takeover have stalled in the past, so there’s no guarantee we’ll see AMD acquiring Xilinx. If it does happen, though, it would be another massive deal in the consolidating chip industry—Nvidia is in the process of buying chip designer ARM for $40 billion.

Xilinx has been focusing heavily on the data center in recent years. A buyout by AMD would give team red an arsenal of new products to compete against Nvidia and its data center GPUs. Xilinx’s Versla Premium line—its next-gen 7nm Adaptive Compute Acceleration Platform (ACAPs) products—is designed for high bandwidth networks and boasts features including 123TB/s bandwidth across its network-on-chip (NOC) and PCIe 5.0 support.

Xilinx products are found beyond the data center. They’re also used in wired & wireless communications, AI, Machine Learning, automotive, industrial, consumer, aerospace, and defense industries.

A Xilinx acquisition would help AMD in its battle against Intel. Back in 2015, Chipzilla acquired Xilinx rival Altera for $16.7 billion, integrating the firm into its Programmable Solutions Group (PSG) and acquiring its FPGA line. Intel will have been watching the Ryzen 5000 reveal last night, which included “the fastest gaming CPU,” and AMD could soon be giving its rival new headaches in other markets.

Image credit: michelmond

Permalink to story.

https://www.techspot.com/news/87045-amd-reportedly-talks-acquire-xilinx-30-billion.html

 

[winjer]

Does AMD have that kind of money?
I bet they will have to take a big loan.

 

[VitalyT]

Are we gonna see another ATI-like purchase, followed by AMD crash and burn?

I can imagine Lisa Su doing a headless chicken run out the door shortly after.

 

[Irata]

Does AMD have that kind of money?
I bet they will have to take a big loan.

I was wondering the same thing. AMD's market cap is around the 100 Billion mark right now if I am not mistaken, but even if they cover part of the aquisition via a share swap that would still be a lot of money.

 

[Vrmithrax]

Are we gonna see another ATI-like purchase, followed by AMD crash and burn?

I can imagine Lisa Su doing a headless chicken run out the door shortly after.

Doubtful. Xilinx is well known across multiple industries, and is not a company in trouble or bleeding money. It is also not a huge step out of a zone of familiarity for AMD, unlike back when a CPU company bought a GPU company. Plus, gaining more strength and exposure for AMD in a wider range of industries, as well as the big boost it could give them in the data center market, could be well worth the investment.

 

[Aryassen]

I have to trust Ms Su on this one. I'm cringing as hell (that is a lot of money), but I also understand that the big money is in the industrial market, and a large (or perhaps the largest) part of that is the datacenters. So the acquisition can make a sound investment, I just don't know enough to judge, and that price-tag makes me feel uneasy....hope it works out...gulp...

 

[zamroni111]

Are we gonna see another ATI-like purchase, followed by AMD crash and burn?

I can imagine Lisa Su doing a headless chicken run out the door shortly after.

I agree. It will be better if amd spend that money to develop tensor processing unit.
Asic will be faster than fpga.

 

[Vrmithrax]

I agree. It will be better if amd spend that money to develop tensor processing unit.
Asic will be faster than fpga.

ASIC is faster, yes. Far less flexible than FPGA, though. For complex hardware that might need to make a lot of changes and updating throughout its lifespan, it's far easier to reprogram an FPGA circuit than it is to physically replace an ASIC circuit.

I use FPGA hardware in control systems, having that flexibility far outweighs the required initial buy-in and development that would be needed for ASIC each and every time there were changes or upgrades performed.

 

[VitalyT]

Doubtful. Xilinx is well known across multiple industries, and is not a company in trouble or bleeding money. It is also not a huge step out of a zone of familiarity for AMD, unlike back when a CPU company bought a GPU company. Plus, gaining more strength and exposure for AMD in a wider range of industries, as well as the big boost it could give them in the data center market, could be well worth the investment.

If you check back, the nature of trouble from ATI was that the company's valuation was completely over-blown. Who's to say that Xilinx isn't trying to pull the same crap now?

 

[poohbear]

Does AMD have that kind of money?
I bet they will have to take a big loan.

Well, they went from a $2 stock in 2016 to a $90 stock in 2020. Lisa Su is taking advantage of the massive increase in valuation by using AMD shares to purchase Xilinx. For AMD to continue to grow and justify their valuation, they have to move into enterprise & industrials. There's only so much growth consumer level CPUs & GPUs can provide for AMD before it stalls. They have to expand if they are to grow. Nvidia did that a long time ago with their enterprise GPU line and their accompanying CUDA software. AMD missed out on that, and Lisa Su wants to rectify that.

If Lisa Su decides this is best for AMD, I wouldn't doubt her one bit after the 180 turn around she's done with AMD.

 

[koblongata]

Whoa Xilinx AMD V.S. ARM nVidia, I cannot tell which will win, it is getting exciting.

 

[koblongata]

ASIC is faster, yes. Far less flexible than FPGA, though. For complex hardware that might need to make a lot of changes and updating throughout its lifespan, it's far easier to reprogram an FPGA circuit than it is to physically replace an ASIC circuit.

I use FPGA hardware in control systems, having that flexibility far outweighs the required initial buy-in and development that would be needed for ASIC each and every time there were changes or upgrades performed.

Can you say like implement the entire OS on a FPGA and patch it as needed as if it is a software?

 

[Vrmithrax]

Can you say like implement the entire OS on a FPGA and patch it as needed as if it is a software?

FPGA is less about complex layered code (such as an OS) and more about logic circuits and hardware interfacing, typically. It's essentially a way to create "virtual" logic circuits which can be reconfigured when needed, vs the old physical circuits requiring physical modification or replacement. And, because it's done at such a base control level, it is much faster than writing code subroutines that must run in the OS itself. With FPGA, you can have specialized functions operating in parallel with no CPU overhead, that your OS can access in real time. The flexibility is pretty insane in control and I/O situations.

 

[Puiu]

If you check back, the nature of trouble from ATI was that the company's valuation was completely over-blown. Who's to say that Xilinx isn't trying to pull the same crap now?

They will try to get as much money from the deal as they can, but deals like this can be paid over many years and not just with money.
As long as AMD does ok in the next few years they'll be able to pay no problem and its probably worth the asking price, especially since it seems like Xilinx is using bleeding edge tech which could help AMD speed up development of new competition products.

God knows how many years and billions it would take AMD to develop what Xilinx already has (patents are most likely a huge issue for them too)