AMD Closes $50 Billion Purchase of Xilinx

AMD on Monday completed the acquisition of Xilinx, creating a company that can offer various types of compute devices, including CPUs, GPUs, and FPGAs. Tom's Hardware reports: AMD CEO Dr. Lisa Su told analyst Patrick Moorhead that the first processor combining Xilinx technologies will arrive in 2023, which is in contrast to the company's previous integration efforts. After AMD bought ATI Technologies in 2006, it took the company five years to build its first accelerated processing units (which included AMD's x86 cores and ATI's GPU). This time AMD inked a long-term development pact with Xilinx and was able to work collaboratively even before the regulators approved the transaction. It remains to be seen what exactly AMD plans to offer, but it is reasonable to expect the new processor to feature AMD's x86 cores and Xilinx's programmable engines.

The move will help AMD to continue expanding its presence in the datacenter sector and offer unique solutions that will combine IP ingredients designed by the two companies. Interestingly, the first fruits of the deal are expected to materialize next year. [...] The Xilinx business will become AMD's Adaptive and Embedded Computing Group (AECG), led by former Xilinx CEO Victor Peng. As a result, Xilinx will maintain its leadership for at least a while. Furthermore, AMD's embedded business will cease to be a part of the company's enterprise and semi-custom unit and will merge into AECG, which might be good news as executives from the enterprise division will now spend more time on EPYC CPUs. "The acquisition of Xilinx brings together a highly complementary set of products, customers and markets combined with differentiated IP and world-class talent to create the industry's high-performance and adaptive computing leader," said Lisa Su, chief executive of AMD. "Xilinx offers industry-leading FPGAs, adaptive SoCs, AI inference engines and software expertise that enable AMD to offer the strongest portfolio of high-performance and adaptive computing solutions in the industry and capture a larger share of the approximately $135 billion market opportunity we see across cloud, edge, and intelligent devices."

Altera and Xilinx

[nonBORG]

Since Intel brought Altera, this is catchup. Not sure that Altera is a huge asset to Intel but possibly it is about to make them buckloads of cash with CPUs that can heavily optimise themselves to the task at hand.

Re:

[AutodidactLabrat]

Altera's architecture substituted large general array links via multiplexers at ever intersection without local qualifications as contrasted with Xilinx. Might help with the VLIW architectures, definitely not with Intel's pseudo-RISC architecture. Barring a major change in their address encoding / decoding, figure this is a wash for them.

Re:

[joe_frisch]

Does Altera have the equivalent of the Xilinx RFSOC devices? For some digital radio applications those provide very dense high performance solutions.

Re:

[dsgrntlxmply]

Around 16 years ago I had a project that used Virtex to prototype modems with CIC direct synthesis at 40 MHz IF, destined for Spartan. The analog RF outside the FPGA was mostly reused from high volume products. MAC blocks were the crucial feature then. I had not looked since then, but their RFSOC stuff today looks remarkably capable.

Re:

[joe_frisch]

We are moving to the RFSOC for almost all our new products. The RF ADC performance is fantastic - better linearity than we could get with stand alone parts. We to readout systems for radio telescopes, along with some other digital radio like proijects

Re:

[Ostracus]

I could see the capability coming in handy if AMD ever runs into its own version of SPECTRE.

Re:

[sabbede]

Is it that much about Intel? They compete with Nvidia too. Maybe this is all about DLSS.

Still not as much as Activision

[Dixie_Flatline]

â¦which seems wild to me. I know very well how big the games industry is, but still.

Why?

[bradley13]

This makes little sense to me. Expending $50 billion, when AMD is continually deep in debt. Seems like it would be more important to concentrate on their core business, where they currently have a thin lead over Intel. Why do officers of big companies have this M&A fetish?

Re:

[flashflood]

This makes little sense to me. Expending $50 billion, when AMD is continually deep in debt. Seems like it would be more important to concentrate on their core business, where they currently have a thin lead over Intel. Why do officers of big companies have this M&A fetish?

You might want to take a look at AMD's balance sheet as it stands today. You are some years behind the times, it would seem: https://www.macrotrends.net/st... [macrotrends.net]

Re:

[r2kordmaa]

Why do you think it's not core business? FPGA-co-processors can pack a very impressive punch for compute heavy tasks. It's just that such systems are not exactly bog-standard right now, but they certainly could become so down the line.

Re:

[DrMrLordX]

It's an all-stock deal. AMD is trading away value on paper. Seen their stock price lately?

Why not increase design-ins of both ...

[Anne Thwacks]

... by reducing the incredible cost of the design software - a massive barrier to entry for new design teams. or, alternatively, provide adequate documentation to the open source developers.

I know they are worried about IP theft, but they must know the cost of entry to the market is huge, as they chose to buy a competitor rather than enhance their existing technology.

Don't you people have some kind of "Monopolies and mergers commission"? How come you let the two biggest players in the market merge?

Re:

[r2kordmaa]

That, very much that. Well, fine they can even keep the pricey dev platforms for all it's advanced features, but ffs, release good enough spec that a foss toolchain can be independently made. Right now the poor chaps have to reverse engineer how the fpga chips really work to build toolchains, that's not entirely normal.

Re:

[DrMrLordX]

AMD is not the biggest player in any market, except maybe consoles.

(and even then, not all the hardware is theirs. Just the SoCs)

ONE wish...

[bradgoodman]

Here's my ONE wish for AMD/Xilinx: Clean up your toolchain mess. The kunky and cumbersome multitude of toolchains has IMHO always been the #1 deterrent for using FPGAs more widely. Look what happened with "graphics cards" once people had the tools they needed to use them for more than just video games...

It won't matter if it's still x86/64

[NoMoreACs]

They can sew all sorts of propellers onto a pig; but it still won't run faster than a racehorse.

Obviously, they are trying to create a built-in Afterburner-like Hardware media CODEC, and possibly a ML "core"; but when that 1% of total instructions are not being called-for (like in 99% of all applications), we're still back to the fat-ass pig: Chewing up cycles and guzzling electrons.

Same as it ever was. . .