AMD To Acquire Xilinx In $35 Billion Stock Deal

[smitty3268]

Yes you can, it's called dilution. Think asset value per share, if Xilinix shares are overvalued then you have diluted your asset value per share in the combined entity, i.e. there are more shares for less worth therefore if they have overpayed existing shareholders take an asset haircut.

Sure, but that doesn't effect AMD because they don't own their stock. It only effects shareholders.

Or, AMD's ability to issue new stock to raise more capital - but it doesn't appear like they are interested in doing that anyway.

I do think they probably overvalued Xilinx, but only because that's almost always the case in these big mergers. It takes overvaluing to get the gears moving behind a deal like this. As long as they don't end up being a lemon, it should be fine. AMD can mostly absorb losing a few % in their stock without too much trouble right now, because of all the good news coming in will likely counteract a lot of the losses.

And Xilinx is unlikely to be a complete lemon. They've had very consistent profits, a steadily rising stock price, and earned more last year than AMD did. They just don't have a large growth market or the public brand name awareness AMD does. If AMD does literally nothing with Xilinx at all, they should still be able to cash in another few hundred million dollars a year in profit, and then apply that towards their R&D.

[pal666]

I wonder if they've overpaid for it especially when you consider Nvidia bought ARM for $40bn

deal is not done. and since arm china is revolting it may never be done

[pal666]

Maybe, it isn't AMD's fault that the graphics drivers contain some bugs and instead the bugs should at least partially be attributed to the programming language in which the drivers are implemented (C/C++)

people select language and then select how they use them. magic languages where bugs are not possible but programs are fast do not exist. slow drivers have little demand

[pal666]

It is exactly the same PL that NVidia is using. Somehow they managed to have great drivers over the last 20 years, something that ATI/AMD did not.

in real world novideo drivers managed to become number one source of bsods

[pal666]

We can shit on nvidia all day long, but their software stack on Linux is superior to whatever amd has to offer today.

their software stack on linux exists only in imagination of nvidiots

[L_A_G]

Xilinx isn't just an FPGA company, they are one of the first companies to adopt chiplets... starting in 2011 and AMD is buying that expertise also in addition to them being a foot in the door of a large segment of hte HPC market. Xilinx was Nvidia's first choice before they bought Mellanox for thier networking hardware (Xilinx has 600G ethernet), Xilinx was Intel's first choice when they wanted an FPGA vendor because they are the leader of the FPGA industry with over 50% market share.

AMD is creating a technology flywheel here.

Xilinx and other FPGA makers have tried getting into the HPC market for quite a while, but so far that's been a total failure so far. As I said in the post you're replying to. The fact that Nvidia and Intel may have wanted to buy them, but the deals never went anywhere or fell trough is just as much evidence that it wasn't worth it to buy them. Either the technology wasn't good enough or then the price was too high.

As for MCM's, considering how much time and money AMD's put into them I have my doubts as to how much Xilinx has to bring here. It may have made sense for Nvidia, Intel or someone else starting from zero or near-zero, but for AMD they're paying a lot a lot for what they already have.

[Anarchy]

Xilinx and other FPGA makers have tried getting into the HPC market for quite a while, but so far that's been a total failure so far. As I said in the post you're replying to. The fact that Nvidia and Intel may have wanted to buy them, but the deals never went anywhere or fell trough is just as much evidence that it wasn't worth it to buy them. Either the technology wasn't good enough or then the price was too high.

As for MCM's, considering how much time and money AMD's put into them I have my doubts as to how much Xilinx has to bring here. It may have made sense for Nvidia, Intel or someone else starting from zero or near-zero, but for AMD they're paying a lot a lot for what they already have.

My suspicion is that AMD is going through a heavy growth and they need more experienced engineers to work on cpus & gpus. Xilins can provide a steady flow of engineers for a while, whatever is left of the carcass in a few years will be sold for peanuts to whoever wants it.

[Anarchy]

their software stack on linux exists only in imagination of nvidiots

i can only lol at that.