Slashdot Mirror
Xilinx and AMD: an Inevitable Match?
itwbennett writes: Steve Casselman at Seeking Alpha was among the first to suggest that Xilinx should buy AMD because, among other reasons, it 'would let Xilinx get in on the x86 + FPGA fabric tsunami.' The trouble with this, however, is that 'AMD's server position is minuscule.... While x86 has 73% of the server market, Intel owns virtually all of it,' writes Andy Patrizio. At the same time, 'once Intel is in possession of the Altera product line, it will be able to cheaply produce the chip and drop the price, drastically undercutting Xilinx,' says Patrizio. And, he adds, buying AMD wouldn't give Xilinx the same sort of advantage 'since AMD is fabless.'
They could simply get an ARM license and design an ARM server CPU with their FPGA in the same chip.
The manufacturing could be done in Samsung or Globalfoundries since they aren't competitors and are probably going to have the next best manufacturing process next to Intel.
There is no evidence whatsoever that Xilinx will buy AMD. It's just some random idiot's speculation.
Before this it was Samsung will buy AMD, the Chinese will buy AMD, Intel will buy AMD, etc.
If AMD wanted to add an FPGA into their design, I am sure that Xilinx would sell them all the parts they wanted. No need to buy the company.
I'm sure someone will correct me if I am wrong, but I thought AMD's license for the x86 instruction set becomes invalidated if they get bought out.
The Intel-Altera deal, while beneficial for Altera shareholders, is not any kind of huge win for Intel. Intel was already Altera's fab partner, and there's very little incremental revenue compared to the cost. $2B/year for a $17B acquisition, even at a modest discount rate, is a questionable ROI.
The reason is that this deal is questionable is that system design considerations vary considerably, and a fat CPU like an Intel Xeon is not always the best match for a networking application with an FPGA that close. Most of these server-side applications are, in any event, I/O bound in a server environment. That means fast backplane technologies for interfacing the various physical layer devices for networking and storage. Integration of programmable logic rather than putting it on a daughter card with a dedicated interface defeats the purpose of the flexibility that the FPGA provides in this environment, and that's to be able to bridge new and emerging standards while standard products eventually come in and take up the slack. Too little programmable logic and you have to replace the entire part. Too much, and you're killing your margins even now that gates are supposedly "free". Why would a system architect bother taking the risk on that without substantial advantages over the lifetime of their rack-mount beast? And this is essentially true whether or not the die is integrated or put in an multi-chip module or 3D die stack. Even if we consider other applications such as artificial intelligence and image processing, there are already alternatives out there including dedicated processors and GPUs that are doing much of this today, and they're off-the-shelf parts without dependency on the host CPU which - again - would be an I/O bound operation that you wouldn't necessarily want to involve the CPU in directly.
Bringing this to Xilinx, AMD - as the article suggests - has even less presence in server. More importantly, AMD is always 1-2 generations behind in process technology versus Intel, which translates to even greater sensitivity to how much FPGA one devotes to the die. There is no Xilinx fab relationship with AMD since it's effectively fabless. Xilinx and Altera also play in other spaces where x86 is either not relevant or insufficiently so to justify integration (e.g. automotive, broadcast). All of the above points for Intel-Altera apply even more for AMD-Xilinx.
Even in 2015, we're still dealing with external GPUs and Ethernet PHYs on small motherboards. Unless an application reaches true ubiquity and the cost-benefit is clear, integration for integration's sake is a losing cause. If Xilinx and AMD merge, it may very well hurt both companies. Stay tuned.
While you technically may be correct, Intel is a company with a history and reputation of abusing it's monopoly position to put competition out of Business to the detriment of consumers and the market as a whole.
Guess what happens as soon as Intel threatens to revoke their license? I'm sure AMD would be allowed to pursue their X86 endeavors while the anti-trust case began it's processing.
And when it fails, we can blame Intel for going out of business.
The point is, Intel plays by different rules and their Altera purchase represents a smaller percentage of their total worth. But the most important reason you shouldn't copy Intel is if there is an x86+FPGA market, you will never be able to beat Intel at it. If Intel wants your niche, they will take it from you. If Intel has already moved there before you even started, now you don't even have the ability to establish a new market, losing the only minute advantage there is.
I recommend trying to come up with a new idea that Intel isn't actively pursuing. Get some customers and lots of patents, then when Intel wants to take it from you, they at least have to do some costly patent settlements.
“Common sense is not so common.” — Voltaire
No it wouldn't. Intel is already a monopoly, anyone buying AMD would retain the x86 license because the government preventing Intel from controlling 99.999% of the x86 market trumps whatever bullshit is in the x86 license agreement.
Xilinx and AMD aren't a good match. Xilinx has not been targetting the server market. They target custom embedded systems. (See DARPA challenge robots). Now that Altera is focusing on the server market and not on embedded, Xilinx can probably take over that entire space. It seems pretty clear that Altera(Intel) will dominate the server market, Xilinx will dominate the embedded market, Actel will dominate the space/irrational government contractor market, and Microsemi(Lattice) will remain largely irrelevant. At this point, I think the only thing that would change things is if one of these companies gets off their ass and open sources their toolchain so the community can make software support for these parts which isn't completely awful/unusuable.
Don't forget that Intel need a AMD64 license for all there x86 64 bits CPU.
Intel's legal agreement with AMD when the original license expired was that AMD could continue using x86 and certain systems (excluded all chipset work and any newly developed tech) but was under the condition that if AMD is ever sold the x86 license goes bye bye. This is a contractual agreement and only the US could stop it and they won't. AMD can't be sold with the x86 license in-tact. Intel would be ecstatic about such a turn of events because they could kill the AMD x86 competition without an iota of government intervention.
The OP you replied to is exactly right, AMD can't be bought. Any speculation that AMD could be purchased by anyone is just garbage. AMD will either survive or it will die, no one outside can buy them without the loss of their primary product (which would make them worthless to buy).
Ok, so Cyrix was bought by National Semiconductor, which in turn sold Cyrix to VIA. I wonder how much of Cyrix ended up going into VIA's line of C7/Nano processors. . . Did anyone on here actually own a Cyrix machine?
that could open up a lot of possibilities to the more adventurous enthusiasts/hobbyists/developers.
Those doors are already open. What you want is a Raspberry Pi, BeagleBone Black, or one of the variants. If you need something with some real programmable logic, get a Cypress PSOC. The top of the line PSOC eval board is $100. The development environment is free for all of the above solutions. The PSOCs even come with some amount of programmable analog.
Waiting for an x86 chip with built in FPGA is a pipe dream, and a stupid one at that. Anything they made would be hideously expensive because it would be low volume (They wouldn’t waste their time putting it into their high volume chips, because the Dells, HPs, etc... of the world wouldn't pay the premium for something that doesn’t improve their bottom line). If you really want compute power and an FPGA, get a Pi2 and an FPGA daughter board. Together they will cost you less than anything Intel will ever be able to build. The only way you can get processor+FPGA to be cost effective is if there is a huge market for the devices, and deep pockets to pay for huge runs. There is only one market like that in our world, and thats phones / tablets. Cypress pretty much owns the processor+programmable market for portable devices. Intel has tried unsuccessfully for years to break into that market, but they cant be cost competitive because they showed up way too late to the party and cant gain enough market share to afford economies of scale, so their solutions remain priced out of the market. Xilinx is in trouble, and no amount of AMD or not can fix that. Intel + Altera is a last ditch effort at the otherwise impossible on Intels part.
I wish I had a good sig, but all the good ones are copyrighted
This is why lots of people are betting on HLS and other ways to improve hardware design at the moment.
AMD still owns x64 - the x86 license applies to the original 32-bit Intel based work, which is rapidly fading, excepting in sections of the embedded market. So if anybody buys AMD, they automatically get the AMD64, and would, under the above clause, only lose the right to manufacture Intel's 32-bit x86 based CPUs. In the case of server chips, it's the AMD stuff that matters, not the x86. Even if Intel could continue to manufacture Intel64 chips, there is nothing Intel can do about any AMD owner's use of AMD's IP.
Yeah, but that won't affect anybody's use of AMD's 64-bit CPU if they buy out AMD.
No, Intel's legal agreement with AMD when the original license expired was something you weren't privy to, nor were you privy to any agreements as a result of Intel using AMD64.
Further, if the agreement is as you imagine it and AMD would lose the license when being bought by another company, AND if AMD was unable to instead continue operating as a subsidiary (keeping the license), the government absolutely would step in and laugh at the prospect of Intel being the only one making x86 CPUs. VIA isn't doing shit and nVidia isn't either (they have the option to). Intel would not be allowed to be the only x86 CPU manufacturer.
X64 is an extension of the x86 instruction set which is copyrighted by Intel. In other words it's a derivative that Intel has control over through their copyright on the x86 instruction set. The value of x64 is entirely dependent on the x86 copyright that Intel holds so it's worthless to anyone else without a license for x86 from Intel. I have no doubt that the contract that allows AMD to use the x86 instruction set copyright includes clauses that will protect Intel and their use of the derivative x64 in the event AMD breaches the agreement by being acquired.
I like AMD but the fact is they can't be purchased because if they are their primary product is lost. You people are living in a fantasy world if you think they could be. The contract that allows AMD to use the x86 copyright is explicit that those rights are lost when the company is acquired. Hell the only other real consumer x86 license is held by VIA which had acquired Centuar and Intel sued them claiming they breached the no purchase agreement. The only reason they weren't terminated is that VIA held a couple key patents that Intel needed so they gave them a 10 year contract extension. But now that Intel is focused on power use VIA's entire market segment has been lost and they haven't developed a new processor since 2011.
There is a chance Intel will loan AMD money to keep them limping along so they avoid anti-trust review but I doubt it highly these days. Intel has a legitimate argument these days that ARM and other processors now provide adequate competition even though Intel still mostly controls the PC segment they are non-existent in the portable tablet/phone category which ships far more CPU's. In fact ARM has begun to harm Intel's margins which is a key sign of competition. The wintel monopoly was crushed by Linux.
If AMD runs out of money and can't get an influx of cash they will likely see much of their business lost and simply go out of business.
It would be nice to see some usable open source FPGA tool chains.
There are plenty of dev boards out there but I found Xilinx's ISE horrible to get started with and while I found Altera's Quartus easier I wouldn't say it was a walk in the park. That's why I got my DE0 nano to do something at least while my papillio is gathering dust.
Getting started with CUDA is way easier, it would be nice to see it reach cuda levels of ease of use and see the same range of books in Amazon and blog articles covering how to do some simple get-started projects.
Nullius in verba
The government didn't do shit when it came out they were bribing OEMs and rigging benchmarks (which they still do to this day) or used their position in one market (X86) to kill a competitor in another market (chipsets) so what makes you think they would give two shits if they became a monopoly?
You seem to forget the USA is an oligarchy and has been so for quite some time. As long as Intel greases the right palms? They can do whatever they want.
ACs don't waste your time replying, your posts are never seen by me.
If the x64 was dependent on x86 copyright, Intel wouldn't have had to sign a mutual copyright exchange agreement w/ AMD. They could have just done what AMD did w/ the 64-bit and claimed it as their own. In fact, IIRC, that's how they were about to start, when Microsoft made it clear to them that they were NOT supporting two x64 instruction sets, and since they already supported AMD's one, that would have to be the default. Which is what forced Intel to come to terms w/ AMD.
You seem to forget that Intel paid $1,000,000,000 for ICC bullshit.
You seem to forget that Intel didn't hold a monopoly in the "chipset market".
You seem to forget that the US government has recently rejected several proposed mergers for major corporations, including telecoms, that would have resulted in monopolies.
You seem to forget that granting AMD the ability to continue to use their license is a much easier thing than stringing up a corporation and punishing them.
You seem to forget that you still have NO FUCKING CLUE what the terms between AMD and Intel are regarding the x86 and AMD64 licenses. You do NOT know the terms, and you have NO REASON to believe that the license would disappear if AMD were bought by another company.
See my response to rahvin112 below
Meh. Only five years ago AMD had very competitive products in the server range. Opterons were efficient and cheap.
.
I remember Windows Azure launching in Europe with mostly Opterons inside and Steve Jobs buying that capacity to launch iCloud.
The servers at my workplace used to run on Opterons, too. Then our provider changed to Intel (you could still get AMD, but you needed to pay extra) and eventual server upgrades lead to total disapperance of AMD from my previous company
I am pretty sure AMD was not a single digit player in the server market back then.
So they technically work the deal the other way. On the books AMD buys Xilinx or they do it as a merger of equals, even if Xilinx is the buyer in reality.
They can just do an AOL/Time kinda deal where on paper AMD buys Xilinx but 'mysteriously' it's Xilinx's management that ends up in charge.
The agreement was made public, or at least a copy of it. It's linked to an quoted at other places in this post and it is exactly as I've said. AMD is purchased and AMD looses it's copyright license to x86 while Intel will only lose their copyright license to x64 if they are acquired.
The government won't do anything about a contractual dispute between Intel and AMD. Nothing at all. Intel has sufficient evidence in their possession to make the case to any jury that they are facing broad competition from the ARM ecosystem and would spank the Government if they were stupid enough to bring an anti-trust action against Intel. This isn't Europe, the government can't do squat without taking Intel to court.
You seem to forget that the $1Billion dollar settlement wasn't 10% of what it Intel's shenanigans cost AMD.
You also seem to forget that the agreement was made public and is available on the internet. Not only that but it's quoted elsewhere in this very post including linked to the source.