Amd acquisition rumors not so unfounded...

[jhu]

At this point, I'm not sure the patents matter all that much. They only last 20 years. I'm not sure, but x86 should be expiring soon? Meaning that AMD could be acquired, and that whole agreement becomes meaningless. And it appears that x64 will be expiring somewhere around 2023. Not that far away really, in CPU development cycles.

A lot of x86 patents have already expired. Want to make an out of order x86 processor with translated micro-ops? Well, all the Pentium Pro patents should expire later this year, so have at it.

 

[III-V]

Scenario I'm thinking of...

AMD declares bankruptcy.
Intel takes a $5B check to court and says "We want it all".

Yep, that is how "democracy" works these days. Well, it's always sort of worked like that, it's just particularly bad at the moment. However, there'd be a rather large public uproar if Intel were to try, and it'd be blocked like the Comcast/Time Warner merger was.

Not sure why anybody thought that AMD's license wouldn't transfer if they were bought out, though. There's no way Intel would be allowed to be the sole major x86 supplier. Doesn't matter what the contracts or court rulings say... things can be overturned.

This is why I've been rather uncaring about the idea of AMD going bankrupt. Just about any potential buyer would be more competent than AMD's BoD. The instant cash infusion would do wonders by itself, too.

 

[jpiniero]

If Intel really thought that AMD could somehow get the license transferred even in bankruptcy they would have hostily bought AMD by now.

There's no way Intel would be allowed to be the sole major x86 supplier.

I don't really see why they wouldn't be. x86 at this point doesn't really matter to anyone other than Intel since they need to protect their cash cow (Corporate laptops and desktops). Everything else Intel does could easily be done on a different architecture; be it ARM or something brand new.

Unless Samsung or anyone wanted to or thought they could crack the corporate market nobody will bother even if Intel was forced to license x86 out.

 

[Valantar]

I honestly don't see how anyone could hope for Samsung buying AMD, condsidering they have demonstrated (clearly!) that they have zero interest in the PC field. Not only do they not make desktops, they recently stopped selling laptops as well, at least in Europe. Why on earth would they then buy a PC processor company? For patents? Mobile graphics? Somehow making AMD X86 into a mobile-only platform? No way. If Samsung buys AMD, that's the end of competition in the X86 field. For ever. And I for one do not want an Intel monopoly, no matter if their CPUs are better as of now.

 

[scannall]

If Intel really thought that AMD could somehow get the license transferred even in bankruptcy they would have hostily bought AMD by now.



I don't really see why they wouldn't be. x86 at this point doesn't really matter to anyone other than Intel since they need to protect their cash cow (Corporate laptops and desktops). Everything else Intel does could easily be done on a different architecture; be it ARM or something brand new.

Unless Samsung or anyone wanted to or thought they could crack the corporate market nobody will bother even if Intel was forced to license x86 out.

Well, except for the large majority of software out there in the world... I'd like to see x86 just die. The lack of competition in mainstream CPU's is not a good thing. But it won't be dying any time soon. It'll be interesting to see what happens when x64 is no longer covered by patents. So it won't need to be licensed.

 

[Valantar]

x86 at this point doesn't really matter to anyone other than Intel

Except for the makers of ... 90% of the world's software, and the users of this software? Sure, ARM is huge in mobile. There are other architectures for servers and the like. But for everything in between the world runs on X86.

Saying "x86 at this point doesn't really matter" is like saying gas-powered cars don't matter, as there are electric cars on the market.

 

[jhu]

It'll be interesting to see what happens when x64 is no longer covered by patents. So it won't need to be licensed.

Likely nothing interesting will happen. The first wave of x64 patents should expire no later than 2023. Even so, all the Pentium (P5) patents expired 2 years ago, yet there's no rush to get into the x86 space. And among those with the relevant patents and cross-licensing agreements, only AMD is remotely competitive. Where are VIA's x86/x64 offerings?

 

[jpiniero]

Except for the makers of ... 90% of the world's software, and the users of this software? Sure, ARM is huge in mobile. There are other architectures for servers and the like. But for everything in between the world runs on X86.

But the popularity is really because of Windows; not x86. Especially now, people mostly use their computer to use a web browser, email, and maybe MS Office.

Intel's success in servers has more to do with their chips than x86.

 

[scannall]

But the popularity is really because of Windows; not x86. Especially now, people mostly use their computer to use a web browser, email, and maybe MS Office.

Intel's success in servers has more to do with their chips than x86.

You are right, the popularity is because of Windows. But aside from the failed Surface RT, it's the only ISA they write an OS for. At that point, it doesn't matter if Intel is popular or not since you really have no choice. OS X is on Intel as well. Though iOS is a subset of OS X and runs on ARM.

I don't see Linux taking over the world, running on POWER8 processors any time soon.

 

[JM Popaleetus]

Be sold would be a big waste of opportunities for AMD. Being sold at a low price(due to current market value) just to fill shareholders pockets and end with many good R&D projects is not a good idea IMO.

Buyer surely can absorb many things of AMD with the deal. Best buyer would be IBM or Qualcomm, and IMO only the second may have interest in the deal...

Qualcomm, Samsung, MediaTek, all come to mind. Nvidia for the CPU division if the C/GPU divisions could be split. In the same spirit, Intel could make an argument to buy AMD's AP and GPU divisions to go more head-to-head against Nvidia.

All players in the mobile computing market basically.

 

[bryanW1995]

Most likely company to buy AMD is Intel. They hardly compete against each other any more, and nobody else wants to try going toe to toe vs Intel in x86.

The ITC would not allow this, I suspect.

Why would the ITC get involved in a theoretically proposed merger between Intel and AMD? Did you mean the FTC?

Regardless, nobody in the world wants to go up against Intel in x86 production, including AMD...while Intel seems like they would be the natural suitor for AMD's gpu division. It's a shrinking market that is surrounded by many fast-growing ones, I'd bet that Intel and AMD could figure out a way to get approval from the FTC/ITC/whomever for a deal if they really wanted one.

 

[RampantAndroid]

Why would the ITC get involved in a theoretically proposed merger between Intel and AMD? Did you mean the FTC?

Regardless, nobody in the world wants to go up against Intel in x86 production, including AMD...while Intel seems like they would be the natural suitor for AMD's gpu division. It's a shrinking market that is surrounded by many fast-growing ones, I'd bet that Intel and AMD could figure out a way to get approval from the FTC/ITC/whomever for a deal if they really wanted one.

ITC/FTC/DOJ, yes. ITC was involved with MS/B&N for example.

Sirius and XM were not allowed to combine; separate, they'd both die. Combined, they could continue...but combining was deemed to be a monopoly....

I think it happened in the end, but was blocked for ahwile.

 

[bryanW1995]

Be sold would be a big waste of opportunities for AMD. Being sold at a low price(due to current market value) just to fill shareholders pockets and end with many good R&D projects is not a good idea IMO.

Buyer surely can absorb many things of AMD with the deal. Best buyer would be IBM or Qualcomm, and IMO only the second may have interest in the deal...

Current market value? This is the 3rd time in 8 years that the stock has dropped down to $2 a share. The market cap is less than $1.8b. If there was enough of an upside then an activist hedge fund would get involved with them, but there's just not much they can do.

Man, this is depressing, I've always been a big AMD fan. I wasn't aware of how bad things had gotten there. AMD's only hope appears to be getting acquired by someone (anyone).

http://finance.yahoo.com/echarts?s=AMD+Interactive#{"range":"10y","showPrePost":false}

Don't be so naive. If work or science came first for AMD they would become a NGO or a research centre, they wouldn't be the for-profit business they are today.

Since when has AMD been a "for profit" business?

 

[RampantAndroid]

Since when has AMD been a "for profit" business?

You know they're publicly traded???

 

[scannall]

You know they're publicly traded???

You may want to get your sarcasm detector looked at.

 

[lamedude]

Even so, all the Pentium (P5) patents expired 2 years ago, yet there's no rush to get into the x86 space.

Win8 requires SSE2 and VS2012 defaulting to SSE2 has caused some problems for AthlonXP users in the FOSS world. I don't see a mostly Windows compatible CPU going far.

 

[Valantar]

You are right, the popularity is because of Windows. But aside from the failed Surface RT, it's the only ISA they write an OS for. At that point, it doesn't matter if Intel is popular or not since you really have no choice. OS X is on Intel as well. Though iOS is a subset of OS X and runs on ARM.

As such, it isn't really Windows any more. It's the fact that Windows has all the applications that people want, including all the ones they've been using for ages. This was what killed RT.

MS is making W10 for ARM - W10 Mobile has Continuity, with a desktop, full featured universal apps, and the like. But no matter what, it won't run whatever 5-10-15-year-old x86/x64 application you might want it to. Unless Intel somehow starts licencing x86 for emulation, in which case - hooray! - you'll be able to run it, it'll just be incredibly buggy and slow.

x86 is here to stay, it'll be very relevant for at least another 10+ years. My guess is far more, considering the way Intel is proving there is such a thing as fast, power-efficient x86. Heck, they're even pushing x86 into phones!

The PC space needs AMD. The 'Intel v. ARM' thing is growing stale, and for the moment, Intel is even winning that. And ARM isn't looking like it'll scale up power-wise any time soon.

 

[DrMrLordX]

To me, it more looks like, Intel is having problems pushing x86 into the phone/tablet space without subsidy, but ARM is failing utterly and completely at pushing into the server sector where it needs to go to expand.

Given the fact that the server sector seems to be a bit healthier in terms of revenue growth than the phone/tablet sector, I'd say Intel's position is the better one overall. So yeah, you could say they are "winning" there.

 

[monstercameron]

To me, it more looks like, Intel is having problems pushing x86 into the phone/tablet space without subsidy, but ARM is failing utterly and completely at pushing into the server sector where it needs to go to expand.

Given the fact that the server sector seems to be a bit healthier in terms of revenue growth than the phone/tablet sector, I'd say Intel's position is the better one overall. So yeah, you could say they are "winning" there.

Define utterly failing?

 

[ShintaiDK]

Define utterly failing?

What progress did ARM make into servers? Oh right, none.

X-Gene was the latest demostration of the complete flop it is.

x86 is king.

 

[cbn]

MS is making W10 for ARM - W10 Mobile has Continuity, with a desktop, full featured universal apps, and the like. But no matter what, it won't run whatever 5-10-15-year-old x86/x64 application you might want it to.

Too bad (at this time) if the Windows mobile device had x86 processor it will not run whatever 5-10-15 year old x86/x64 desktop app either. (re: To run in that Continuum desktop mode the app must be made into a universal app....which is a real shame because I don't think many of our traditional apps will ever make it into that MS app store.)

Now maybe if Intel makes a mobile processor which is in great demand, MS may decide to break part of that x86 desktop app compatibility problem in order to garner more sales.

 

[DrMrLordX]

Define utterly failing?

Still no notable market share to report, unless it's all being hidden behind opaque ODM deals that the customer(s) want kept quiet.

We keep getting reports that ARM will take n% of the server marketshare by <insertyearhere> without actually being shown which product or products will be the ones taking over that market share. The latest blurb in that vein I saw was this one.

You've still got Applied Micro and Cavium out there, doing what they do, but so far, it hasn't amounted to much. HP announced that they were shipping 64-bit ARM servers last September, and Lenovo announced an ARM prototype server this past February. Other than that, all's quiet on the ARM front. The ARMy hasn't been able to budge the usual server market share numbers reported by IDC and suchlike. Either that, or they haven't seriously tried yet.

AMD saw how things were shaping up for the ARM server market and acted accordingly, by delaying K12 and axing SkyBridge.

All those 64-bit ARM products might well be wonderful, it's just that there isn't much to indicate widespread (or even significant) adoption of existing ARM server tech. There is the odd possibility that ARM servers are shipping in quantity to ODM customers on the down-low, but it's hard to track that stuff. You would have to guess which CPU supplier was fulfilling those orders and connect dots from there.

 

[superstition]

Not sure why anybody thought that AMD's license wouldn't transfer if they were bought out, though. There's no way Intel would be allowed to be the sole major x86 supplier.

VIA is still releasing vaporware (e.g. Isaiah 2).

Anyway... I think it's droll that the "good faith" thing is always brought up, like it will magically force Intel to play nicely and not abuse its power &#8212; as if the government can be relied upon to do the right thing out of its inherent altruism.

Anyway... I wish Apple would return to its innovative mindset (now long gone) and acquire AMD. But, that's won't happen.

 

[DrMrLordX]

Anyway... I wish Apple would return to its innovative mindset (now long gone) and acquire AMD. But, that's won't happen.

I like being able to actually buy AMD processors, so no thanks.

 

[myocardia]

I like being able to actually buy AMD processors, so no thanks.

If Apple bought AMD, you'd still be able to buy AMD processors. The single-core, cache-free Semprons would be $999, though.