Gideon
Supreme [H]ardness
- Joined
- Apr 13, 2006
- Messages
- 4,189
X64 is owned by AMD and licensed to Intel for use. AMD may not be obligated to give those rights to another company, depends on the wording in the contract.
![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
-
Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.
Qualcomm bid for Intel
- Thread starter philb2
- Start date
X64 is owned by AMD and licensed to Intel for use. AMD may not be obligated to give those rights to another company, depends on the wording in the contract.
- Joined
- Mar 1, 2003
- Messages
- 2,610
That would be the most beautiful train wreck to watch.
Intel is a company that has escaped major disaster for a long long time. I am not gong to lie I am enjoying them laying in the bed of their making at this point. It sucks for competition, but then Intel has NEVER played fair anyway. Perhaps in a way its better for competition industry wide if Intel was no more.
They couldn't meet demand way back in the 486 days... so they licensed a handful of companies to make x86 chips. Every single one of them got deals with big legal poison pills in them. Intel annoyed with AMD success in '91 tries to strangle them... get sued by AMD, and AMD wins securing them their first royalty free license. (cause the court found Intels behavior gross enough to justify zero royalties).
Cryix white rooms x86.... Intel sues them. Cyrix wins after proving there designs were their own. In the court fillings Intel ends up with detailed analysis and design plans for Cyrix product. Intels Pentium is having issues with basic math... so what does Intel do, well they have the plans for that far superior Cyrix core. Hmmm lets just ape their design and call it the Pentium Pro. Cryix sues... and LUCKY for Intel after years of court battles and customers staying away from buying Cyrix due to them being painted as a copy cat, Cyrix needs the money. So accepts a paltry little settelment, but scores a lifetime Intel royalty free cross patent. OF course Intel inserts another contractual poison pill, cause you see that agreement states Cryix must use Intel licensed fabs. So when Cyrix doesn't die as expected and instead National Semi buys them. Intel drops all their previous National Semi contracts... meaning they can't fabricate Cyrix in their new state of teh art fabs as they intended. (National wasn't actually stupid they had new cutting edge Intel licenced fabs... Intel got into a legal fight wiht Semi cancelling all their intel deals to prevent them fabbing Cyrix) Instead Cryix is forced to fabricate at IBM... who Intel also sues (and looses). End result Cyrix is sold off to Via... who now through a 15% ownership in Zhaoxin is fabricating royalty free x86 chips in China. Which has contributed to Intel loosing what used ot be a healthy market selling outdated Intel product too.
Pentium 4... Intel used what was left of their 486 marketing cash to survive the netburst screwup. They tried to do battle with x86 extensions like MMX... and in the end their mis step meant x86-64 was developed by AMD and forced Intel to stop trying to extend their way out of competition.
Then they came back with a revised core and new branch predictor. Which eveyrone remembers fondly. Intel again got lucky no one noticed they cheated on the design of those cores. If Specter and Meltdown had been discovered when the Sandy bridge cores were a year or two old and not 5 years old... they would be remembered a lot differently. Turns out AMDs contemporary chips which are not well rememebred, would have been rememerbed much differently if Sandy had took a 40% performance hit in a security patch while it was current. Intel was also saved by spending a ton of cash to keep researchers quite. Intel is lucky the right researchers found those issues.. those exploits for years prior to the public announcements and fixes. IMO it was only ever announced cause at least 5 independent researchers had figured it out. The first security paper on cache timing was published in 1995. Apple added kernel space layout randomization way way back in 2012. The exploits were officially "discovered" in '16... and it wasn't till '18 that it was made public after giving Intel a chance to at least mostly mitigate the issue on a couple generations of CPUs. Intel was able to cover their own idiotic design... which probably existed cause some engineer needed to hit a desing performacne target and no one questioned how they did it.
Then there is Pat and the fabs... announcing your accepting fab business, while also competing on the design and sales side with almost every potential customer was just plain insane. Really the balls it takes to even think that was a good play. Imagine if Coke built extra bottling factories in every market and then announced to investors HEY we are going to transiton to a fabrication company... we are looking for other soda companies to bottle their product in our plants. Great idea right ? They are going to line up cause everyone knows Coke factories are the best soda factories. How could they refuse us? Combine that with the obvious issues with 13th and 14th gen... cause you know why not push 400 watts through a consumer CPU. Why not everything else in the world is disposable why not CPUs?
I'll turn the rant off... my only point is, Intels downfall is 100% of their own making, and might actually be deserved. If they fall much further someone will buy them up even if x86 dies, they hold enough other interesting IP. The fabs will get spun off no matter what its already Intels own plan if no one upends that plan. IMO Intel fabs are destined to travel the same road Global foundries did. They aren't likely to attract big high volume production. They will end up being boutique / old process volume production much the same way GF ended up going.
They couldn't meet demand way back in the 486 days... so they licensed a handful of companies to make x86 chips. Every single one of them got deals with big legal poison pills in them. Intel annoyed with AMD success in '91 tries to strangle them... get sued by AMD, and AMD wins securing them their first royalty free license. (cause the court found Intels behavior gross enough to justify zero royalties).
Cryix white rooms x86.... Intel sues them. Cyrix wins after proving there designs were their own. In the court fillings Intel ends up with detailed analysis and design plans for Cyrix product. Intels Pentium is having issues with basic math... so what does Intel do, well they have the plans for that far superior Cyrix core. Hmmm lets just ape their design and call it the Pentium Pro. Cryix sues... and LUCKY for Intel after years of court battles and customers staying away from buying Cyrix due to them being painted as a copy cat, Cyrix needs the money. So accepts a paltry little settelment, but scores a lifetime Intel royalty free cross patent. OF course Intel inserts another contractual poison pill, cause you see that agreement states Cryix must use Intel licensed fabs. So when Cyrix doesn't die as expected and instead National Semi buys them. Intel drops all their previous National Semi contracts... meaning they can't fabricate Cyrix in their new state of teh art fabs as they intended. (National wasn't actually stupid they had new cutting edge Intel licenced fabs... Intel got into a legal fight wiht Semi cancelling all their intel deals to prevent them fabbing Cyrix) Instead Cryix is forced to fabricate at IBM... who Intel also sues (and looses). End result Cyrix is sold off to Via... who now through a 15% ownership in Zhaoxin is fabricating royalty free x86 chips in China. Which has contributed to Intel loosing what used ot be a healthy market selling outdated Intel product too.
Pentium 4... Intel used what was left of their 486 marketing cash to survive the netburst screwup. They tried to do battle with x86 extensions like MMX... and in the end their mis step meant x86-64 was developed by AMD and forced Intel to stop trying to extend their way out of competition.
Then they came back with a revised core and new branch predictor. Which eveyrone remembers fondly. Intel again got lucky no one noticed they cheated on the design of those cores. If Specter and Meltdown had been discovered when the Sandy bridge cores were a year or two old and not 5 years old... they would be remembered a lot differently. Turns out AMDs contemporary chips which are not well rememebred, would have been rememerbed much differently if Sandy had took a 40% performance hit in a security patch while it was current. Intel was also saved by spending a ton of cash to keep researchers quite. Intel is lucky the right researchers found those issues.. those exploits for years prior to the public announcements and fixes. IMO it was only ever announced cause at least 5 independent researchers had figured it out. The first security paper on cache timing was published in 1995. Apple added kernel space layout randomization way way back in 2012. The exploits were officially "discovered" in '16... and it wasn't till '18 that it was made public after giving Intel a chance to at least mostly mitigate the issue on a couple generations of CPUs. Intel was able to cover their own idiotic design... which probably existed cause some engineer needed to hit a desing performacne target and no one questioned how they did it.
Then there is Pat and the fabs... announcing your accepting fab business, while also competing on the design and sales side with almost every potential customer was just plain insane. Really the balls it takes to even think that was a good play. Imagine if Coke built extra bottling factories in every market and then announced to investors HEY we are going to transiton to a fabrication company... we are looking for other soda companies to bottle their product in our plants. Great idea right ? They are going to line up cause everyone knows Coke factories are the best soda factories. How could they refuse us? Combine that with the obvious issues with 13th and 14th gen... cause you know why not push 400 watts through a consumer CPU. Why not everything else in the world is disposable why not CPUs?
I'll turn the rant off... my only point is, Intels downfall is 100% of their own making, and might actually be deserved. If they fall much further someone will buy them up even if x86 dies, they hold enough other interesting IP. The fabs will get spun off no matter what its already Intels own plan if no one upends that plan. IMO Intel fabs are destined to travel the same road Global foundries did. They aren't likely to attract big high volume production. They will end up being boutique / old process volume production much the same way GF ended up going.
Geez. No end of ideas here. Think that Intel's or AMD's lawyers should read this thread?
A feel the played fair or not narrative is a bit of a strawman, what company in position of not playing fair always do.... (Was IBM playing a more fair game when they were the big guys...)
https://investors.coca-colacompany.com/about/coca-cola-system
They are really good at putting liquid in bottle and distributing them enough to sell it as a service,
https://en.wikipedia.org/wiki/Coca-Cola_FEMSA
I do not imagine it is current to for them to bottle pepsico product too, but a energy drink, water, regional juice or Dr. Pepper company sure. But they did split the company between the design coca and bottlers (called independent coca-cola bottlers), quite similar to some of Intel plans.
On cloud, database and virtual machine providers.... not among regular folks and it is not like AMD (or Arm or IBM) was impact free from Specter.
Coke is the biggest bottler in the world by far, that the famous coca-cola system:
https://investors.coca-colacompany.com/about/coca-cola-system
They are really good at putting liquid in bottle and distributing them enough to sell it as a service,
https://en.wikipedia.org/wiki/Coca-Cola_FEMSA
I do not imagine it is current to for them to bottle pepsico product too, but a energy drink, water, regional juice or Dr. Pepper company sure. But they did split the company between the design coca and bottlers (called independent coca-cola bottlers), quite similar to some of Intel plans.
Most certainly not., I'm old enough to remember the saying, "You never got fired for buying IBM." And I remember some of their sales tactics, like going to the board of directors when the head of IT (or what they were called at the time) wanted to buy Storage Technology plug-compatible disk or tape drives that would plug into an IBM CPU channel controller, Worse yet an Amdahl CPU that replaced the entire IBM CPU and eliminated the possibility of that customer buying IBM drives.
Every company pushes the limits of what is legal sure. I think my point is everyone suggesting Intel must survive for "competition" sake might be a bit off on that one. Intel has been massively anti competition. AMD only exists cause investors were willing to bank roll years of legal fights with Intel. Everyone else that was an actual potential has been choked out. I didn't even get into all the third party chipset companies Intel chased off. Remember back when you could buy a MOBO with a third party chipset. I remember having an Nvidia nforce chipset at one point. Perhaps... Intel dying is what competition actually needs. Maybe it is time for actual real competition again like back in the 90s... when there were actual alternatives like MIPS chips around. Sure maybe ARM becomes the defacto if x86 dies... maybe someone builds a super Risc-V chip. Perhaps windows pushes Microsoft ARM hardware and makes x86 widnows second class. Intel imo has been bad for the industry for 20 years.
On meltdown and specter... AMD took a 3% performance hit sure.
https://www.extremetech.com/internet/291649-intel-performance-amd-spectre-meltdown-mds-patches
Ok lol I didn't expect the coke turn there... perhaps a bad analogy. Then again it also sort of makes the point I guess. Your right Coke bottles but for smaller places like a boutique producer.
Qualcomm’s potential acquisition of Intel would likely encounter thorough regulatory review, especially due to Intel’s U.S.-based foundry operations that produce chips for defense applications. As of this month, 2024, Intel reportedly qualified for as much as $3.5 billion in federal grants to make semiconductors for the Defense Department. It’s unclear whether Qualcomm would want to keep Intel’s manufacturing business, which has been a financial burden. Qualcomm has experience navigating such challenges, as seen in its unsuccessful two-year attempt to acquire Dutch semiconductor company NXP in 2016, largely due to delays in China’s approval process.
https://epsnews.com/2024/09/23/intels-future-uncertain-as-challenges-mount/
https://epsnews.com/2024/09/23/intels-future-uncertain-as-challenges-mount/
I am guessing the difference with coke is that bottling is low capital compared to fabs.
Remember real men had fabs comment.
(Yet things turned in 2006 when AMD acquired ATI & Nvidia released CUDA)
I believe Coke franchises/outsources its bottling requirements. In India they even sued a 'bottler' who didn't cease the production of their inhouse cola brands
yes pretty much all of it the top coca-cola company make the sirups concentrate, but many bottlers that win their business are called coca-cola bottling Europe, coca-cola bottling south america, etc.... say:
https://en.wikipedia.org/wiki/Coca-Cola_Europacific_Partners
Which is not coca-cola, but in part owned by them (sometime they own the majority like Coca-Cola bottlers Africa), it is a bit well known because Buffet own a lot of Coke and talked about it over the years.
Unlike Intel model, Coca-cola mostly acquired part of already existing bottler all around the world to become a giant in it.
Last edited:
- Joined
- Jul 1, 2004
- Messages
- 8,440
is Qualcomm capable of running a CPU business? could they make it worse?
Qualcomm Said to Wait for US Election to Decide Intel Move
Qualcomm’s deliberations are ongoing and there’s no certainty that the company will decide to pursue an offer for Intel and the timing could change, the people said. Representatives for Qualcomm and Intel declined to comment, while the State Administration for Market Regulation in China didn’t respond to requests for comment.
DukenukemX
[H]F Junkie
- Joined
- Jan 30, 2005
- Messages
- 11,276
That's scary that they're waiting after US elections. They afraid to alter the election, or waiting to see which candidate wins and would favor their acquisition?
Probably want to know if Lina Khan (it is a particularly harsh FTC at the moment) is still there or someone else, as it would be something the FTC would look for sure. The process would be so long that it could not affect the almost right now election.
Last edited:
It has nothing to do with the election, they can frame it that way to the public and be all like OOOh we have concerns there.
Really it is all about the Q3 financial reports which will come out after the election, if they are bad and the stock goes down, Qualcomm has the potential to save Billions on its offer, but if they make the offer now, even if the financials are bad the stock prices would be propped up by their offer.
So say "We'd like to wait until after the election" as a means of deflecting the actual issue in an attempt to get a better deal later, because they can't just come out and say "We're waiting for the price to dip lower".
On the other hand if they wait for the next Intel quarterlies. They might end up dealing with competing offers. Qualcomm isn't the only company contemplating buffing up their patent library.
It is a short list of companies that would be allowed to purchase Intel, and that's not a secret.
The Military contracts Intel holds, and its place in the US strategy to update its own electronic manufacturing sector make it something that will be very tightly controlled.
Broadcom is a big no, as they were denied the ability to purchase Qualcomm previously because of their supposed Chinese connection
https://www.cnbc.com/2018/03/12/tru...ting-broadcoms-bid-to-take-over-qualcomm.html
Nvidia would be a big no, as would AMD (but fuck that would be funny), Amazon, Apple, and Microsoft, all no because of conflicts of interest.
The big banking conglomerates who could reasonably purchase Intel would likely be denied for similar reasons to Broadcom.
There is a small chance that IBM could put forward the capital to purchase Intel if they wanted to go that route, which would be fun, but there are not many companies that have the money and the lack of conflicting interests to have a hope of purchasing Intel.
I don't think your analysis is quite correct. Apple I don't see the upside for them no... and as far as anti compete goes, maybe its shot down, maybe not honestly. I don't see why they would be interested. Microsoft is 100% interested... and it would go through as well. Yes AMD would complain but I don't see why it would get stopped either. Amazon... no conflict there. I don't see why they would be interested either, I also don't see why it wouldn't pass regulatory issues. Nvidia... really don't see what holds that sale up frankly. Intel is not competing with Nvidia really... Nvidia has held off on going hard into CPU markets. I think they were holding off ARM moves until they owned ARM. Once that sale fell through I find it interesting their CPU moves have been small to say the least. Maybe they have a consumer mediatek part in the works. Really it all seems like stand by work. I do believe Jensen is going to make an offer on Intel design. He doesn't want their fabs. Intel will continue to exist... which makes regulatory interference a lot less likely. Intel will just be a 100% fab company at the conclusion of the sale. For Nvidia they add x86 for when the industry moves on mass to inference.
No one is in the market for the whole shebang. Intel already set the table for the company to split in two. The offers Intel shareholders may be looking at won't actually include a wholesale sale. Should Intel post disappointing quarterlies... expect a lot of odd offers. If Qualcomm is waiting imo its more to see what other offers are coming. The first company to make an offer is going to end up having to rebid imo.
Nvidia is playing ARM smart by being quiet, the ARM - Qualcomm lawsuit kicks off Dec 16, 2024, and depending on how that goes Nvidia may soon become the only non-Apple 3'rd party still allowed to sell customized ARM SoCs (Nvidia got a 40-year extension to its license when the sale fell through).
After Nvidia failed to purchase ARM, ARM took the list of things that everybody feared Nvidia would do as a personal challenge and updated their licenses accordingly. So should the Qualcomm lawsuit fall through, Qualcomm, Broadcom, Samsung, AMD, MediaTek, and everybody else other than Apple will lose their ability to design their custom ARM solutions once their licenses expire, and those expiry dates are fast approaching (2028 is the last of them I believe).
Nvidia's attempt to purchase Intel would be blocked for the same reasons its purchase of ARM was blocked with a few extras sprinkled in for good measure.
Amazon and Microsoft would be denied because of their data center operations alone, either of them owning Intel could allow them to prioritize their operations over those of rivals.
Last edited:
Nvidias arm purchased was denied because ARM licenses nvidias direct competition. In what way is Intel Nvidia competition? AI? Nviidas legal will laugh that idea out of court. What Nvidia competition relies on Intel ? No one. As crazy as it sounds Nvidia is more likely to get approval buying Intel design then ARM.
While I agree somewhat, I think the potential Monopoly that could exist if Nvidia owned Intel would scare regulators to the point that they would run the clock out on the deal.
Say Nvidia did buy Intel, and then shortly after ARM wins the Qualcomm lawsuit.
Then we have Nvidia with an overwhelming GPU lead, a commanding AI lead. The only 3’rd party ARM licensing rights, the contract to manufacture the ARM chips in the Intel fabs. And to top it off the dominant (by volume) x86 platform it would give them such an overwhelming market force that if they even looked in a direction any market authority considered anti competitive they would face calls to split them up.
There’s too much in the air for Nvidia to buy Intel and not receive severe market pushback.
Regulators would try to force Nvidia to make legal agreements saying they wouldn’t do a bunch of things that would limit their ability to unify their platform and ultimately hinder their ability to properly utilize Intel.
"In an unusual effort to aid an electronics company that supplies it with vital parts, the International Business Machines Corporation announced yesterday that it will pay $250 million to purchase 12 percent of the Intel Corporation." This was in 1982. In 1981 IBM used the Intel 8088 CPU in its PC product.
https://www.nytimes.com/1982/12/23/...82 capital expenditures of about $130 million.
True. OF course that assumes Nvidia wants Intels fabs. I doubt highly they want Intel fabs. No one wants intel fabs. That is what makes these waters dangerous for Intel imo. No one is in the market for INTEL... they only want Intel.
There is also the possibility of interesting investment deals where a company like Nvidia just flat out says its buying 40% of Intel... with plans to heavily invest in the Fab business IPO when it happens. (and expect some NV-Intel tech team ups out of the deal)Intel continues in almost every ones potential offer... as a fab company. I don't think anyone is offering to buy Intel whole including the fabs. I would not be shocked to see two companies team up for that matter. I could think of a few scenarios were a company like Qualcomm makes an offer for only part of the design business, and another company offers to pick up the other half / third what have you.
Nvidia buying the whole thing? Fab included. I could see that having regulation issues. On the other hand Intel has missed A LOT of gov contract time tables. Nvidia might get some unexpected support if they were willing to take on Intel fab. Its possible they US gov sees Nvidia taking over management of the fab as a better solution then trying to fix Intels obvious broken culture. Pat was supposed to fix the fab business... the gov might be fine with Jensen taking a shot at that. It also means Nvidia is far more likely to fabricate future AI parts in the US. I know the regulatory body is not the gov dept with outstanding Intel silicon contracts... I think their might be pressure to push that deal through though if Jensen says he wants to push a US fab and make it #1.
- Joined
- Jul 13, 2006
- Messages
- 2,243
This is a fairly epic rant. Bravo
- Joined
- Jul 13, 2006
- Messages
- 2,243
Why is Nvidia a no?
- Joined
- Jul 13, 2006
- Messages
- 2,243
My understanding is that having a monopoly and abusing the monopoly are two different issues. You can have a monopoly and be fine as long as you don't abuse the market share to keep out competition.
But that is for breaking up existing companies. I don't fully understand laws regulating companies being bought.
Building yourself a platform or product that becomes a monopoly then not abusing it is one thing. But buying yourself one is another, regulators tend to block them running the assumption that they are doing it with the express desire to abuse it.
Look at the agreements that they forced Microsoft into for the Activision Blizzard purchase as a starting point and that wasn’t even putting Microsoft in a monopolistic position, it was just a big purchase.