Nvidia's Plan for ARM to Take X86 's Throne | Exclusive

For all the jocularity. Lets be clear. nVidia has a freakin good track record with the switch and I can't imagine Nintendo is anything but happy ATM.

I do wonder if they can find a ballance between the new power hangry GPU and whatever denver morphs into...
 
FLECOM said:
member when MIPS was going to take the throne away from x86?

member when PowerPC was going to take the throne away from x86?

member when SPARC was going to take the throne away from x86?

member when Alpha was going to take the throne away from x86?

member when PA-RISC was going to take the throne away from x86?

member when RISC-V was going to take the throne away from x86?
Click to expand...

VECTOR guys said that old-RISC (MIPS, Power, SPARC, Alpha, PA-RISC) couldn't replace them. It is impossible, they said.

Old-RISC guys (MIPS, Power, SPARC, Alpha, PA-RISC) said that x86 couldn't replace them. It is impossible, they said.

d-by-RISC-microprocessors-in-turn-displaced-by-x86.png

x86 guys say now that ARM will never replace x86. It is impossible, they are saying us.

It is kind of Ironic how history repeats. :D

FLECOM said:
Microsoft used PowerPC for the xbox 360 after switching away from x86 in the original xbox... and back to x86 they went
Click to expand...

Microsoft returned to x86 because ARM was only 32 bits when they started the design of the PS4.

FLECOM said:
the microsoft surface RT was also arm based, how did that go?

unless microsoft can make their own version of windows 10 with an x86>ARM emulation layer (rosetta2) that's as efficient as apples (and therefore bring over decades of x86 software), and the big guys (lenovo, hp, dell, microsoft) release some arm based hardware that's competitive i just don't see it happening

I see pcs just going away before x86 goes away, ironically replaced by arm devices
Click to expand...

Microsoft has ported Windows server to ARM and is using ARM servers internally. The world doesn't start and end at PCs
 
Last edited:
LukeTbk said:
Is that true, I feel that if x86 stay alive the next 100 years that would not necessarily mean ARM market share on smartphone, tablet, cars, television and so on would become nearly nill, ARM is so omnipresent, I think they are selling over 25 billions of them a year now, that more than 3 by humans on earth in just a year. And for many of those application, it is not obvious shifting to x86 would bring much (there is little legacy software useful for them and so on).

The talk about never (about C, or X86) do sound like really premature, if computing was 1,000, 2,000 year's old, statement like that could make sense, but thinking that in 3,000 year's C/C++ and X86 will still be significant and expressing it as a certain fact is pushing it, imo (or even just by the end of this century).
Click to expand...
We're talking more about human nature than technology. ARM could emulate x86 to some degree of success, but so can x86 emulate ARM. Realistically neither can emulate each other 100% effectively without some specialized hardware. We know that x86 is not really CISC anymore, but RISC with built in CISC to RISC conversion. So x86 can evolve in many ways. AMD, who is not entirely affiliated with x86, can go ahead and build a ARM/X86 core that can effectively run both really well. AMD can build a ARM core with a x86 conversion. Intel could as well but we know Intel isn't a big fan of ARM.

The main problem with x86 is licensing, as in nobody else can make x86 CPU's. Even if they could, they likely won't get the latest design that Intel has and would need to engineer the CPU from scratch. This is the main appeal of ARM since anyone can buy the license and use their latest design. This is why ARM is in smartphones, tablets, cars, televisions, and etc. Not because it's good but because it's good enough. Nvidia will certainly push ARM to the performance level of x86, but I'm not sure how they will handle licensing. Nvidia is not historically known for sharing technology, even to license.
 
DukenukemX said:
We're talking more about human nature than technology. ARM could emulate x86 to some degree of success, but so can x86 emulate ARM. Realistically neither can emulate each other 100% effectively without some specialized hardware. We know that x86 is not really CISC anymore, but RISC with built in CISC to RISC conversion. So x86 can evolve in many ways. AMD, who is not entirely affiliated with x86, can go ahead and build a ARM/X86 core that can effectively run both really well. AMD can build a ARM core with a x86 conversion. Intel could as well but we know Intel isn't a big fan of ARM.

The main problem with x86 is licensing, as in nobody else can make x86 CPU's. Even if they could, they likely won't get the latest design that Intel has and would need to engineer the CPU from scratch. This is the main appeal of ARM since anyone can buy the license and use their latest design. This is why ARM is in smartphones, tablets, cars, televisions, and etc. Not because it's good but because it's good enough. Nvidia will certainly push ARM to the performance level of x86, but I'm not sure how they will handle licensing. Nvidia is not historically known for sharing technology, even to license.
Click to expand...
Intel regularly licenses the x86 ISA, and AMD regularly licenses the x86-64 ISA (including to Intel).

The trick here is the long tail of software and compatibility. We'll have to see what happens - and not just in the low-power space where ARM is competing well, but in the full-fat server space as well (where ARM tends to, at least historically, use the same power and TDP as x86 CPUs while offering little performance benefit - obviously pre M1 and Nvidia purchase, of course).
 
All the talk of emulation is a bit silly at this point in the software world. Emulation isn't required. Old software is legacy and sure perhaps they can add a emulation for it so you can still run it with the expectation that some stuff will run like ass.

However Apple has show if you force your developers to use the proper tools.... you can just recompile at install. Write your software for X framework... and the OS can recompile those libraries for any arch it supports.

MS has been having a hard time getting their developers onboard with their newer frameworks... but that is exactly why they where pushing it. Apple forced it and it looks like it works exceptionally well.

The windows world is going to follow because MS is already itching knowing their surfaces now look like even more hot garbage compared to what Apple is doing. They just need a proper chip to compete with M1. Yes Nvidia sees it coming... no other reason to spend what they did picking up ARM. They could have made a chip with just a license. They are looking for the mind share that comes with being the modern day "intel" controlling the cloners licenses. Its all about the mind share.... they want ARM to = Nvidia. And considering what they paid for that.... I have ZERO doubt they already have a new chip in the works, and it may be a lot closer to prime time then people would think. I suspect by this time next year we will be talking about M2 vs Nvidias new super duper Windows friendly octo Core ARM + Ampere+ graphics solution.
 
To follow on what I posted above. Apple isn't doing anything software wise that the larger industry can't do. Linux has had distros like this for years... where everything you install compiles when you install it. Those distros and the majority of their software repositories tend to compile just fine on almost any arch. x86 ARM Power MIPS. If your compiling from source and libs the target arch isn't really a big deal. Sure specific things run faster on X or Y arch but they compile and run fine.

Being based on BSD, Apple acts a lot like a Linux distro... them adding a install time compile option was actually pretty trivial, making sure developers where using the same frameworks simply makes it a 1000% more consistent. Projects like Arch Linux ARM and Gentoos ARM distro run quite well, and for the most part they just compile from source after changing the target arch and a bit of meta data here and there. MS just needs to push more of their big software vendors to conform a bit more. (and really I think most are already where they need to be)

If Nvidia where to build a ARM chip / MOBO setup that we are used to seeing in the PC world with PCIx slots / M.2 storage options ect I do think things like gaming on ARM are not at all out of the question.
 
lopoetve said:
Intel regularly licenses the x86 ISA, and AMD regularly licenses the x86-64 ISA (including to Intel).
Click to expand...
Who does Intel license their x86 tech to? I don't see too many IBM or Cyrix x86 CPU's anymore. There is Via with Zhaoxin but we know that isn't going anywhere.
 
DukenukemX said:
Who does Intel license their x86 tech to? I don't see too many IBM or Cyrix x86 CPU's anymore. There is Via with Zhaoxin but we know that isn't going anywhere.
Click to expand...
Couple of Chinese companies, AMD, various virtualization companies. Few embedded processor manufacturers. License price used to be public, don’t know if it is anymore. There was a bit of a hullabaloo a while back when it came out they had joint ventures and had licensed it to the Chinese.
 
DukenukemX said:
We're talking more about human nature than technology. ARM could emulate x86 to some degree of success, but so can x86 emulate ARM. Realistically neither can emulate each other 100% effectively without some specialized hardware. We know that x86 is not really CISC anymore, but RISC with built in CISC to RISC conversion. So x86 can evolve in many ways. AMD, who is not entirely affiliated with x86, can go ahead and build a ARM/X86 core that can effectively run both really well. AMD can build a ARM core with a x86 conversion. Intel could as well but we know Intel isn't a big fan of ARM.
Click to expand...

And still rumors point to being Intel who is developing a hybrid CPU that can run both x86 and ARM code. The same source that leaked the hybrid AMD GPU Intel CPU.

DukenukemX said:
The main problem with x86 is licensing, as in nobody else can make x86 CPU's. Even if they could, they likely won't get the latest design that Intel has and would need to engineer the CPU from scratch. This is the main appeal of ARM since anyone can buy the license and use their latest design. This is why ARM is in smartphones, tablets, cars, televisions, and etc. Not because it's good but because it's good enough. Nvidia will certainly push ARM to the performance level of x86, but I'm not sure how they will handle licensing. Nvidia is not historically known for sharing technology, even to license.
Click to expand...

Licensing is only one of the problems. The main problem is that x86 is an outdated and inefficient ISA.

ARM is in "smartphones, tablets, cars, televisions, and etc." because an equivalent x86 core would be much bigger, expensive, and power hungry.

mobile cpu core size.png
 
lopoetve said:
Couple of Chinese companies, AMD, various virtualization companies. Few embedded processor manufacturers. License price used to be public, don’t know if it is anymore. There was a bit of a hullabaloo a while back when it came out they had joint ventures and had licensed it to the Chinese.
Click to expand...
Intel did ?

No there is two Chinese companies that have or are making x86 chips.... one got a license deal with AMD (which the US gov shut down). The other has their own license from Via. Via for those that are not old enough or forgot bought Cyrix when they fell on hard times... thanks to Intel making its extremely hard for them to do business. However Cyrix clean roomed x86, they never licensed it... Intel sued them multiple times and LOST every time. The Cyrix chips owed NONE of their tech to Intel (legally proven) they simply understood x86 instructions. Cyrix ended up with a never ending license (which is why Via bought them)... because Cyrix sued Intel when the Pentium pro/II chips came out. They sued because its Intel that copied Cyrix designs, anyone familiar with x86 workings saw that not only did Cyrix have a strong case its pretty obvious that Intel had cracked a few Cyrix chips and realized their design was superior. They copied how Cyrix dealt with name registers (A method that is 10x faster and scalable... which the older Intel method was not)... they also copied Cyrixs internal power management. (something previous Intel chips didn't do at all, Intels implementation is almost identical to the last gen Cyrix chips) In fact now it IS the same... cause Intel knowing they where about to loose and probably owe Cyrix (or whatever company picked up its bones, had Intel bought them they would have been in trouble with anti trust regulators) Billions of dollars and probably royalties for as long as they used those designs. Today Intel and Via (Cyrix) have a cross license for the life of both companies. The license survived the death of Cyrix cause Intel is still using Cyrix patents.

Intel has never really licensed chip manufacturers to compete with them. Cyrix didn't ask they white roomed. AMD was licensed to make overflow chips for cheaper markets in a time when Intel couldn't keep up with computing chip demand in the 80s... and there where actual competing ISAs that would fill the void if x86 couldn't. AMD and Intel had tons legal fights... and much like Cyrix it got settled with cross license deals cause AMD did invent a ton of x86 tech not simply x86-64. The virtualization companies don't have chip design licenses.

Intel loved having competition so much they even made sure Nvidia couldn't make chipsets for x86. Intel HATES Nvidia... and Nvidia has zero love for Intel. Nvidia received massive payments for Intel for years over GPU tech. The idea that Nvidia wants to compete with Intel directly is a no brainer. Of course they do. Nvidia tried to wiggle into computing beyond the GPU and got shut down by Intel. I have no doubt they would love now to push ARM into a position to pound x86... and have Intel licensing from them.

EDIT - the only other major x86 play I remember that wasn't just intel trying to supply specific industrial markets held by other ISAs and was consumer facing was Transmeta. They also did not license anything from Intel. Based on the legal ground work laid by Cyrix and AMD... they simply created a hardware level x86 translator for their RISC type chip. They went out because frankly they where a decade ahead of their time. Had they came a little later they could have probably stolen all the early ARM mobile wins. They where just too early... no one cared that much about 40-50% better power efficiency for 20% less performance when they launched. Their tech also survives.... in fact on topic sort of Nvidia has a life time license for their longrun tech. (I'm pretty sure some version of it is still used by Nvidia GPUs)
 
Last edited:
FLECOM said:
member when MIPS was going to take the throne away from x86?

member when PowerPC was going to take the throne away from x86?

member when SPARC was going to take the throne away from x86?

member when Alpha was going to take the throne away from x86?

member when PA-RISC was going to take the throne away from x86?

member when RISC-V was going to take the throne away from x86?
Click to expand...


We’re all also using Linux because it took the crown from Windows too...
 
Crap, I had it backwards ChadD - AMD had/has licensed THEIRS, vs Intel doing the licensing. You're right.
https://www.infoworld.com/article/3...l-could-create-more-x86-rivals-for-intel.html

Derp. Apologies. I will say that the virtualization companies DO have far more advanced licenses than you suspect - but no interest in making silicon.

Agreed on the hate between Nvidia/Intel though. Nvidia tends to bully their way around as-is, so that definitely goes both ways :p
 
  • Like
Reactions: ChadD
like this
lopoetve said:
Crap, I had it backwards ChadD - AMD had/has licensed THEIRS, vs Intel doing the licensing. You're right.
https://www.infoworld.com/article/3...l-could-create-more-x86-rivals-for-intel.html

Derp. Apologies. I will say that the virtualization companies DO have far more advanced licenses than you suspect - but no interest in making silicon.

Agreed on the hate between Nvidia/Intel though. Nvidia tends to bully their way around as-is, so that definitely goes both ways :p
Click to expand...

x86 license deals have always been a mess. Mostly cause ya Intel doesn't really want to do it... they have only ever licensed when they had no other choice. It was never because they wanted to further x86 or had some dream of a world where anyone but them supplied CPUs. They dealt with AMD cause the alternative was strong offerings from IBM(Power) and Alpha could have filled the void, as they just couldn't produce enough chips to power the PC revolution alone. I can't remember who but I think they licensed x86 to a few other players in very niche markets... industrial control stuff in the 80s that type of thing. (nothing powerful or consumer facing)

You could be right about some of the virtual players... I suspect though Intel after their experience with AMD in the 80s has some pretty strong wording in their agreements with them that would make it difficult. Perhaps not impossible but very very difficult.

Intel and NV are both corp jerks no doubt. Ya Intel has always felt very suckered by the 1.5B they had to pay Nvidia over GPU tech they felt they invented. (and well they did lol) Nvidias purchase of 3DFX gave them access to patents from Real3D (early Military patents from the 70s/80s). Intel GAVE 3DFX those patents to settle a different dispute. Intel worked with Real3D for years on tech that became the Intel iGPUs when the military contractors decided to exit that play, they gave Intel all the patents... seemed the right thing to do they sold them to Intel for a fraction of their worth. Everyone assumes Nvidia bought 3DFX cause they wanted the tech or something... but anything they got on that end was just a nice to have. Really they wanted those Real3D patents cause they knew it was the basis of every Intel CPU with on board video sold. They thought they where going to pressure Intel into a x86 license, is my opinion. Intel was forced to pay Royalties.... cause it was preferable then having to compete against both AMD and Nvidia. Well seems like now there probably going to have to do both.... and it could end up being the end of x86 and perhaps even Intel. I know that sounds crazy... but hey IBM was once IBM... now there ibm (we also sell Linux folks).
 
Last edited:
I don't think Nvidia will ever actually acquire ARM. Too many regulatory hurdles.
 
juanrga said:
Linux took the crown on supercomputers, servers, phones, tablets... It is only on desktop where Windows is still the king.
Click to expand...


I was being sarcastic. BSD Unix and Linux made lots of in roads into the consumer space via Mac OS and Android respectively. I was referring to how every year we see articles about how “this is the year of the Linux desktop”. Unfortunately it’s never the case regardless of how much better and easier Linux has become.
 
Revenant_Knight said:
I was being sarcastic. BSD Unix and Linux made lots of in roads into the consumer space via Mac OS and Android respectively. I was referring to how every year we see articles about how “this is the year of the Linux desktop”. Unfortunately it’s never the case regardless of how much better and easier Linux has become.
Click to expand...
Desktop and consumer use cases are different (still) than mobile, and significantly different from server/etc.

Once someone skins Linux to look close enough to a mobile device, and convinces people thats all they really need...
 
FLECOM said:
Microsoft used PowerPC for the xbox 360 after switching away from x86 in the original xbox... and back to x86 they went

the microsoft surface RT was also arm based, how did that go?

unless microsoft can make their own version of windows 10 with an x86>ARM emulation layer (rosetta2) that's as efficient as apples (and therefore bring over decades of x86 software), and the big guys (lenovo, hp, dell, microsoft) release some arm based hardware that's competitive i just don't see it happening

I see pcs just going away before x86 goes away, ironically replaced by arm devices
Click to expand...

I would say judging the merits of the tech on how well Microsoft can implement it is a pretty crappy measurement.
 
Revenant_Knight said:
I think Apple will be very against it. They probably regret not pushing Open RISC right now.
Click to expand...

Apple already has everything they need from ARM in perpetuity, the rest of what they build with their chips is their own IP.

The deal will go through.
 
Mega6 said:
Chinese will bitch and complain like they do about everything and maybe not approve. It won't matter.
Click to expand...

Yeah no matter how they feel about it one way or another, they have no legal ownership of ARM so their opinion really means squat. Most they can do is say they won't allow future ARM sales in China, which would be shooting themselves in the foot and hurt them more than allowing it seeing as the amount of devices that use ARM, and can't just be switched over overnight. They create a counterfeit ARM, that leads to further economic sanctions and penalties against them. They're caught between a rock and a hard place with this. Sucks to be them.
 
  • Like
Reactions: Mega6
like this
ChadD said:
x86 license deals have always been a mess. (...) They dealt with AMD cause the alternative was strong offerings from IBM(Power) and Alpha could have filled the void, as they just couldn't produce enough chips to power the PC revolution alone.
Click to expand...
Oh, I'm sure they could have, but IBM didn't want to rely on just one manufacturer and forced Intel into bringing AMD in as a second source.

ChadD said:
Intel has always felt very suckered by the 1.5B they had to pay Nvidia over GPU tech they felt they invented.
Click to expand...
As far as I know that settlement wasn't over patents, it was nVidia suing Intel over anti-competitive practices --- as everyone else did at the time.

ChadD said:
Nvidias purchase of 3DFX gave them access to patents from Real3D (early Military patents from the 70s/80s). Intel GAVE 3DFX those patents to settle a different dispute. Intel worked with Real3D for years on tech that became the Intel iGPUs when the military contractors decided to exit that play, they gave Intel all the patents... seemed the right thing to do they sold them to Intel for a fraction of their worth. Everyone assumes Nvidia bought 3DFX cause they wanted the tech or something... but anything they got on that end was just a nice to have. Really they wanted those Real3D patents cause they knew it was the basis of every Intel CPU with on board video sold. They thought they where going to pressure Intel into a x86 license, is my opinion. Intel was forced to pay Royalties....
Click to expand...
But why would Intel pay royalties for patents they own? Intel didn't give 3dfx patents, they cross-licenced each others portfolios. I can't find information of Intel paying royalties to nVidia for anything. But certainly not those Real3D patents.
 
marginalerror said:
Oh, I'm sure they could have, but IBM didn't want to rely on just one manufacturer and forced Intel into bringing AMD in as a second source.


As far as I know that settlement wasn't over patents, it was nVidia suing Intel over anti-competitive practices --- as everyone else did at the time.


But why would Intel pay royalties for patents they own? Intel didn't give 3dfx patents, they cross-licenced each others portfolios. I can't find information of Intel paying royalties to nVidia for anything. But certainly not those Real3D patents.
Click to expand...

https://arstechnica.com/information...intel-processor-die/?comments=1&post=21198785
"NVIDIA wrote in to tell us that our original headline was not accurate. An NVIDIA spokesperson said, "Licensing a technology is different than incorporating an entire processor. The settlement provides Intel with access to our IP and patents, such as Sandy Bridge which already uses NVIDIA technology. The license enables Intel to extend that model for the next 6 years.""

https://www.computer.org/publications/tech-news/chasing-pixels/famous-graphics-chips-Intel740
"By late 1999, Intel did two things, shut down the i740 project, and acquired the assets of Real3D from Lockheed Martin. As Real3D crumbled, ATI hired many of the remaining employees and opened up an Orlando office. Prior to the sale of its assets to Nvidia, 3Dfx had sued Real3D over patent infringements. Intel settled the issue by selling all of the intellectual property back to 3Dfx, which ultimately ended up in Nvidia’s hands. "


So yes Nvidia ended up holding the patents that Intel had used when they co developed the i740 with tech originally from GE Aerospace... developed for consumer use with Intel and Read3D(a division of Lockheed Martin). Intel was still using those patents when Nvidia and Intel settled. Its the basis of every Intel integrated GPU since. i740->810->HD ect ect ect. Its all derived from their Real3D partnership... which Nvidia ended up owning. When Nvidia themselves said in 2001 "such as Sandy bridge which already uses Nvidia technology." that is what they are saying. Intel was using their own igpu... Nvidia made it clear that was their line of attack. The lawyers at Intel paid Nvidia off rather then giving them what they where really after.... Jensen said at the time, oh yes we are going into CPUs ARM CPUs cause the world already has too many x86 players??? I mean come on that was BS.... he didn't get his way he was angling for a x86 license and figured getting a hold of the patents Intel relied on for their igpus which at that point was the majority of their CPU line he could hold them over a barrel. I assume he realized Intel would rather see them in court and drag it out over a decade before allowing that and took the cash. Both companies at the time said they where happy... but why would Intel be happy about paying Nvidia 1.5 billion for their own tech ? And NV didn't really get what they wanted either. Fast forward to the present ya Nvidia can't wait to bring ARM chips that can compete vs Intel and beat them up... and Intel would still love to see Nvidia burn. The next few years should be good fun. :)
 
lopoetve said:
Desktop and consumer use cases are different (still) than mobile, and significantly different from server/etc.

Once someone skins Linux to look close enough to a mobile device, and convinces people thats all they really need...
Click to expand...


I just don't see that happening right now though. Microsoft is slowly moving closer and closer to making desktop Windows free, and the mind share that would be required to get people move away from it is staggering. Apple has been trying since...before Windows. We've had OS/2 Warp, Mac OS, BeOS, a half dozen really good Linux distros, and most recently Chrome OS. We've even had Steam OS for gaming. Getting past that mind share is going to take a monumental shift in society itself right now. I can only think of two companies that stand a chance in pulling that market share and they are Google and Apple. At least for the US and European markets.

I do think though that desktop computer may evolve to just become docking stations for phones and tablets. In that case, aside from artistic professional work, high security/scientific work, and games (for now), we might see Windows as a primary OS falter. Microsoft needed to really commit to the phone market, and they Zuned it.
 
staknhalo said:
Apple already has everything they need from ARM in perpetuity, the rest of what they build with their chips is their own IP.

The deal will go through.
Click to expand...

That's a big statement. What do you know that we don't? I don't recall Apple's ARM licensing contract being public. We don't know what they have, how long they have it, or what their plans are. Developing their own IP would literally mean switching architecture unless they have some sort of specialized contract. I'm not saying they don't, but unless you have access to that info, we don't know.

I suspect the deal will go through, but I still don't see a benefit for Apple with Nvidia now holding the keys to the ARM castle.
 
Revenant_Knight said:
That's a big statement. What do you know that we don't? I don't recall Apple's ARM licensing contract being public. We don't know what they have, how long they have it, or what their plans are. Developing their own IP would literally mean switching architecture unless they have some sort of specialized contract. I'm not saying they don't, but unless you have access to that info, we don't know.

I suspect the deal will go through, but I still don't see a benefit for Apple with Nvidia now holding the keys to the ARM castle.
Click to expand...
Yes Apple has a license for what they have now forever >.< The only thing Nvidia could try to do to screw them over would be to update a new edition of the ISA get the entire market to really support it... and then force Apple to license that. However seeing as Apple controls their entire stack they really could care less. If Nvidia did that they would be screwing themselves over not Apple. Considering the lock in Apple has with developers I feel if Nvidia started that fight they wouldn't like the outcome. Apple could extend ARMv8 all sorts of non official ways and force their extensions into the software world. IF Nvidia tried to start that war.

Apple is one of 15 companies that hold Architecture 64bit ARM licenses. AMD also holds one.... as does Qualcomm and Samsung. There is NOTHING Nvidia could do to screw them over. They have the rights to the ISA to use in their own designs forever. What they paid for that right... ya who knows pretty sure no one has disclosed how much they paid ARM for their Arch licenses.

The only companies Nvidia could really screw over license wise would be the smaller players who use pre built ARM cores... companies like say Mediatek.
 
Last edited:
juanrga said:
And still rumors point to being Intel who is developing a hybrid CPU that can run both x86 and ARM code. The same source that leaked the hybrid AMD GPU Intel CPU.
Click to expand...
According to Linus Tech Tips, that's how the M1 is achieving decent x86 emulation. In that it does have hardware that converts x86 code to ARM.

Licensing is only one of the problems. The main problem is that x86 is an outdated and inefficient ISA.

ARM is in "smartphones, tablets, cars, televisions, and etc." because an equivalent x86 core would be much bigger, expensive, and power hungry.
Click to expand...
If Apple can make the M1 do x86 efficiently then it's obvious that AMD and Intel could do the same thing. It's still emulation and it is 2/3 the performance of native ARM apps, but it can obviously be done. Remember that modern x86 CPU's are just RISC with a CISC to RISC converter. Also every time someone says how bad and old x86, then suddenly we have new x86 CPU's that dominate in performance.

The problem is that Intel dominated the market for too long and they got lazy about increasing performance. There's a reason why 2500K's and 2600K's are still good CPU's to use because the amount of improvements we've seen since 2012 has been minor. Also why AMD's new Ryzen 5000 series are not good deals over h 3000 series. The improvements are just not that much.

What Apple did with the M1 was catch up to Intel in performance, but beat them in power efficiency hands down. With Nvidia I expect them to destroy Intel in performance and power consumption. But now AMD and Intel have a serious problem on their hands, because the market isn't a simple duopoly anymore. AMD and Intel now have to compete with Apple and Nvidia, and possibly anyone who can pay the ARM license. We may just get an explosion of cheaper and faster CPU's, which is great for consumers.
 
Revenant_Knight said:
That's a big statement. What do you know that we don't? I don't recall Apple's ARM licensing contract being public. We don't know what they have, how long they have it, or what their plans are. Developing their own IP would literally mean switching architecture unless they have some sort of specialized contract. I'm not saying they don't, but unless you have access to that info, we don't know.

I suspect the deal will go through, but I still don't see a benefit for Apple with Nvidia now holding the keys to the ARM castle.
Click to expand...

Apple was a founder company for ARM

As a part of selling away their shares (to softbank?), Apple made sure that they had a perpetual license to ARM
 
Is ARM not a European company? Why did we end up talking about China? China gets no say here.
 
FLECOM said:
member when MIPS was going to take the throne away from x86?

member when PowerPC was going to take the throne away from x86?

member when SPARC was going to take the throne away from x86?

member when Alpha was going to take the throne away from x86?

member when PA-RISC was going to take the throne away from x86?

member when RISC-V was going to take the throne away from x86?
Click to expand...
member when a 2 TRILLION dollar company switched all their products to run on any of those above....

i'll wait
 
You must log in or register to reply here.
Back
Top