Report: Apple to Move a Part of its Embedded Cores to RISC-V, Stepping Away from Arm ISA

[AleksandarK]

According to Dylan Patel of SemiAnalysis sources, Apple is moving its embedded cores from Arm to RISC-V. In Apple's Silicon designs, there are far more cores than the main ones that power the operating system and end-user applications. For example, embedded cores are present, and there are 30+ in M1 SoCs responsible for all kinds of workloads not related to the operating system. These tasks are usually associated with other functions such as WiFi/BlueTooth, ThunderBolt retiming, touchpad control, NAND chips having their own core, etc. They run their own firmware and power everything around the central cores that run the OS, so the whole SoC functions appropriately.

It appears that a lot of these cores are based on Arm M-series or lower-end A-series IP that Apple is currently looking to replace with RISC-V. Given that a large portion of software runs on the main big.LITTLE configuration, other secondary SoC tasks can migrate to a different ISA like RISC-V, with a small firmware adjustment. Given that these cores can be placed with custom IPs, Apple would save licensing fees if custom RISC-V cores were used. Additionally, developing firmware for these cores at an Apple engineering team size shouldn't be a problem. Of course, we have no information about when these custom cores will appear inside Apple Silicon. Even when they are used, no formal announcement is expected given that the main cores remain to be powered by Arm ISA, with everything else invisible to the end-user.


View at TechPowerUp Main Site | Source

 

[stimpy88]

The ARM isa has pretty much reached the peak of what it's capable of. We have seen that Apple, who has the most advanced and performant ARM design, has not been able to make any significant IPC gains for a few years now. The A16 has single digit IPC improvements over the A15, and A15 had only small gains over A14, with most gains being achieved by simple clockspeed bumps.

I will concede that lack of competition in the mobile space may also be a factor in this, but I will remind you that Apple is also using these designed in their laptops, so they are competing with AMD and Intel on that front.

So, I think it's time for the next generation of low power CPUs, and I don't think it will be ARM based.

 

[R0H1T]

This probably has more to do with having (extra) control & not paying anymore licensing fees to ARM than they absolutely have to! Apple being cheapskate as usual

 

[Nanochip]

I wonder if ARM’s lawsuit against Qualcomm spooked Apple? If so, ARM may well begin to decline as RISC-V begins to ascend.

 

[Wirko]

Quite a journey, even if we ignore everything before the Motorola 68000.

 

[Bomby569]

These license fees are peanuts for a company like Apple, still money is money i guess.

 

[SOAREVERSOR]

I wonder if ARM’s lawsuit against Qualcomm spooked Apple? If so, ARM may well begin to decline as RISC-V begins to ascend.

The way I read it they will still be using ARM for the main bits, just moving many of the other functions over to RISC-V.

 

[bonehead123]

If ARM has really reached it's peak capability, and as long as this doesn't turn out to be anutha AIM clusterf*ck like it was before, I am good with whatever they come up with to keep the tech moving forward....

If you don't know what AIM was or what happened, grab a barf bag & hold on to your seat when you google it... man I was using ALOT of pepto & Aspirin back then....hehehe

 

[john_]

Nvidia failed to buy ARM. But what if in the future ARM is bought by someone else or the owner of ARM starts changing the rules? I think that is the reason why Apple could be using ARM as an intermediate step before moving everything to Risc-V. Nvidia's efforts to buy ARM probably spooked Apple execs. So maybe we see the first step of Apple moving from ARM to Risc-V. And who knows after 10-15 years we could see Apple devices using chips that are based on some kind of Apple proprietary architecture, unique to Apple.

 

[AnotherReader]

If ARM has really reached it's peak capability, and as long as this doesn't turn out to be anutha AIM clusterf*ck like it was before, I am good with whatever they come up with to keep the tech moving forward....

If you don't know what AIM was or what happened, grab a barf bag & hold on to your seat when you google it... man I was using ALOT of pepto & Aspirin back then....hehehe

Apple has far more resources than any two of the participants in AIM had in the 1990s.

 

[r9]

These license fees are peanuts for a company like Apple, still money is money i guess.

That's a lot of peanuts though.
Plus any savings turn into big fat checks for the execs So yeah it's worth the hustle.

The ARM isa has pretty much reached the peak of what it's capable of. We have seen that Apple, who has the most advanced and performant ARM design, has not been able to make any significant IPC gains for a few years now. The A16 has single digit IPC improvements over the A15, and A15 had only small gains over A14, with most gains being achieved by simple clockspeed bumps.

I will concede that lack of competition in the mobile space may also be a factor in this, but I will remind you that Apple is also using these designed in their laptops, so they are competing with AMD and Intel on that front.

So, I think it's time for the next generation of low power CPUs, and I don't think it will be ARM based.

For IPC improvements you can optimize something so much without adding more instructions=transistors and the more instructions they add the closer they get to x86 and it will defeat the purpose.
RISC/ARM is the future at least should be.
Apple proved that there is nothing that ARM can't do even with apps that are not native. So if the world switched to ARM right this second and everybody just made ARM version of their software in about a year x86 would be totally useless especially in the mobile world.
Take Steam Deck that tiny APU probable uses same power as M1 Pro/Max and is like 20 times slower.
With ARM you can have Ryzen 5800 / RTX 3070 in something small like the Steam Deck today where with the x86 in about 10 years.

 

[AnotherReader]

That's a lot of peanuts though.
Plus any savings turn into big fat checks for the execs So yeah it's worth the hustle.


For IPC improvements you can optimize something so much without adding more instructions=transistors and the more instructions they add the closer they get to x86 and it will defeat the purpose.
RISC/ARM is the future at least should be.
Apple proved that there is nothing that ARM can't do even with apps that are not native. So if the world switched to ARM right this second and everybody just made ARM version of their software in about a year x86 would be totally useless especially in the mobile world.
Take Steam Deck that tiny APU probable uses same power as M1 Pro/Max and is like 20 times slower.
With ARM you can have Ryzen 5800 / RTX 3070 in something small like the Steam Deck today where with the x86 in about 10 years.

That isn't about ARM/RISC vs x86. ISA doesn't matter as much as RISC cultists think it does. Besides Apple, the other ARM chips have unimpressive performance per watt vs AMD at the same performance node. Apple looks better, because they are a node ahead, and they optimize for power. If Intel or AMD designed their cores to have sub 4 GHz clocks, they could use smaller and slower transistors and burn less power, but they opted to push for peak single threaded performance for poorly optimized code by choosing high clock speeds.

One of your predictions is already correct though: nobody cares about x86 in the mobile world unless you count laptops.

 

[Denver]

That's a lot of peanuts though.
Plus any savings turn into big fat checks for the execs So yeah it's worth the hustle.


For IPC improvements you can optimize something so much without adding more instructions=transistors and the more instructions they add the closer they get to x86 and it will defeat the purpose.
RISC/ARM is the future at least should be.
Apple proved that there is nothing that ARM can't do even with apps that are not native. So if the world switched to ARM right this second and everybody just made ARM version of their software in about a year x86 would be totally useless especially in the mobile world.
Take Steam Deck that tiny APU probable uses same power as M1 Pro/Max and is like 20 times slower.
With ARM you can have Ryzen 5800 / RTX 3070 in something small like the Steam Deck today where with the x86 in about 10 years.

It wouldn't be a problem to build a big x86 APU, drop the clockrate and get better efficiency, but that chip would be expensive and in limited supply.. that's not the purpose of SteamDeck or AMD. They have a huge market to feed, not just elitists who pay thousands of dollars to be limited to the apple OS.

 

[usiname]

That's a lot of peanuts though.
Plus any savings turn into big fat checks for the execs So yeah it's worth the hustle.


For IPC improvements you can optimize something so much without adding more instructions=transistors and the more instructions they add the closer they get to x86 and it will defeat the purpose.
RISC/ARM is the future at least should be.
Apple proved that there is nothing that ARM can't do even with apps that are not native. So if the world switched to ARM right this second and everybody just made ARM version of their software in about a year x86 would be totally useless especially in the mobile world.
Take Steam Deck that tiny APU probable uses same power as M1 Pro/Max and is like 20 times slower.
With ARM you can have Ryzen 5800 / RTX 3070 in something small like the Steam Deck today where with the x86 in about 10 years.

M1 pro and the rest m1 toys are not faster in anything. 6900hs has similar consumption as m1 pro and is way faster with node disadvantage. Not to mention all limitations of ARM vs x86

 

[Wirko]

And who knows after 10-15 years we could see Apple devices using chips that are based on some kind of Apple proprietary architecture, unique to Apple.

Apple Core!

 

[enzolt]

Where is the source?

 

[First Strike]

Apple's involvement will not be necessarily good for RISC-V. Apple keeps a very great record of adding custom instructions / instruction behaviors without even publicly documenting them, even when they are using ARM which is supposed to be proprietary (See how much trouble Linux for M1 group went through). And now we have the fragmentation-prone RISC-V. Currently, the only thing to keep RISC-V from fragmenting is the standard committee who has been ratify extensions after extensions, profiles after profiles. But Apple is sure to ignore that.

 

[watzupken]

The ARM isa has pretty much reached the peak of what it's capable of. We have seen that Apple, who has the most advanced and performant ARM design, has not been able to make any significant IPC gains for a few years now. The A16 has single digit IPC improvements over the A15, and A15 had only small gains over A14, with most gains being achieved by simple clockspeed bumps.

I will concede that lack of competition in the mobile space may also be a factor in this, but I will remind you that Apple is also using these designed in their laptops, so they are competing with AMD and Intel on that front.

So, I think it's time for the next generation of low power CPUs, and I don't think it will be ARM based.

If you look at Apple’s product over the last half a decade, its never about pushing the performance, design nor features. They basically left those behind. The SOC is still one of the best, but it is a far cry from what it used to be a decade back. Their strategy is more about pushing some niche features, and trying to lock you in their ecosystem to make you stick around despite bringing very little improvements year on year. I feel Apple is on the path of decline because they play it too safe and don’t really bring meaningful improvements to the table. That’s why people have been comparing Steve vs Tim’s Apple. It is almost like 2 companies. If Apple started with Tim, I am doubtful it will be successful.
In my opinion, it is not that ARM have reached some sort of “peak” but rather the fab improvements have slowed down drastically. Name of the node is not the same as the actual product, and while something like a 7nm to 4nm sounds like a big improvement, it is not. With fab improvement falling behind, the improvement in SOC is also falling behind. You can tell that modern SOC be it from the likes of Qualcomm, Mediatek or Apple, they throttle fairly significantly under sustained load. So you can make a ridiculously power SOC, but the fab just can’t accommodate it.

 

[ypsylon]

I would speculate that Arm-China may be at play here. China flat-out refused to allow IPO by Softbank and they control the majority stake (according to reports 20% Arm, 28% Softbank and almost all of the rest is Arm-China). Last thing Apple needs is China being whacked by some US IT-related sanctions and they'll remain entangled by a supplier which will be unable to supply anything.

That's what you get for short sighted lack of diversification for the sake of profit$$$.

 

[Wirko]

Apple's involvement will not be necessarily good for RISC-V. Apple keeps a very great record of adding custom instructions / instruction behaviors without even publicly documenting them, even when they are using ARM which is supposed to be proprietary (See how much trouble Linux for M1 group went through). And now we have the fragmentation-prone RISC-V. Currently, the only thing to keep RISC-V from fragmenting is the standard committee who has been ratify extensions after extensions, profiles after profiles. But Apple is sure to ignore that.

The good side of Apple's isolated and walled garden is that their custom ISA extensions can't easily infect the rest of the ARM (or RISC-V) ecosystem.

I also don't see any danger coming from custom extensions - as long as they don't break compatibility with basic ISA and standard extensions defined by Arm. Here I suppose that Arm demands full compatibility for any product to be branded or advertised as Arm.

 

[defaultluser]

sound just like everyone else approach to risc-v:replace some random m4 series with a custom microcontroller

this completely ignores the hpc cluster that is optional vector unit, and clunky variable-length vectors

 

[Darmok N Jalad]

This probably has more to do with having (extra) control & not paying anymore licensing fees to ARM than they absolutely have to! Apple being cheapskate as usual

No, it’s controlling your own destiny. ARM was almost bought by NVIDIA, which Apple appears to be on bad terms with for a long time. Less dependence on licensing when you can do it yourself, the better off you can be, provided you can design something competitive. Apple has had its share of performance delivery problems from past vendors, be it IBM/Moto or Intel, heck, even AMD. ARM is more a licensing issue, and one can get behind a barrel there too.

 

[Nanochip]

M1 pro and the rest m1 toys are not faster in anything. 6900hs has similar consumption as m1 pro and is way faster with node disadvantage. Not to mention all limitations of ARM vs x86
View attachment 261870

With those outliers skewing the results, the average is the wrong tool to use. Use the median instead.

 

[Aquinus]

No, it’s controlling your own destiny. ARM was almost bought by NVIDIA, which Apple appears to be on bad terms with for a long time. Less dependence on licensing when you can do it yourself, the better off you can be, provided you can design something competitive. Apple has had its share of performance delivery problems from past vendors, be it IBM/Moto or Intel, heck, even AMD. ARM is more a licensing issue, and one can get behind a barrel there too.

^ This. If Apple is considering RISC-V it's only because nVidia spooked them when they tried to acquired ARM. Something like that could have massive ramifications for Apple, so diversifying gives them some flexibility. Using RISC-V for all of the controllers associated with their SoCs is not a bad idea. Even if someone were to acquire ARM, that at least lessens the possible impact when it comes to licensing and any changes to costs or terms. From a safety perspective, this would make a whole lot of sense to be completely honest.