Apple should hire back Nuvia employees

[leman]

ARM stopped the Nuvia's license when they learned Qualcomm will take over. Qualcomm does not owe ARM money. ARM does it because they feared Qualcomm will stop using ARM's reference design after acquring Nuvia. It shows ARM sees Nuvia as a serious competition to their reference design.

That's an extremely weird way to be interpreting these things. And yes, according to ARM Qualcomm over them money for the license as they do not allow architectural licenses to be transferrable via startup purchases. That's exactly ARM's argument. They had a licensing agreement with a startup, not Qualcomm, so by acquiring the startup the license becomes void. It's just business as usual. I don't have a personal horse in this particular race, but ARM's arguments make perfect sense to me here. I mean, if you buy a coffee chain franchise and later sell your shop to a big supermarket chain that won't automatically allow them to use that coffee chain logo, would it?

If they were indeed worried about Nuvia becoming a competitor, why did they sell an architectural license to Nuvia in the first place? How can you reconcile these facts?

Yeah, we will see if Apple makes big changes in the performance cores in 2023. I am not optimistic about that. It has been 3 years since Firestorm came out and we are just seeing 18% improvement over those 3 years.

If you want to focus on just one metric among many, sure. Unfortunately such a narrow view does not add credibility to your argument.

 

[albertserene]

Wow, 1985! Which one? Where’d you get it?

Mac 512K

 

[albertserene]

That's an extremely weird way to be interpreting these things. And yes, according to ARM Qualcomm over them money for the license as they do not allow architectural licenses to be transferrable via startup purchases. That's exactly ARM's argument. They had a licensing agreement with a startup, not Qualcomm, so by acquiring the startup the license becomes void. It's just business as usual. I don't have a personal horse in this particular race, but ARM's arguments make perfect sense to me here. I mean, if you buy a coffee chain franchise and later sell your shop to a big supermarket chain that won't automatically allow them to use that coffee chain logo, would it?

If they were indeed worried about Nuvia becoming a competitor, why did they sell an architectural license to Nuvia in the first place? How can you reconcile these facts?



If you want to focus on just one metric among many, sure. Unfortunately such a narrow view does not add credibility to your argument.

ARM customers can license the ARM architecture or specific processor implementations. Companies like Apple, Nvidia, Nuvia only license the ARM architecture and develop their own processors. Companies like Qualcomm license the processor IP as well. ARM gets more loyalty fee from Qualcomm as a result. If Qualcomm uses the Nuvia processor design, they do not pay to ARM for the processor IP. This will result in loss of revenue for ARM. That's why ARM file the lawsuit against Qualcomm. It's all about money here

 

[leman]

ARM customers can license the ARM architecture or specific processor implementations. Companies like Apple, Nvidia, Nuvia only license the ARM architecture and develop their own processors. Companies like Qualcomm license the processor IP as well. ARM gets more loyalty fee from Qualcomm as a result. If Qualcomm uses the Nuvia processor design, they do not pay to ARM for the processor IP. This will result in loss of revenue for ARM. That's why ARM file the lawsuit against Qualcomm. It's all about money here

Of course it's about money. ARM doesn't want Qualcomm to get a cheap architectural license via a startup purchase. I am certain that they would be more than happy to sell one for Qualcomm for an appropriate price.

 

[Anaxarxes]

Ever since Apple lost its semiconductor employees to form Nuvia, the development process of Apple silicon has come to the grinding stop. The M2 is only 18% quicker than M1 after almost 2 years gap. The A16 is virtually the same as A15. It seems Apple is unable to further improve upon its ARM architecture. With the lawsuit between Qualcomm and ARM, Apple should take this opportunity to hire back those employees. Or else, it might just have to become a licensee of ARM's reference design.

Oh boy. Love those tinfoil hat business consultants that develop action plans for trillion dollar companies from their basement.

 

[fakestrawberryflavor]

Everything with chip design is on a 2-3 year development cycle. What happened 2-3 years ago? There was a major disruption in literally every market, every career field.

I’m not worried about their capabilities or future products. They have enough money to weather the storm and buy the expertise they need to deliver something successful.

My last comment is, general purpose CPUs are pretty fast already. If there is a focus on AI & ML engines, Real time ray tracing, etc, that would be nice too.

 

[heinzel]

ARM customers can license the ARM architecture or specific processor implementations. Companies like Apple, Nvidia, Nuvia only license the ARM architecture and develop their own processors. Companies like Qualcomm license the processor IP as well. ARM gets more loyalty fee from Qualcomm as a result. If Qualcomm uses the Nuvia processor design, they do not pay to ARM for the processor IP. This will result in loss of revenue for ARM. That's why ARM file the lawsuit against Qualcomm. It's all about money here

I am not sure whether anyone outside ARM and its IP customers knows what the specific price tag for the license negotiated between ARM and each chip manufacturer is. It sounds like Nuvia, being a new kid on the block, received very favorable licensing terms for its specific intended business model, and ARM is not willing to let those terms or the business model be transferred to Qualcomm without requiring a new license agreement. So, the dispute seems about licensing terms first, and about money second. Perhaps ARM loses some money if a chip maker switches to an architectural license (which I assume to depend on volume, do tell if you know anything about any actual terms), but that probably also depends on which core licenses get replaced (do you happen to know how this works in detail?). Regardless, they’ll continue to receive royalties for every chip sold, which makes up more than half of their revenue. If ARM loses grip on their licensing and on enforcing licensing terms, however, their entire business model falls apart. It sounds like Qualcomm tried to say that they aren’t required to discuss any licensing terms with ARM since both they and Nuvia had license agreements with ARM before they bought Nuvia, but those terms were negotiated based on each individual companies business plan and volumes. Since those have now changed for both I think that’s ARM why is litigating now. Of course in the end it’s about money, but not as simple as just that.

Here’s some (old, but still relevant) info on ARM’s licensing business: https://www.anandtech.com/show/7112/the-arm-diaries-part-1-how-arms-business-model-works/3

 

[souko]

I did take my time reading your thread, and all I get is you are pessimistic about the Apple silicon progress, because you expect apple Silicon to have 100% performance skyrocket every year or something crazy. What’s bad about the rest of the world catching up? Apple must be always successful 100%, flawless? Not to mention that’s already false (butterfly keyboard for example).

Im not defending Apple here. I just merely point out exponential growth you are looking for can’t happen all the time. Did you not see the huge performance jump between last Intel MacBook vs M1 MacBook Air? Did you not notice The amazing power efficiency on Apple silicon chips? Achieving better CPU performance at the power budget of a freaking whole computer is really something to brag about all the time?

Take your 6800U as an example, https://www.cpubenchmark.net/cpu.php?cpu=AMD+Ryzen+7+6800U&id=4923 shows the 6800U still lags behind on performance compared to M1 Pro and M1 Max in some settings, tho TDP is Only 15W, which is indeed impressive. https://www.ultrabookreview.com/36030-amd-ryzen-7-u-laptops/ But, that‘s just one processor designed for ultra books. Does it run games as well as other higher end chips? Likely not. Does it handle more intensive tasks better? I bet not. AMD and Intel still has a ton to do with their power efficiency, not to mention the desktop Apple Silicon hasn’t been released yet. (M1 ultra doesn’t count)

Just like Intel and AMD, apple won’t stand still in this Silicon performance race. I don't know why you are disappointed, just because they might move a bit slower for the time being. And about benchmark, while it can provide a clear comparison between different processors and architectures, do you know the benchmark program might be written to favor one vendor’s processor over the other? Same for GPU?

The AMD 6800U actually has a fairly powerful RDNA2 iGPU. It can achieve 3379 Gigaflops which is only slightly below M2's 3550 Gigaflops. Running at only 15W, It shows that x86 camp has improved its power efficiency.

Anyway, this discussion has gone too far. I hope I am wrong and that Apple can surprise us next year with new chips. Oh yeah, I have been an Apple fanboy since 1985 when I got my first Macintosh.

So, 6800U power consumption is about 50W, not 15W. https://www.notebookcheck.net/AMD-Ryzen-7-6800U-Processor-Benchmarks-and-Specs.589897.0.html

 

[albertserene]

So, 6800U power consumption is about 50W, not 15W. https://www.notebookcheck.net/AMD-Ryzen-7-6800U-Processor-Benchmarks-and-Specs.589897.0.html

TDP is 15W to 28W. 50W is the system power consumption with external monitor.

 

[Xiao_Xi]

SemiAnalysis has published some articles about the lawsuit between Qualcomm and ARM.

Is Arm Desperate? Qualcomm Claps Back At Arm’s Potentially Frivolous Lawsuit

Licensing Details Revealed

www.semianalysis.com

Arm Changes Business Model – OEM Partners Must Directly License From Arm

No More External GPU, NPU, or ISP’s Allowed In Arm-Based SOCs

www.semianalysis.com

Arm’s Nuclear Option – Qualcomm Must Cancel Next-Generation Products If Arm Succeeds

Arm denies allegations of changing its business modelIf you have been keeping up with the biggest soap opera in the tech world, you know Arm sued Qualcomm. Qualcomm more recently made some pretty b…

www.semianalysis.com

So far, Nuvia has proven nothing. It has published a white paper and named its future SoC. That is.

 

[BellSystem]

Stagnation?!?!?!? WHAT???!!?!? The M1 Ultra came out in March 2022 and the M2 in June of 2022. I think the problem is somebody's unrealistic expectations of Apples release and development cycles. What do you think their targets are??? New processor every 4th Tuesday of the month?? You have to sell the current products long enough to at least make back the development expenses and make some profit. I'm sure Apple has plenty waiting in the wings....but nothings moving until the sell through the current product targets. Like someone said, they are probably waiting to release the new Pro. They certainly are not going to release an M2 Ultra Platinum 6G SuperSport Mac Mini until the Pro comes out with it. I'm sure the Pro is also timed with Studio sales targets. Who knows.

TLDR: calm down!

 

[leman]

TDP is 15W to 28W. 50W is the system power consumption with external monitor.

TDP is not the same as power consumption. In CPU tests the 15W 6800U uses roughly 20 watts. I would assume that the same limit would also apply to the GPU. Of course, the TFLOPS you quote are unlikely from the 15W mode.

 

[souko]

TDP is 15W to 28W. 50W is the system power consumption with external monitor.

TDP is not maximum power CPU can draw. TDP for Intel and AMD is nothing saying number. 50W is only CPU in multicore Cinebench. M1 Pro 8-core draws 23W here, for example.

EDIT: Here is another test https://www.notebookcheck.net/AMD-R...iew-Zen3-beats-Intel-Alder-Lake.623763.0.html

 

[maflynn]

Stagnation?!?!?!? WHAT???!!?!? The M1 Ultra came out in March 2022

I think you should have picked a different variant of the M1. The ultra is just to M1 Pros Maxs lashed together. Its impressive but its basically two processors put together.

 

[Ethosik]

I think you should have picked a different variant of the M1. The ultra is just to M1 Pros lashed together. Its impressive but its basically two processors put together.

It’s two M1 Max. And having two significantly increases the memory bandwidth which in the apps I use I can tell a difference.

 

[maflynn]

It’s two M1 Max. And having two significantly increases the memory bandwidth which in the apps I use I can tell a difference.

Thanks for the correction, I mis-spoke.

 

[Pankeborg]

Ever since Apple lost its semiconductor employees to form Nuvia, the development process of Apple silicon has come to the grinding stop. The M2 is only 18% quicker than M1 after almost 2 years gap. The A16 is virtually the same as A15. It seems Apple is unable to further improve upon its ARM architecture. With the lawsuit between Qualcomm and ARM, Apple should take this opportunity to hire back those employees. Or else, it might just have to become a licensee of ARM's reference design.

In case this is meant seriously, and not as sarcasm. 18% is a lot, especially when your chips are already way faster than what 90-95% of your customers could ever want. Apple is not in the business of making ultrafast computers for gamers with a performance fetish, so they should not be judged like it.

 

[Tyler O'Bannon]

Ever since Apple lost its semiconductor employees to form Nuvia, the development process of Apple silicon has come to the grinding stop. The M2 is only 18% quicker than M1 after almost 2 years gap. The A16 is virtually the same as A15. It seems Apple is unable to further improve upon its ARM architecture. With the lawsuit between Qualcomm and ARM, Apple should take this opportunity to hire back those employees. Or else, it might just have to become a licensee of ARM's reference design.

When they don’t shrink the die size or significantly increase core count, this is likely going to be standard. I’m modest single core increase. It’s probably never going to be a bigger jump like 50% year over year.

 

[Kimmo]

I never said Apple Silicon is a terrible chip. It was the best chip when it was released. But the competition doesn't stand still and it's time to take the next big jump. We will see how Apple's semiconductor team can come up with without Gerald William III at the helm.

Gerald, is that you?

 

[Bug-Creator]

Every tech evolves slower and slower with every generation and there is always a hard ceiling on how good something can get no matter what.

The development of CPU/GPU "cores" show this just fine and most big jumps come from external factors. Faster I/O more&smaller transistors allowing for more cores or running at higher clocks.

1 field that is really showing progress is special accelerators for en/de-coding, neural networks etc which won't contribute to simplistic benchmark score.

Something I (as an uneducated layman) as potential field of growth is implementing these in some kind of hybrid mode between a fixed layout and FPGA allowing for smaller function blocks to be stringed together forming complex function.

Point is, Moore's law is dead, hence everybody need to think outside the box to proof Moore's law ain't dead...

 

[johnsc3]

I think a lot of talented engineers are tied up making the chipsets for the new Mac Pro. If you look at the rumours that is going to be a beast, and it requires different engineering for the outward facing / upgradeable parts. The effort of making the transition to Apple Silicon across the entire line-up shouldn’t be underestimated.

Thats why you’re temporarily seeing less focus on the A-series cores, and less of a step from M1 to M2. Once the new Mac Pro has finished testing, those engineers will be released back onto other projects and we will see whether Apple has managed to retain the important talent.

Well said.

 

[falkans]

You got all baited imo.

I'll give you a fancy one, just hear me:

Working on a new kind of device which is a keyboard and the cover is a screen.

Enjoy your we.

 

[daabido]

Mac 512K

Luxury.

 

[SlCKB0Y]

So what are you saying? Newbies are not entitled to their opinions? Another conspiracy theory? The Mac's competition is with x86 architecture. Raptor Lake already overtake M1 Ultra's performance. On the phone SoC, Snapdragon 8 gen 2 already overtake A16 in Antutu. As I have said, I am a big fan of Apple. It's time to move forward. The world doesn't standstill.

Initial indicators are that Apple with their A16 does not need to worry about SD 8 gen 2: https://www.macrumors.com/2022/11/17/iphone-14-pro-a16-outpreforms-snapdragon/

 

[SlCKB0Y]

I don't care for Apple anymore

And yet, here you are??