Qualcomm RISCs, Arm Pulls: The Legal Battle for the Future of Client Computing | Hacker News Reader

Qualcomm RISCs, Arm Pulls: The Legal Battle for the Future of Client Computing

thechipletter.substack.com

55 pointschmaynard21 comments

Only half of this article is free, the other is paid.

So the ARM v. Qualcomm litigation and the business moves preceding it seem to basically be hardball license contract negotiation via lawsuit. Qualcomm is by far ARM's largest customer and the ARM license is fundamental to Qualcomm's CPUs. Neither can really afford to live without the other but for both, significant business and strategic value is tied to getting more per core (ARM) or paying less per core (Qualcomm).

I suspect that eventually they'll end up settling with each other once the court case gets to a point that narrows the probable range of the outcome (thus creating the approximate upper and lower boundaries to negotiate within).

I think I would rather characterise it as hardball action by Qualcomm that got an expected response from Arm. The fact that Qualcomm told everyone except Arm in advance of the Nuvia deal is a tell.

I think settlement is unlikely - we are two years into the lawsuit. Arm really doesn’t want Oryon cores dominating - I think independent of price - and Qualcomm probably feels it has little to lose.

(I’m the author of this post)

I'm sure that the lawyers decided that ARM must enforce its licenses, especially when a license holder has explicitly violated the terms in a public fashion. If they don't at least automatically try and enforce their position through legal processes, ARM risks devaluing their IP and losing future negotiating power and give.

I'd be surprised if there was anything but cold hard contract negotiations going on behind the scenes that will shortly end in an amicable settlement and another 10 years of ARM based QCOM chips.

> Neither can really afford to live without the other

I imagine that ARM without QCOM will be totally fine. Rockchip is happy to take the low end, MTK will take the middle, AAPL already owns the high end.

Which "end" is taken by whom does little to explain how ARM will be alright after losing its largest client.

Particularly, as it would mean that client would likely switch to RISC-V entirely, thus further accelerating that ecosystem. Which no doubt would hurt ARM.

It would take years for Qualcomm to catch up. They’d be in a much worse than for example Intel is in now.

By the time they had a competitive RISC-V core (just look at how lo no it took them to ship Oryon and it would be much harder) Qualcomm would be either bankrupt of sold for parts.

The implicit assumption is that they are starting RISC-V efforts today.

Yeah, I’d say the same even if they had already started months ago, maybe even a year or two.

Qualcomm bought Nuvia back in 2021, so they didn’t start from scratch. They’d only released their first product this year and it’s not even that competitive (it’s good but certainly not an “x86 killer”.

And developing a new ARM core should be significantly easier to begin with.

They are also a public corporation it would be virtually impossible for Qualcomm to actually not make some sort of a deal with ARM, the market and their board would just force them to.

Why would ARM be making more per single Apple chip than they could get from Qualcomm even in the worst case?

ARM’s business model is not that great to begin with, losing Qualcomm would have a massive impact (obviously same applies to both companies ).

I doubt HN is interested in reading this. Especially half of the post are materials against Qualcomm.

> I doubt HN is interested in reading this. Especially half of the post are materials against Qualcomm.

Considering Qualcomm's business practices of the past, I would rather consider the material against Qualcomm to be beloved by a significant fraction of the HN readers. :-)

You would expect that. Except 99% of comments on the subject has been practically speaking "ARM is dead". ARM shouldn't sue Qualcomm.

Of course it seems that this hardball behaviour from ARM is likely a bad idea concerning its long-term business perspective, but I guess there is a significant group of people on HN which wouldn't be too sad about that because they hope this way RISC-V (which they love) becomes more widely used.

This is not wrong.

If it helps RISC-V, it really is for the greater good.

And if it helps RISC-V it helps China so.. even greater, greater good.

This is a great article explaining the reasoning of both sides of the battle.

And it is a very important turning point: riscv getting traction, desktop and mobile arms everywhere, x86 crisis in general.

I found it useful, so I'm glad I found it here.

Speak for yourself. If 'HN' doesn't want to read it, that's what downvoting is for.

What makes Qualcomm somehow sacrosanct and immune from analysis or critique?

>What makes Qualcomm somehow sacrosanct and immune from analysis or critique?

None of what was written is new. And it is not Qualcomm is immune to analysis. It is the other way around. Any evidence that benefits ARM and harm Qualcomm in this case has been completely ignored on HN.

> Any evidence that benefits ARM and harm Qualcomm in this case has been completely ignored on HN.

This feels like beating around the bush.

Can you give examples or citations of “evidence that benefits ARM and harm[s] Qualcomm,” so that we on HN may avoid completely ignoring it? Otherwise it feels like we’re doomed to tilt at windmills in your view, but we have no way of knowing how to disabuse ourselves of notions you only vaguely gesture at.

Reply on news.ycombinator.com