Arm Refutes Custom Chip Production Ambitions, Wants to Destroy Qualcomm's Nuvia IP

[AleksandarK]

A high-stakes trial between technology giants Arm and Qualcomm has revealed deeper tensions in the semiconductor industry, as Arm seeks the destruction of chip designs from Qualcomm's $1.4 billion Nuvia acquisition. The case, being heard in Delaware federal court, centers on a licensing dispute that could impact the future of AI-powered Windows PCs. Arm CEO Rene Haas took the stand Monday, adding allegations that Qualcomm violated licensing agreements following its 2021 acquisition of chip startup Nuvia. The issue is whether Qualcomm should pay Nuvia's higher royalty rates for using Arm's intellectual property rather than its own lower rates. Internal documents revealed Nuvia's rates were "many multiples" higher than Qualcomm's, with the acquisition potentially reducing Arm's revenue by $50 million.

During cross-examination, Qualcomm's legal team challenged Arm's motives, suggesting the dispute is part of a broader strategy to confront a customer increasingly viewed as a competitor. When presented with documents outlining potential plans for Arm to design its own chips, Haas downplayed these ambitions, emphasizing that Arm has never entered chip manufacturing. Allegedly, Arm sent letters to Qualcomm's customers, including Samsung, warning about possible disruption if Nuvia's IP design before acquisition in 2021 must be destroyed. Haas defended these communications, citing frequent inquiries from industry partners.




The trial's outcome could have significant implications for the PC market, where Microsoft and others are counting on Qualcomm's new AI-capable chips to compete with Apple's advances in laptop market. With Arm seeking no monetary damages despite Qualcomm's estimated $300 million annual licensing fees, the focus remains on the fate of Nuvia's designs. Jury deliberations are expected to begin Thursday, with Qualcomm CEO Cristiano Amon potentially taking the stand before the trial concludes.

View at TechPowerUp Main Site | Source

 

[windwhirl]

with the acquisition potentially reducing Arm's revenue by $50 million

50 million.

I feel like the amount of money involved here is so small considering the size and the importance of these companies, that I feel like slapping the back of the Qualcomm CEO's head for not just paying ARM that money, instead of making all this unnecessary mess and getting Qualcomm's licenses revoked.

For the record, Qualcomm's FY2024 ended with 10 billion operating income, so it's not like they're not able to pay.

 

[kondamin]

kinda lose my sympathy if you want to destroy a promising architecture.

 

[NoneRain]

50 million.

I feel like the amount of money involved here is so small considering the size and the importance of these companies, that I feel like slapping the back of the Qualcomm CEO's head for not just paying ARM that money, instead of making all this unnecessary mess and getting Qualcomm's licenses revoked.

For the record, Qualcomm's FY2024 ended with 10 billion operating income, so it's not like they're not able to pay.

$50 million is nothing. They're not aiming for the money.

 

[londiste]

Microsoft and others are counting on Qualcomm's new AI-capable chips to compete with Apple's advances in laptop market.

Why would Microsoft care? There are AMD and Intel both with their own AI hardware. Nvidia also exists plus there are other competitors to choose from if need be. I do wonder who the others are in laptop market betting on AI right now?

 

[Geofrancis]

All this is going to do is rapidly accelerate the transition to RISC V. Arm are already loosing large chunks of the microcontroller market, seagate moved all their hard drives to riscv, all the esp32 micro controllers are now riscv, along with a lot of other companies that are now moving over to RISCV and now with this lawsuit ARM are aiming at the high end customers. there are going to be few reasons left to use them.

 

[windwhirl]

All this is going to do is rapidly accelerate the transition to RISC V. Arm are already loosing large chunks of the microcontroller market, seagate moved all their hard drives to riscv, all the esp32 micro controllers are now riscv, along with a lot of other companies that are now moving over to RISCV and now with this lawsuit ARM are aiming at the high end customers. there are going to be few reasons left to use them.

I don't see that happening as fast as you seem to imply, to be honest. Lots of things need to change everywhere first, such as Android and Apple moving away from ARM and carrying with them all the software stuff that currently expects an ARM chip to work.

Plus I'm not sure about how production ready RISC-V is at the Linux kernel and such, for mobile phone usage.

That said, if RISC-V ends up being something better than ARM for the consumer down the line, I'm all for it.

 

[igormp]

All this is going to do is rapidly accelerate the transition to RISC V. Arm are already loosing large chunks of the microcontroller market, seagate moved all their hard drives to riscv, all the esp32 micro controllers are now riscv, along with a lot of other companies that are now moving over to RISCV and now with this lawsuit ARM are aiming at the high end customers. there are going to be few reasons left to use them.

The microcontroller and general purpose CPU spaces are totally different, so I don't think it's a good point of comparison.
Most ARM customers also just buy their core IPs instead of building their own. ATM I can only think of Apple and Qcom doing so currently in a large, consumer-facing scale (also fujitsu, but that's really niche).

Plus I'm not sure about how production ready RISC-V is at the Linux kernel and such, for mobile phone usage.

Kernel-wise it's ok, I have a single core SBC with proper upstream support. Issue is mostly getting the device trees for the all different SBCs (not much different from ARM), but lots of software are still not properly optimized, and the actual cores behind most RISC-V CPUs aren't really high-performing.

 

[ScaLibBDP]

I don't see that happening as fast as you seem to imply, to be honest. Lots of things need to change everywhere first, such as Android and Apple moving away from ARM and carrying with them all the software stuff that currently expects an ARM chip to work.

Plus I'm not sure about how production ready RISC-V is at the Linux kernel and such, for mobile phone usage.

That said, if RISC-V ends up being something better than ARM for the consumer down the line, I'm all for it.

>>...Plus I'm not sure about how production ready RISC-V is at the Linux kernel and such, for mobile phone usage.

RISC-V ISA is too fragmented because of countless number of useless extensions. I've been working with Linux for RISC-V ( Ubuntu / Debian / Fedora ) for almost 2 years and at the kernel level it looks very good. A lot of software packages have been ported ( at least what is needed for my work as a C/C++ developer ).

>>...That said, if RISC-V ends up being something better than ARM for the consumer down the line, I'm all for it.

I do Not think RISC-V is better and I would describe the current state as Stagnation after a Fireworks-like Start.

RISC-V International can Not do anything with regards to RISC-V ISA fragmentation because companies continue releasing new RISC-V ISA extensions and do Not care about Stability of the Whole RISC-V architecture.

Even if Google has informed everybody that RISC-V for Android would be a Tier-1 work there is an absolutely strange silence. No any news from Google.

What about NVIDIA and Alibaba? Both companies do a lot with RISC-V but they do Not try to enter PC and HPC markets with RISC-V solutions seriously.

At the RISC-V Summit North America 2024 I did Not see anything that I would describe as "Finally! There is a good progress to bring RISC-V to consumers".

RISC-V Summit North America 2024

Share your videos with friends, family, and the world

www.youtube.com

 

[unwind-protect]

RISC-V will so take over. Even earlier than I thought.

Licensing - pay money to be locked in.

 

[Fatalfury]

i think ARM is self-harmed its image & reputation thats it had build upon over the years.
a common man would think ARM to be like an "Open Source" Architecture which was the Father of Chip/Architecture design

So ARM was like a father to Qualcomm, Mediatek, Apple chips,Exynos etc.
But now things dont look Good...
ARM just cant kill Qualcomm in a day like they think...or can they?

 

[igormp]

RISC-V will so take over. Even earlier than I thought.

Licensing - pay money to be locked in.

FWIW most devices using RISC-V out there also license cores. The "royalty-free" licensing part is mostly relevant to companies doing their own core designs in-house.

 

[unwind-protect]

FWIW most devices using RISC-V out there also license cores. The "royalty-free" licensing part is mostly relevant to companies doing their own core designs in-house.

Sure, but you would have the option to switch around inside the RISC-V universe without putting your customers through an architecture change.

 

[trsttte]

If you can't compete, anti-compete

Qualcomm are no saints, but neither is ARM, they have a monetization problem but that's their problem, cases like this just makes them look bad. Qualcomm doesn't look any better, but when they show up with receipts like these, that's very damaging for ARM's case.

 

[thestryker6]

50 million.

I feel like the amount of money involved here is so small considering the size and the importance of these companies, that I feel like slapping the back of the Qualcomm CEO's head for not just paying ARM that money, instead of making all this unnecessary mess and getting Qualcomm's licenses revoked.

For the record, Qualcomm's FY2024 ended with 10 billion operating income, so it's not like they're not able to pay.

You're working on the assumption that $50 million is what they lost due to Qualcomm not using Nuvia's license agreements. The more likely source of that figure is that this is what Arm would have made without the acquisition had Nuvia moved forward. If Nuvia's royalty rates were "many multiples" of Qualcomm's and that only came to $50 million with Qualcomm's volume then Qualcomm has the best architecture license ever.