Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Arm is canceling Qualcomm's chip design license

bloomberg.com

Post reply on HN
Print view View original

101–110 of 480 posts

Re: Arm is canceling Qualcomm's chip design license

#101

Post by snvzz »

hajile wrote: post #60
svnt wrote: post #57

Earlier quoted context omitted.

You’re suggesting that Snapdragon processors would switch to RISC-V and that would be no big deal? Presumably Qualcomm is committed to numerous multi-year supplier agreements for the arm chipsets.

Qualcomm pitched Znew quite a while ago. It mostly ditched 16-bit instructions and added a bunch of instructions that were basically ripped straight from ARM64. The idea was obviously an attempt at making it as easy as possible to replace ARM with RISC-V without having to rework much of the core. https://lists.riscv.org/g/tech-profiles/attachment/332/0/cod...

An attempt that failed miserably. (it was formally rejected a year ago)

But, by now, it is expected that Qualcomm's RISC-V designs have been re-aligned to match the reality that Qualcomm does not control the standard.

Re: Arm is canceling Qualcomm's chip design license

#102

Post by fargle »

> If Arm follows through with the license termination, Qualcomm would be prevented from doing its own designs using Arm’s instruction set

i'm not sure this is true. certainly "chip" IP has been a real legal quagmire since, forever.

but it was my understanding that you could neither patent nor copyright simply an "instruction set".

presumably what you get from ARM with an architecture license would be patent licenses and the trademark. if so, what patents might be relevant or would be a problem if you were to make an "ARMv8-ish compatible" ISA/Architecture with a boring name? i haven't seen much about ARM that's architecturally particularly unique or new, even if specific implementation details may be patent-able. you could always implement those differently to the same spec.

to further poke at the issue, if it's patents, then how does a RISC-V CPU or other ISA help you? simply because it's a different ISA, doesn't mean its implementation doesn't trample on some ARM patents either.

if it's something to do with the ISA itself, how does that affect emulators?

what's ARM's IP really consist of when you build your own non-ARM IP CPU from scratch? anyone have examples of show-stopper patents?

Re: Arm is canceling Qualcomm's chip design license

#103

Post by snvzz »

topspin wrote: post #77
1123581321 wrote: post #61

Earlier quoted context omitted.

It’s not a big hurdle so long as they hire Jim Keller to rapidly improve yet another architecture. (Only half-joking.)

Jim Keller is actually working on this right now at Tenstorrent.

Ascalon (claiming Zen5-tier performance) is done, and has been available for licensing for about a year now.

It was then made public that LG bought a license right away.

Re: Arm is canceling Qualcomm's chip design license

#104

Post by Joel_Mckay »

chx wrote: post #45
Joel_Mckay wrote: post #33

Next week, Qualcomm will likely announce a 64 core RISC-V RVA23. ARM really shouldn't pursue an aggressive posture with lines outside iOS or Win11 ecosystems. The leverage won't hold a position already fractured off a legacy market. =3

1. I am fairly sure games and other performance sensitive apps are using the Android NDK which is not available for RISC-V. 2. I am fairly sure a competitive RISC-V CPU is not days or weeks but years away.

I think it is more of a "chicken and egg" ordering problem.

1. The RISC-V design standard fragmentation issue has been addressed.

2. A reasonable mobile class level SoC will be available for integration after any large production run of the chips.

If ARM forces #2 out of silliness, than it also accelerates #1 in the market.

In general, there is plenty of use-cases even if a chip is not cutting edge. =3

Re: Arm is canceling Qualcomm's chip design license

#105

Post by philistine »

nemothekid wrote: post #62
hajile wrote: post #52

Earlier quoted context omitted.

Qualcomm doesn't have nearly as much to lose as ARM does and they know it. Qualcomm is almost certainly ARM's biggest customer. If ARM loses, Qualcomm doesn't have to pay out. If ARM wins, Qualcomm moves to RISC-V and ARM loses even harder in the long-term. The most likely outcome is that Qualcomm agrees to pay a slight bit more than they are currently paying, but nowhere near what ARM is demanding and in the meantim…

> Qualcomm moves to RISC-V and ARM loses even harder in the long-term. I think long term is doing a lot of heavy lifting here. How long until: 1. Qualcomm develops a chip that competitive in performance to ARM 2. The entire software world is ready to recompile everything for RISC-V Unless you are Apple I see such a transition taking a decade easily.

That's what's magical about Apple. It was a decade-long transition. All the 32-bit code that was removed from macOS back in 2017 was all in preparation for the move in 2019.

Re: Arm is canceling Qualcomm's chip design license

#106

Post by snvzz »

Atotalnoob wrote: post #76
gjsman-1000 wrote: post #8

Earlier quoted context omitted.

No kidding; and while RISC-V is a massive improvement, I hate to be the wet blanket, but RISC-V will not change signed boot, bootloader restrictions, messy or closed source drivers, carrier requirements, DRM implementation, or other painful day to day paper cuts. Open source architecture != Open source software and certainly != Open source hardware; no matter what the YouTubers think.

Until arm or risc-v fix standardize the bootloader, it’s always going to be a big deal for each arm/risc device added anywhere…

Relevant RISC-V specs were released years ago and implementations follow them.

I know of no boards that have application processors and yet not implement SBI. Furthermore, everybody seems to be using the opensbi implementation.

ARM and RISC-V are not the same.

Re: Arm is canceling Qualcomm's chip design license

#107

Post by duxup »

snvzz wrote: post #13

Bloomberg disappoints by failing to mention RISC-V at all in the entire article. They have to be doing this deliberately, as it's hard to explain otherwise.

If the comments in here are correct, RISC-V is really not an option at this time due to performance.

Re: Arm is canceling Qualcomm's chip design license

#109

Post by Joel_Mckay »

svnt wrote: post #65
chx wrote: post #45

Earlier quoted context omitted.

1. I am fairly sure games and other performance sensitive apps are using the Android NDK which is not available for RISC-V. 2. I am fairly sure a competitive RISC-V CPU is not days or weeks but years away.

> 2. I am fairly sure a competitive RISC-V CPU is not days or weeks but years away. And chasing a moving target fueled by the largest technology companies on the planet.

Tying products to Googles ecosystem is usually financially risky. Not a good long-term strategy for startups. =3

Re: Arm is canceling Qualcomm's chip design license

#110

Post by monocasa »

dagmx wrote: post #90
snvzz wrote: post #80

Earlier quoted context omitted.

This. It is evident to anybody paying attention that Microsoft has RISC-V support well underway. But even if they had to start from scratch, it would be much easier, thanks to ARM having paved the way.

How do you imagine ARM having helped?

The x86 emulator for one.
Post reply on HN