Live data from Hacker News

Arm is canceling Qualcomm's chip design license

bloomberg.com

291–300 of 480 posts

Re: Arm is canceling Qualcomm's chip design license

#291

Earlier quoted context omitted.

Without any ARM specific optimizations, most apps wouldn’t even have to recompile and resubmit. Android apps are uploaded as bytecode, which is then AOT compiled by Google’s cloud service for the different architectures, from what I understand. Google would just have to decide to support another target, and Google has already signaled their intent to support RISC-V with Android. https://opensource.googleblog.com/2023…

I remember when Intel was shipping x86 mobile CPUs for Android phones. I had one pretty soon after their release. The vast majority of Android apps I used at the time just worked without any issues. There were some apps that wouldn't appear in the store but the vast majority worked pretty much day one when those phones came out.

Yep! I had the Zenfone with an Intel processor in it, and it worked well!

Re: Arm is canceling Qualcomm's chip design license

#292
post #64

Earlier quoted context omitted.

(Naive question) then is it really a big hurdle for a company who knows how to make arm chips to try making riscv chips?

The question is more of will your customers agree to go along with this major architectural shift that sets you back on price-performance-power curves by at least five years and moves you out of the mainstream of board support packages, drivers, and everything else software-wise for phones. Also we should not pretend that ARM is just going to sit there waiting for RISC-V to catch up.

> The question is more of will your customers agree to go along with this major architectural shift that sets you back on price-performance-power curves by at least five years and moves you out of the mainstream of board support packages, drivers, and everything else software-wise for phones.

Embedded is moving to RISC-V where they have low performance needs.

One example is the Espressif line of CPUs - which have shipped over 1B units. They have moved most of their offerings to RISC-V over the last few years and they are very well supported by dev tools: https://www.espressif.com/en/products/socs

Re: Arm is canceling Qualcomm's chip design license

#294
post #138

Earlier quoted context omitted.

For Apple it was do or die the first few times. Until x86, if they didn’t move they’d just be left in the dust and their market would disappear. The ARM transition wasn’t strictly necessary like the last ones. It had huge benefits for them, so it makes sense, but they also knew what they were doing by then. In your examples (which are great) Intel wasn’t going to die. They had backups, and many of those seem guided m…

In a way that's also true for the x86->ARM transition, isn't it? I had an MacbookAir 2018. And.. "it was crap" is putting it very, very mildly. Yes it was still better than any Windows laptop I got since and much less of a hassle than any Linux laptop that I'm aware of in my circle. But the gap was really, really small and it cost twice as much. But the most important part for the working of the transition is probabl…

I don’t think so. I’ve got a 2019 MBP and yeah, the heat issue is a big problem.

But they weren’t going to be left in the performance dust like the last times. Their chip supplier wasn’t going to stop selling chips to them.

They would have likely had to give up on how thin their laptops were, but they could have continued on just fine.

I do think the ARM transition wasn’t strictly good, it let them stay thin and quiet and cooler. They got economies of scale with their phone chips.

But it wasn’t necessary to the degree the previous ones were.

Re: Arm is canceling Qualcomm's chip design license

#295
post #243

Earlier quoted context omitted.

This is a bit off topic, but has anyone demonstrated it's possible to design a big RISC-V core, that's performance competitive with the fastest x86 and ARM designs?

Well, Tenstorrent, Andes and others have their respective designs... On the in-order side, I can see on-par performance with the ARM A5x series quite easily.

After a bit of digging, I found that the SiFive P670 has performance equivalent to the Galaxy S21, or the desktop Ryzen 7 2700, which is not too bad and definitely usable in a smartphone/laptop form, so competitive with the 2021 era designs. Definitely usable. It's not clear what's the power level is.

Re: Arm is canceling Qualcomm's chip design license

#296

Earlier quoted context omitted.

> 1. Qualcomm develops a chip that competitive in performance to ARM Virtually all high performance processors these days operate on their own internal “instructions”. The instruction decoder at the very front of the pipeline that actually sees ARM or RISC-V or whatever is a relatively small piece of logic. If Qualcomm were motivated, I believe they could swap ISAs relatively easily on their flagship processors, and…

I am fan of the Jeff Geerling Youtube series in which he is trying to make GPU (AMD/Nvidia) run on Raspbery Pi. It is not easy - and they have linux kernel source code available to modify. Now imagine all Qualcomm clients have to do similar stuff with their third party hardware, possibly with no access to source code of drivers. Then debug and fix for 3y all the bugs that pop up in the wild. What a nightmare. Apple a…

Hardware drivers certainly can be annoying, but a hobbyist struggling to bring big GPUs’ hardware drivers to a random platform is not at all indicative of how hard it would be for a company with teams of engineers. If NVidia wanted their GPUs to work on Raspberry Pi, then it would already be done. It wouldn’t be an issue. But NVidia doesn’t care, because that’s not a real market for their GPUs.

Most OEMs don’t have much hardware secret sauce besides maybe cameras these days. The biggest OEMs probably have more hardware secret sauce, but they also should have correspondingly more software engineers who know how to write hardware drivers.

If Qualcomm moved their processors to RISC-V, then Qualcomm would certainly provide RISC-V drivers for their GPUs, their cellular modems, their image signal processors, etc. There would only be a little work required from Qualcomm’s clients (the phone OEMs) like making sure their fingerprint sensor has a RISC-V driver. And again, if Qualcomm were moving… it would be a sea change. Those fingerprint sensor manufacturers would absolutely ensure that they have a RISC-V driver available to the OEMs.

But, all of this is very hypothetical.

Re: Arm is canceling Qualcomm's chip design license

#297
post #286

Earlier quoted context omitted.

Does Qualcomm police it's IP at all? Traditionally they've been known as a tech company that employs more lawyers than engineers, if that tells you anything. I'd probably go up against IBM or Oracle before I tugged on Qualcomm's cape. Good luck to ARM, they'll need it.

Oracle v Qualcomm would be epic.

The public will likely be the loser of such a battle. :-(

Re: Arm is canceling Qualcomm's chip design license

#298

Qualcomm is known for having a particularly aggressive & hardball-style legal department to enforce its patents on core telecom IP. I believe the most likely outcome is they just settle the dispute here. Arm fighting hardball with hardball. Which would not really affect the ecosystem of phones using Qualcomm arm chips, it would just change the margins / market cap of Qualcomm. Yes, longterm Q might invest in their ow…

The Qualcomm online defenders are something else too. Qualcomm have been acting badly for years, including attempting to turn RISC-V into Arm64 but without the restrictions. You cannot trust people that behave like this, where everything they do is important and everything you do is worthless. The funny thing is Qualcomm do have some wildly impressive tech which is kept secret despite being so ubiquitous, but they ha…

> attempting to turn RISC-V into Arm64 but without the restrictions

This flew past me, do you have a link?

Re: Arm is canceling Qualcomm's chip design license

#299

And Qualcomm is the only competitive ARM chip on the market besides Apple's. And now they are being taken out by ARM. Is it really that expensive to re-license things? This seems self-defeating.

The issue is that Qualcomm wants to switch to ARM chips where they don't pay ARM much money.

When you're making ARM chips, you can either license the instruction set or you can license whole core designs. Most people license whole core designs. You build a chip using an ARM Cortex X4 core, 3 Cortex A720 cores, and 4 Cortex A520 cores and call it a day. But when you're using ARM's core designs, you have to pay them a lot to license those core designs. But when you're licensing the instruction set (and designing your own cores), you pay ARM a tiny licensing fee. This is what Apple does.

In this case, the story goes:

A startup called Nuvia wanted to create custom ARM cores for servers and negotiated a deal with ARM that was very favorable since ARM would like to grow its server marketshare. The agreement included a stipulation that they couldn't sell their IP to another company for that other company to build ARM cores built on Nuvia IP (according to ARM). Qualcomm argues that they have an instruction set license so they're allowed to build custom cores based off that license. ARM says that Nuvia's instruction set license means that Qualcomm can't.

I don't know what the cost difference is between core licenses and instruction licenses, but some places seem to think it's around 5x. Qualcomm has around 35% of the ARM chip market, but crucially a huge portion of the more expensive (and profitable for ARM) flagship cores. It's possible that Qualcomm is half of ARM's business (maybe more). If Qualcomm shifts to their own core designs and starts paying ARM 20% of what they're paying now, that could wipe out 40% of ARM's revenue.

If Qualcomm can shift from more expensive core licenses to cheaper instruction licenses, it would wipe out a huge portion of ARM's business. Worse, if those cores are better and become the de-facto flagship cores, it'd wipe out even more of ARM's business as companies like Samsung and Google might feel the need to buy Qualcomm chips (with better cores) rather than buying ARM's Cortex X cores for their flagship phones.

Likewise, Qualcomm is unlikely to stop at smartphones. They're already moving into laptops which will make it harder for any company using ARM-designed cores to get a foothold there. Qualcomm could move into servers in the near future and offer something better than the ARM Neoverse cores that are used by AWS's Graviton, Google's Axion, and Ampere.

So it's an enormous threat to ARM's business and ARM feels like it gave Nuvia a sweetheart deal because they were a small startup looking to enter a mostly new market rather than disrupting their current business - and they gave them a license with restrictions on it. Then Nuvia sold to Qualcomm who is using that IP to stop paying ARM - and ARM thinks that goes against the restrictions that Nuvia agreed to.

Re: Arm is canceling Qualcomm's chip design license

#300

Earlier quoted context omitted.

The Qualcomm online defenders are something else too. Qualcomm have been acting badly for years, including attempting to turn RISC-V into Arm64 but without the restrictions. You cannot trust people that behave like this, where everything they do is important and everything you do is worthless. The funny thing is Qualcomm do have some wildly impressive tech which is kept secret despite being so ubiquitous, but they ha…

> attempting to turn RISC-V into Arm64 but without the restrictions This flew past me, do you have a link?

Brucehoult mentions it further down.

> This time last year they were all over the RISC-V mailing lists, trying to convince everyone to drop the "C" extension from RVA23 because (basically confirmed by their employees) it was not easy to retrofit mildly variable length RISC-V instructions (2 bytes and 4 bytes) to the Aarch64 core they acquired from Nuvia. At the same time, Qualcomm proposed a new RISC-V extension that was pretty much ARMv8-lite.

This is enough of a philosophy change to break existing RISC-V software, and so is purely motivated by a desire to clone IP they supposedly licensed as honest brokers.

Post reply on HN