Question Qualcomm's first Nuvia based SoC - Hamoa

[senttoschool]

Then ARM should compete by producing better designs.

Agreed. They should and they will. But they can also simultaneously sue Qualcomm.

 

[FlameTail]

Then ARM should compete by producing better designs.

This. Competition is always good for the consumer, even though it may not be good for ARM in their perspective🤪

 

[Gideon]

Licensing core designs is much more profitable than just licensing the ISA for ARM. Hence, if Nuvia designs are as good as Apple's, then ARM will lose a huge customer in Qualcomm, and will lose royalties from companies like Mediatek due to Nuvia dominating non-Apple sales.

Nuvia is a huge threat to ARM.

There is also some middle ground. They could always renegotiate the ISA licence in a way that it isn't an outright ban. Like asking for some revenue per chip sold.

 

[igor_kavinski]

Nuvia's design based on ARM ISA is the problem. They can create their own ISA and end this legal battle.

 

[igor_kavinski]

Like asking for some revenue per chip sold.

Which Qualcomm wouldn't agree to.

 

[KompuKare]

Nuvia's design based on ARM ISA is the problem. They can create their own ISA and end this legal battle.

Probably far too late for this, but anyone pouring lots of money into designing a new CPU from the clean sheet is going to really consider RISC-V. And that's even if ARM's ISA is better. For some the risk is just to great - spend ages and huge resources designing a world-leading CPU implementation and then have a chance for lawyers to get it banned? Better to go the extra step and avoid ARM.

 

[Gideon]

Probably far too late for this, but anyone pouring lots of money into designing a new CPU from the clean sheet is going to really consider RISC-V. And that's even if ARM's ISA is better. For some the risk is just to great - spend ages and huge resources designing a world-leading CPU implementation and then have a chance for lawyers to get it banned? Better to go the extra step and avoid ARM.

The way things are headed, this is something we will probably see at some point. The problem is RISC-V is actually quite fragmented as far as implementations go (much more than ARM) and getting ARM servers to the point they are took 10 years of work more-or-less. Just ask Jon Masters who is mostly responsible for it.

The OS ports are just the tip of a software iceberg, warns Jon Masters. He has spent the last nine years working on a standard version of Red Hat Linux for Arm servers. So far, just two commercial systems have been announced as certified to run it.

The process involved identifying a set of low-level hardware primitives that are assumed in the x86 world for things like how interrupts and power states work. Those de facto standards were articulated in a 50+ page document, then applied to the Arm architecture.

A separate effort documented x86 boot standards across Linux and Windows and applied them to Arm cores. They included obscure but crucial details of BIOS, power, and multiprocessing functions.

“I worry that the excitement of a new thing like RISC-V misses the real problems” of defining such support details, said Masters, who has been working on Linux for Arm servers since 2011.

RISC-V backers “will learn that no one got PCI right in Arm for the first few generations — to do PCI right [in a non-x86 host system] is tricky,” he said.

Server customers want support for both Windows and Linux, said Richard Jones, an emerging technologies specialist at Red Hat who worked on the RISC-V port of Fedora. In addition, they need a single image of an OS kernel and an applications binary interface that remains stable for years, Jones said in a talk at a RISC-V event last year.

A bootstrapped version of 64-bit Debian is available for RISC-V, but the architecture won’t be natively supported in the upcoming version 10 of the code. RISC-V still lacks support for an LLVM compiler, Java, and languages such as Golang and Rust, said a Debian developer at a recent event.

For its part, Arm led the formation of Linaro, a collaboration that has worked for several years on low-level software for servers as well as systems for telcos and the internet of things. Linaro helped pave Arm’s expansion into a wide variety of areas but has required dedicated engineering time from a wide variety of members at an estimated cost of $100 million.

RISC-V “needs standards and something like a Linaro, and they know this,” said Masters. China’s burgeoning interest in RISC-V could accelerate such efforts, perhaps collapsing them to a five-year journey. However, the opportunity for servers is arguably closing with two strong x86 companies and Arm server alternatives available from Ampere, Cavium, and Huawei, he said.

“Cloud server vendors are hedging their bets with Arm, but they don’t want 10 alternatives, so I’m not sure a RISC-V server opportunity really exists,” said Masters. “I’m sure startups will get funded, but I think they will build domain-specific accelerators.”

Luckily this article is nearly 4 years old.

BTW I'm not saying RISC-V takeover it won't happen, far from it. As it currently stands it probably will happen and I'm quite certain they'll manage the turnaround (in servers at least) faster than the 10 years it took ARM (as ARM has already made up some groundwork).

But RISC-V won't be a solution for most in the near-term (2-3 years) time. There will probably be some early birds, but it will take 5+ years to gain any serious traction.

 

[hemedans]

The core designs they produced before were no better (and sometimes worse) than ARM designed cores. Some will claim the difference is that Qualcomm cores (from Nuvia) are better now, but I still think there must be some language in these agreements about transfer of IP since that's what ARM keeps harping on in their filings and what Qualcomm's filings seem to dismiss.

If the language was absolutely crystal clear in either direction it would be tough to fight in court, so I'm guessing there's enough ambiguity that lawyers on both sides can attempt to argue their side is right. Too bad we can't see that language, if nothing else so we can armchair lawyer it and attempt to guess which side has a stronger position.

Not true though, Krait were miles better than Cortex A15, and in those area where A15 perform better it used more power.

Snapdragon 800 (MSM8974) Performance Preview: Qualcomm Mobile Development Tablet Tested

www.anandtech.com

Arm had good core from cortex A72 to A78, but before that period their cores were average. One of the reason Sd 810 were bad is due to Arm design cortex A57.

 

[nicalandia]

What is LLC ?

Last Level Cache

 

[DrMrLordX]

So wait

where does Oryon fit into all of this?

 

[Roland00Address]

But legally QC designs beating Arm designs now and not before changes nothing on the legal front.
If Arm was happy before, it will be harder to argue suddenly there's a problem.

The details of a contract, one we can not see matter. And the details may be stupid, and included to benefit one party and not the other party, but that is the point once it is in a contract one does not get to renegotiate like that.

And remember we are not dealing with one contract here, there was separate contracts with different language for Qualcomm and Nuvia, and Arm wants these two different legal entities to remain separate, and if Qualcomm can not do that they must destroy all the Nuvia IP that it acquired.

It really stinks that there will always be a fog of war that we can not see what is in the legal contract. Likewise we really do not care what is in the legal documents, we just want the best external results no matter what they are. Shrug it is all Kant and the 1780s all the way down, we Kant know what we want to know.

 

[Doug S]

Which Qualcomm wouldn't agree to.

Though they seem to have no problem at all asking for it when selling their chips!

 

[senttoschool]

There is also some middle ground. They could always renegotiate the ISA licence in a way that it isn't an outright ban. Like asking for some revenue per chip sold.

Who knows what the current terms are for ISA? For all we know, Qualcomm and ARM agreed to a lifetime license for their ISA already and it can't be changed.

Seeing how ARM is suing their biggest customer, it seems more likely than not that the Qualcomm ISA agreement was lifetime or at least decades.

 

[Thala]

Luckily this article is nearly 4 years old.

Depends on how you see it. Despite the article being 4 years old, most of the things stated there still hold true. That shows you how slow things are moving and that a decade is already underestimated.
In fact as was pointed out, the window of opportunity might be closing for RISC-V, as there is not much interest (and not much gain) in a 3rd server ecosystem.

 

[Thibsie]

Depends on how you see it. Despite the article being 4 years old, most of the things stated there still hold true. That shows you how slow things are moving and that a decade is already underestimated.
In fact as was pointed out, the window of opportunity might be closing for RISC-V, as there is not much interest (and not much gain) in a 3rd server ecosystem.

For such markets and for now, maybe.
But if ARM get their licensees angry, things might change.
For sure, smarthpone market would be easier to move to RiscV than DC market.

 

[FlameTail]

When are we going to move to Quantum processors ?

😆

 

[moinmoin]

In fact as was pointed out, the window of opportunity might be closing for RISC-V, as there is not much interest (and not much gain) in a 3rd server ecosystem.

Personally I think the space to watch is not commercial data centers but science and education. And the latter will show its effect only in the long run, slowly affecting other markets afterward.

 

[Nothingness]

Personally I think the space to watch is not commercial data centers but science and education. And the latter will show its effect only in the long run, slowly affecting other markets afterward.

I will just quote the answer I made to a claim you made months ago:

Sorry but you are overestimating the impact of the ISA. I studied and played with MIPS at Uni (back in a time it was used in workstations and servers, yes I'm old). If RISC-V had existed back then, I would surely have used it rather than MIPS, but that wouldn't haven't changed anything as I never used MIPS in my career in CPU design later on. When you have learned an ISA, you can switch to any other in very little time, or you are missing some of the qualities that make a good engineer.

 

[moinmoin]

@Nothingness Good link back to the previous extensive discussion on that topic in case others are interested in it. Agree to disagree on that particular point.

 

[eek2121]

RISC-V is catching up. It takes time. If I were ARM I would be worried. Qualcomm could always drop the ARM license and build a RISC-V chip.

Right now, you can get an SBC with a RISC-V chip that performs as well as the Cortex A-72, and who knows what we will see in the future?

 

[FlameTail]

I wonder which foundry they are going to use to make this Hamoa SoC?

 

[DrMrLordX]

I wonder which foundry they are going to use to make this Hamoa SoC?

Probably TSMC

 

[soresu]

RISC-V is catching up. It takes time. If I were ARM I would be worried. Qualcomm could always drop the ARM license and build a RISC-V chip.

They can, but they would run into exactly the same problems that Intel did breaking into the Android market - native ARM binaries for many popular apps.

I'm not saying it's an insurmountable problem, but it certainly won't be easy.

That being said I wouldn't be surprised to find that Jim Keller's Tenstorrent company weren't also working on a straight up RV CPU core along with their RV based ML accelerator.

 

[NostaSeronx]

That being said I wouldn't be surprised to find that Jim Keller's Tenstorrent company weren't also working on a straight up RV CPU core along with their RV based ML accelerator.

Ascalon has been known for sometime...

One Ascalon:

Two Ascalon:

Two Ascalon's Pipeline:

Two Ascalon LP:

 

[VirtualLarry]

Then ARM should compete by producing better designs.

BAN ARM.