Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Concerns About an Apple Silicon Mac Pro
- Thread starter RedTheReader
- Start date
- Sort by reaction score
Exactly what I was getting at
Exactly what I was getting at
To an extent but my point was they don’t need to own the ARM IP - they might *like* to own it, but they don’t *need* to own it to build combined CPU and GPU tech. They’re doing it already and this looks pretty promising.
Nvidia is not wrong that their deep pockets and aggressive pushing could make a potentially stronger ARM ecosystem if everyone trusted them to behave themselves. I mean if you just look at the upsides, sure there are a number of them. But the downsides are why everyone is looking at this deal in askance.
Having more control over the CPU IP would allow Nvidia to integrate systems further. Right now they are pretty much limited to whatever they can license. I can't find the source right now, but some time ago Nvidia engineers were complaining that the industry is running into limitations of current designs — for example, memory interfaces are simply not fast enough to do Ml on large scale, and making RAM faster is not efficient. One would need novel solutions here such as on-RAM processing, tighter integration, CPU/GPU hybrids, all that kind of stuff.
You are right of course, they don't "need" it right now, but it can become a very valuable asset down the road when paradigms change.
Oh, given Nvidia's business practices I am sure they will do their best to mess things up for everyone if the acquisition goes through. It would be a terrible thing for general purpose computing. No wonder there is so much resistance against it.
Nvidia was completely shut out on the two largest government contracts.
Frontier went to AMD.
Aurora went to Intel.
I already wrote that they already license ARM and have ARM CPUs. But ARM right now is unfocused in my opinion. I think Nvidia will focus ARM better in the consumer and server market. I think Nvidia sees more potential for ARM than just a company that designs general CPUs for others to license. I think ARM can better compete with Apple, Intel, AMD in the hands of Nvidia.
I support the Nvidia acquisition of ARM. I think it'll be better for competition because it enables another company to compete in the SoC space.
I don't see it that way.
I don't believe Nvidia will mess with the ARM licensing model. I think anyone with a current ARM license will be safe. And I actually see better ARM designs in the future if ARM has access to Nvidia IPs and resources.
I see it as adding another integrated SoC competitor to Apple, Intel, AMD, Qualcomm, Mediatek, etc. I think it'll be good.
I'd like to see what Nvidia could do with ARM.
Everyone is well aware of Nvidia's Grace CPU. It's just an example of how Nvidia needs custom CPUs and can't rely on AMD/Intel CPUs.
But even with Grace, Nvidia is just relying on stock Neoverse cores. The components and interconnects around the core are custom, but the core is still just ARM Neoverse. This is probably where Nvidia wishes it can control the roadmap of ARM to better serve its goals.
I have no problem whatsoever with the idea of Nvidia owning CPU IP and adding tot he competition. Competition is always good. But I wouldn't want to give Nvidia control over the single most ubiquitous CPU IP on the market — no matter how many safety checks one puts in. ARM IP should not be controlled by a corporation that has a horse in the race. That's the beauty of the ARM business model — they don't make any products themselves.
So are you saying that ARM should never be allowed to sell to Apple, Google, Microsoft, Facebook, Amazon, Intel, AMD, Qualcomm, Mediatek, Huawei, Nvidia, etc? They can only sell to someone who doesn't make chips or is in tech? That screws ARM employees (who own equity) and ARM shareholders. I don't like the idea of restricting businesses like this. Let's open up the market. Free market.
In reality, if Nvidia f's up ARM, there is RISC-V and Intel is opening up x86. There are competing ISAs. This is the beauty of competition.
But companies like Apple would not be affected by Nvidia's acquisition of ARM anyways. Apple has a perpetual architecture ARM license and Apple doesn't even use any stock ARM designs.
What prevents Nvidia from designing its SOC using ARM ISA? Can Nvidia pay only for the ISA and not for the ARM's designs?
The lack of competition is what resulted in severe stagnation in the x86 market in the last decade. Intel and AMD controlled the computer and server market for so long without a 3rd competitor. A duopoly was barely competition. It's better now that Apple can compete.
I'd love to see Nvidia enter the ring too. Nvidia has been the most innovative semi-conductor company in the US, in my humble opinion.
Nvidia has an architectural license and uses it to build custom cores for Tegra. They don’t need control over ARM’s roadmaps to execute their own custom cores and roadmaps.
In fact they would be severely limited in what they could do. To ensure the deal goes through, Nvidia is in part promising that ARM’s current customers wouldn’t be at any disadvantage to Nvidia if Nvidia buys ARM. Difficult to see how that is possible if Nvidia is directing the core designs.
If all Nvidia wanted out of the deal was CPU IP that they controlled and they didn’t believe their own custom core designers were up to it with an ISA license … they could set up a joint venture company with ARM to develop the cores which would obviate all the criticisms. Just an aside such an arrangement is why ARM exists in the first place. ARM was a joint venture between Acorn, Apple, and VLSI for exactly such a purpose. Obviously the company grew beyond that.
Yup. See above.
Last edited:
Apple, Qualcomm, and Mediatek all use ARM cores and none need to own ARM.
Actually in part yes. Nvidia claims that they didn’t go out to buy ARM that SoftBank came to them. I believe them because the first reports were that SoftBank approached Apple first who turned them down. Why? Because again Apple doesn’t need to own ARM *and* reportedly they believed they’d never be allowed to own ARM for much the same reasons as Nvidia is being scrutinized now. Too much conflict of interest and they have less than Nvidia!
The only solution I’ve seen that involves an ARM customer owning ARM that could work was a consortium deal where all the customers pony up in a coop to own them. Otherwise ARM is no longer a neutral party and the ecosystem relies on that. So this is less about Nvidia in particular than the obvious conflict of interest between a customer of the ecosystem representing a broad number of other customers also owning the ecosystem.
Last edited:
Nvidia can compete without buying ARM, plenty of CPU design teams out there and they have it own core designs and designers already.
As for Frontier and Aurora, small number stats: Nvidia lost two contracts but have won a bunch more and likely will in the future. Just recently they won contracts for the Polaris and Perlmutter supercomputers.
There have been so many times in so many contexts where people have said, “person, company, country X would never do Y, that would be self defeating!” Years after Y happened, “well really, who could’ve predicted that?” The goals of Nvidia, ARM, the ARM ecosystem are not always aligned. What they consider messing with or harmful are not always in agreement.
Last edited:
Huch, what gave you that idea? I am saying that Nvidia owning ARM is a conflict of interests. ARM's business is selling CPU licenses to various hardware companies, while Nvidia business is building hardware. Nvidia is not interested in licensing IP to their competitors.
Me neither. But its not a free market if you are closing up technologies. Let's say that the contract prevents Nvidia from revoking licenses. Fine. But what's to stop them from refusing to license future ARM IP? You want latest and greatest CPU? Buy the Nvidia phone! Samsung and Qualcomm can go screw themselves.
A lot of companies, organisations and individuals have invested time and resources into making ARM ecosystem great. Giving control over the core technology to an interesting party is exactly the opposite of opening up the market.
If only it would be so simple to bring up a competitive CPU architecture...
Apple wouldn't be affected. Everyone else might.
Nothing. They likely want ARM to access the talent.
You’re not getting it.
Nvidia thinks ARM can do way better. Despite being such an important architecture, its shareholder value has not increased proportionally. ARM has barely made a dent in the server and supercomputer market for example despite many years of promise. AMD and Intel is still kicking ARM's ass there. For laptops, only Apple has successfully used ARM.
Nvidia thinks it can do better with ARM.
Why restrict the sale of ARM, which hurts the shareholders and employees of ARM, in order to increase the shareholder value of Qualcomm, for example?
Where is it in the law that ARM can only be bought via a consortium? And will this consortium pay as much as Nvidia is willing to pay?
Many businesses that get acquired have a conflict of interest. It doesn't mean governments should stop them. Free market.
Then you don't support a free market then. You're supporting a conditional free market. In reality, if Nvidia wants to destroy ARM by refusing to give out future licenses, then other companies will respond by adopting something else. Nvidia will just screw ARM, which they will have spent $40 billion to acquire.
It isn't. But Nvidia acquiring ARM will have no impact on a company like Apple. If Nvidia signals that it wants to close ARM to outsiders, customers will have plenty of time to adopt something else.
And ARM/Nvidia are affected by not having this deal go through. You seem to be unfairly favoring one side. In my opinion, consumers will benefit from this deal. I think Nvidia will help accelerate ARM's adoption in laptops, desktops, servers, and supercomputers, giving everyone another option besides AMD, Intel, and Apple. In addition, I think Nvidia IP, money, and leadership will make stock phone ARM designs more competitive against Apple.
Last edited:
The huge and very possible risk with Nvidia acquisition is that somehow they can force developers in to some proprietary stuff like CUDA and have even more monopolistic power.
Nothing stops Nvidia to build great custom CPU (or whatever tech) by licensing ARM IP, if Apple can make this why they want to do differently? Nvidia have foresee that efficient architectures are the future, right now ARM is the major player in this field and most companies can freely (I mean without artificial proprietary restriction) benefit from ARM design/license.
Nvidia don't want to destroy ARM, they just want to profit from this situation and to force everyone in to their stuff like they have done with CUDA etc.
They can promise whatever they want but antitrust regulators from all over the world are trying to investigate/stop this acquisition for very good reasons.
When they acquired Mellanox nobody was concerned because there are alternatives and competitors, but when you are trying to buy something used by almost any human being, the story is a whole lot different.
Nothing stops Nvidia to build great custom CPU (or whatever tech) by licensing ARM IP, if Apple can make this why they want to do differently? Nvidia have foresee that efficient architectures are the future, right now ARM is the major player in this field and most companies can freely (I mean without artificial proprietary restriction) benefit from ARM design/license.
Nvidia don't want to destroy ARM, they just want to profit from this situation and to force everyone in to their stuff like they have done with CUDA etc.
They can promise whatever they want but antitrust regulators from all over the world are trying to investigate/stop this acquisition for very good reasons.
When they acquired Mellanox nobody was concerned because there are alternatives and competitors, but when you are trying to buy something used by almost any human being, the story is a whole lot different.
Last edited:
It sounds like Softbank no longer wants to own ARM. So really the question becomes who should buy it to "pick up the slack" as it were. I mean the alternative is Softbank shutters ARM which seems worse, right?
Of course no one at apple ever thought of any of this, the Apple silicon transition has been a complete failure, if only they had listened to you they could be selling water-cooled scalder lake Intel laptops with 4 hours of battery life
While your points are completely valid, when apple does release Apple silicon Mac pros, they will be fine for the target market. Rumors also have apple releasing some new hardware for the existing line as well
While your points are completely valid, when apple does release Apple silicon Mac pros, they will be fine for the target market. Rumors also have apple releasing some new hardware for the existing line as well
Seeing to these guys like to move, and given Johnny Srouji’s initial comments on AS, I would not be surprised if there are two new processors introduced- “M2/Pro/Max” and another group of processors for the desktops with Apple cranking out releases, similar to Intel’s ‘tick-tock’, but desktop to mobile. In my opinion, the MacPro being announced after the iPhone and pre-orders the year for a very late 2022 release, or a possible 2023 release to further ease the recovering supply chain, if in case there are setbacks (Covid, silicon, yield, and etc.) However, we waited 5/6 months for an apple silicon Mac after the chip was announced. Either play would not surprise me.
It’s an interesting situation but, personally, I’m not too concerned for a few reasons.
First and most obvious, Apple wouldn’t have chose to make this jump unless it saw a future of growth in not only features but performance. Remember, AS is more than simply raw power or efficiency per watt, it’s the benefits of the ‘whole widget’, as Apple itself said.
Second, Apple likely isn’t in a rush to release a replacement for the 2019 7,1 Mac Pro. This is because, as rumours have pointed out, they may have designed a different type of performance computer - in this instance a Mac mini Pro/Pro mini. It likely won’t be modular, but it’s a performance-driven machine that suits a customer segment. It will be mean that the niche segment that does need internal modularity can still have a Mac.
This ties into the third reason, which is that Apple can keep the 7,1 Mac Pro for as long as it wants and still claim that the transition is complete. They don’t have to update that machine, but they can wait until AS is a stage where it matches or exceeds what the Xeon and AMD chips are currently offering, and then use AS for that chassis.
And I’m probably in the minority here, but there’s a good chance that the supposed Mac mini Pro is actually a Mac mini with a Pro/Mac chip in it. Even the current M1 Mini would have a suitable thermal envelope for those chips, so it’s a safe bet.
First and most obvious, Apple wouldn’t have chose to make this jump unless it saw a future of growth in not only features but performance. Remember, AS is more than simply raw power or efficiency per watt, it’s the benefits of the ‘whole widget’, as Apple itself said.
Second, Apple likely isn’t in a rush to release a replacement for the 2019 7,1 Mac Pro. This is because, as rumours have pointed out, they may have designed a different type of performance computer - in this instance a Mac mini Pro/Pro mini. It likely won’t be modular, but it’s a performance-driven machine that suits a customer segment. It will be mean that the niche segment that does need internal modularity can still have a Mac.
This ties into the third reason, which is that Apple can keep the 7,1 Mac Pro for as long as it wants and still claim that the transition is complete. They don’t have to update that machine, but they can wait until AS is a stage where it matches or exceeds what the Xeon and AMD chips are currently offering, and then use AS for that chassis.
And I’m probably in the minority here, but there’s a good chance that the supposed Mac mini Pro is actually a Mac mini with a Pro/Mac chip in it. Even the current M1 Mini would have a suitable thermal envelope for those chips, so it’s a safe bet.
Why wouldn't better efficiency be desirable in a desktop machine to? Especially companies which run dozens or hundreds of Mac Pros surely wouldn't say no to powerful but yet efficient machines. Better efficiency = less power consumption = more money saved.
Last edited: