Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Qualcomm CEO Outlines Plans to Compete With Apple Silicon, Handle Apple's Shift to In-House Modems
- Thread starter MacRumors
- Start date
- Sort by reaction score
I like to see healthy competition. That ends up being good for the consumer. Let's get ready to rumble.
The main issue with anyone not just Qualcomm, is that making an ARM SOC is just part of the performance gains Apple is seeing…. It’s also tied to having control of the entire pipeline which includes the OS.
AMD, Intel, Qualcomm will still be subject to a random version Windows or Linux, which will come with thousands of manufacturers fighting/ arguing over which drivers, implementations, hardware placement on boards, etc. are the best for their products.
Which as we know will cause tons of inefficiencies and incompatibilities.
And this doesn’t even account for the multiple board suppliers who think “their way” is the best way.
Soooo many different variables will end up just making Windows ARM on a Qualcomm chip run like crap.
AMD, Intel, Qualcomm will still be subject to a random version Windows or Linux, which will come with thousands of manufacturers fighting/ arguing over which drivers, implementations, hardware placement on boards, etc. are the best for their products.
Which as we know will cause tons of inefficiencies and incompatibilities.
And this doesn’t even account for the multiple board suppliers who think “their way” is the best way.
Soooo many different variables will end up just making Windows ARM on a Qualcomm chip run like crap.
For my use and the discounts I got, the M1 mini was the most cost effective solution and it is completely silent. I probably save another 100$ on power between the M1 and decrease in AC costs. When I include the resale value of the Mac Mini the savinging on Power will cover the depreciation of the computer. I'm not an Apple fanboy, there is a reason my last mac was purchased in 2012, the M1 Mac was the first time in a long time where I found superior value in a Mac.
It will be interesting to see how many people would be interested in a. Qualcomm ARM chip running windows or Linux. I could see it even if M$FT did it or even Intel. It will probably be interesting to see how Apple’s in house 5G modems compete with Qualcomm modems at the beginning because if Apple doesn’t get that right (I’m sure they will) on the first shot will that severely impact iPhone sales. I’d say yes. That though is only an opinion
You know why intel was on top a decade ago and now apple is on top and intel behind for quite some time?
Because of the same brain. That man knows silicone like no one else
Snapdragons are not in any way crap compared to Apple silicon. They compete very well. We are talking about a 10-15% differential and sometimes Qualcomm comes out ahead on the benchmarks.
One thing people forget is that Qualcomm has experience with high performance, custom ARM cores and silicon.
They had the Centriq. It had 48 cores running at 2.6GHz with six channels of DDR4 and 60MB of L3.
Qualcomm is not new to desktop and server chips.
What is the use case and profitable business plan for Qualcomm? Isn't the entire smartphone market all running on ARM SOC style chips(I believe so). That is a mature market. So is Qualcomm going after the windows desktop/laptop world? That's a big market but it also will require a pretty serious change in direction in that market.
Apple obviously has a large enough market within its own products.
Apple obviously has a large enough market within its own products.
So won’t the Qualcomm tax where you have to license all sorts of unrelated IP doom it?
Apple = Qualcomm in my opinion. And all the legal disputes between the two companies are public corporate theatre intended to keep anti-trust monopoly lawsuits at bay. Apple is using Qualcomm as a front to now produce windows compatible chips based on M1 tech to capture the entire windows market from Intel and others. All while attracting no attention whatsoever as the public continues to assume Apple and Qualcomm are rivals and have bad sentiment towards each other. Genius Apple.
Keep an eye out for Apple-style security and privacy coming to these new Qualcomm chips coincidentally similar to Apple’s Secure Enclave. Windows is likely already aware of what’s about to happen and that’s why there was so much vagueness surrounding the recent Windows 11 CPU security requirements. I could be wrong on all these points but I stand by this speculative hypothesis.
Keep an eye out for Apple-style security and privacy coming to these new Qualcomm chips coincidentally similar to Apple’s Secure Enclave. Windows is likely already aware of what’s about to happen and that’s why there was so much vagueness surrounding the recent Windows 11 CPU security requirements. I could be wrong on all these points but I stand by this speculative hypothesis.
Then why did they hire all those Apple engineers? Imagine if another company did that to Qualcomm. Can you say lawsuit?
I know is hard to digest, but A-series chips are up to twice faster -yes, 100% more- than Snapdragons when normalized by energy consumption, and around 30-40% faster all in all (i.e. the Snapdragons consuming more energy), here, have a look:
The iPhone 12 & 12 Pro Review: New Design and Diminishing Returns
In regards to Intel, you can read along the bottom: "...the new A14 is massively impressive and grows the performance gap compared to the competition. Apple has now reached higher single-threaded performance within their phone SoCs than what Intel can deliver in any of their designs..."
It just takes a quick glance to picture the gap between Qualcomm & Apple:
(a) I suspect there is massive truth to that. Good ideas are important, but so is great execution with all that entails -- making cautious decisions when appropriate, make extreme decisions when appropriate, forcing the correct pace, having a plan B, being willing to spend money, etc etc.
QC doesn't have a great track record in most of these execution issues, and just having good ideas doesn't change that.
(b) There ARE *many* very good ideas in Apple's SoCs. Before I began investigating the M1 in great detail, I expected mainly to find that Apple was the first to implement good ideas that had already been suggested in academia. I did find a lot of that, but I also found a variety of very clever ideas that I've never seen in the literature. So, absolutely, Manu and GW3 are damn impressive people (John Bruno I know less about; I didn't see his name on any of the patents I looked at). BUT, and I cannot stress this enough
- even with all the cleverness in the M1, there remains a large reservoir of academic cleverness that is still not yet implemented AND
- there's a LARGE pool of people at Apple coming up with new ideas. As I've worked my way through the patents (a huge job, not close to done!) I've seen the evolution of the names. You see the same names over and over again in the early PA Semi patents up to 2014 or so, but I guess once the A7 shipped (2013) and it became clear that something extremely interesting was happening at Apple, every ambitious and inventive SoC/NoC/CPU/GPU/fabric engineer in the world has considered going there, and the pool of names on the patents has just exploded.
"Those skilled in the art" know a few names from the earlier days of CPU design -- Andy Glew, Glenn Hinton, Jim Keller, and now GW3. But note how the names keep changing. There is only one GW3, like there is only one Jim Keller -- but there are constantly new people with the vision to do things differently, and as long as a company can find those people and is willing to give them what they need, life goes on.
Dude, you are just embarrassing yourself. Stop now.
There are people on this site who have forgotten more tech details than you know, and you don't want them drowning you in a hundred pages of "here EXACTLY is how Apple's SoC is completely different [and superior] to anything else on the market".
Not sure where you learned math, but the fastest 888 is roughly 3400 on GB 5. Last years a14 is 4200. That is way more than 10% against last years SOC. Am I missing something?
Honestly they could considering Qualcomm co-designed and manufactures the SQ series of ARM SoCs the Surface Pros currently use.
And it's a given there will be a "Snapdragon Compute Platform" for Windows 11 PCs to succeed the one for Windows 10.
The existence of x86 Hackintoshes did not cause Apple to significantly reduce the price of Macs.
I see no reason why the existence of ARM Hackintoshes would be any more successful in that endeavor.
I want Windows on Arm to succeed, (although I have my doubts), but I’d rather Qualcomm is not the one to do it. They are a monopolist that has held back Android phones for years. Much rather that Microsoft partners with AMD (they do semi-custom SoC, had the ARM K12, good graphics) or NVIDIA (did Tegra ARM SoC, buying ARM, good graphics) for WoA.
macOS isn’t going to run on other ARM chips because:
1. M1 uses iBoot, other ARM chips use other booting processes
2. macOS does not have graphics drivers for Qualcomm GPUs
Why haven’t those people left Apple right now, when you can run macOS on bog-standard x86 chips in laptops? Why hasn’t that $599 Dell already pressured Apple to lower prices on Intel Macs?
Qualcomm already does all that with the 8cx- essentially an overclocked phone SoC. It’s not remotely competitive with Apple. With NUVIA, it seems that they are going for PCs first, then bringing the cores down to phones, then maybe servers. As for 5G, I hope not. Bundling modems in with their SoC is what allows them to monopolize the Android market in the west. It would be a shame for them to the same for PCs (although of course they will).
Qualcomm isn’t a threat to Apple. They’re a threat to Intel.
The ARM plan is to replace the current desktop market owned by Intel with QC chips.
Of course QC today have the chips inside (some) tablets and chrome books, but they believe they can replaced Intel in laptops and AiO's.
It's not a crazy thought! Obviously
- Apple has already implemented phase 1 of their version of this
- QC believe that they can ship a SoC at around Apple levels
- MS are, in usual MS fashion, dithering and straddling, but seem kinda sorta prepared to really support Windows on ARM (just as soon as they can do it without any longer having to care about what Intel thinks)
So the issue that matters is step 2 of this plan. Can Manu and GW3 (with a team of engineers, maybe three years of work, and some QC IP to build upon) create something that matches Apple A15/M2?
Can they build something that's better than Snapdragon today?
Can they build something better than Intel?
Clearly better than Snapdragon today.
Probably better than Intel (honestly, my god, it's no longer a high bar. How the mighty have fallen! And they seem unable to get their act together and get up; they just keep digging a deeper hole.)
Better than Apple? There I am doubtful. The Apple bench of IP and engineers is now so solid that they can just cruise for a few years implementing only obvious improvements and still be ahead of everyone.
But I do look forward to seeing what QC ship!
Last edited:
They tried, but Intel couldn't deliver modems with the quality or feature-support that Qualcomm could and it was hurting sales of iPads and iPhones.
Hence why they tried to sue them to offer better terms to adopt them again and when that failed, settled out-of-court to do so.
Of course, Apple also bought Intel's modem business to give them the foundation to develop their own modems with the quality and feature-support of Qualcomm modems and once they do so, they be able to once-again "ditch" them.
Great write up.
If QC can do it then I'd want to own their stock. That is a seriously big, lucrative market. But I suspect it's A LOT easier said than done (to take the desktop/laptop market over)
I think Apple has, the base model Mac's with M1's beat high end intel chips, IMO this is like a price drop.