The M1 as a computer is too locked down for me, and it seems like for you. But the M1 as a demonstration of the power of modern ARM chips is absolutely unmatched. I have seen a lot of people use the lockdown argument to justify a disregard of ARM overall without realizing that no, the open version of this technology will be incredible.
> the M1 as a demonstration of the power of modern ARM chips is absolutely unmatched
I feel like it's more a demonstration of Apple's competency in vertical integration, cultivating talent, reducing the impact of politics on product/technical quality and ensuring productivity within and between technical teams. Aspects where Intel has abjectly failed and hasn't found any solutions for.
Intel has tried repeatedly to push into new areas, but they've failed every time. Meanwhile, Apple has somehow figured out how to push into new market segments successfully in a crazy consistent way and the M1 is a clear indication of how healthy and functional they are as a business. Like, when was their last real failure?
Apple is absolutely leading on this front, starting in the phone space. That being said, I don't think their lead over other distributors is especially large. Apple has taken ARM seriously, and built performance chips in their mobile devices that if not for thermal constraints would handily compete with x86 systems. That being said, the same is true of a Snapdragon. Apple ARM, and certainly M1, is better than the competition, but that's because no one else has looked for performance from ARM. The chip designers haven't really been pushed yet. Now that M1 has demonstrated what can happen when the technology is freed from the phone, we'll see people catch up.
I believe Apple has significant advantage to design CPU cores compared to Qualcomm. They acqhired PA Semi team and has been designing great performance cores continuously compared to ARM Cortex (Especially for Single threaded). Qualcomm no longer develop own CPU core since ARMv8 transition.
Oh I'm not against ARM at all. In fact I have many systems running it :) Raspberries but also a Pinebook. I'm also getting an M1 mini from work.
I just don't like the almost-religious overhyping of everything Apple. They're doing a great job. They're not perfect.
For example: Moorhead makes a good point IMO. Yes he's talking about software issues. But in the case of Apple, you can't separate those two. The OS doesn't work on other hardware and (in this case) the hardware doesn't work with other OSes. This works to their (huge) advantage in some cases, being able to optimise each component perfectly. But it can be a drawback too like it is here.
> For example: Moorhead makes a good point IMO. Yes he's talking about software issues. But in the case of Apple, you can't separate those two. The OS doesn't work on other hardware and (in this case) the hardware doesn't work with other OSes. This works to their (huge) advantage in some cases, being able to optimise each component perfectly. But it can be a drawback too like it is here.
Moorhead didn't make that point.
Moorhead made the point that a small amount of software that's mostly rarely used on the Mac platform isn't working perfectly. And even the mass market software he's pointed at, like Adobes is already in beta test with native versions about to be released.
> I just don't like the almost-religious overhyping of everything Apple. They're doing a great job. They're not perfect.
I'm in the same boat as you and the other OP up-thread. I'm amazed by what M1 has accomplished, but less so because of what it represents for itself (and for Apple) and more that it shows ARM is capable of.
As an architecture, ARM is impressing the Hell out of me as of late. Sure, it's not touching Intel and AMD in terms of raw performance, but once you start looking at the performance-per-watt, the story is absolutely different. x86's power budget is looking awfully bloated by comparison.
What happens if NVIDIA's acquisition of ARM goes through? Do they start producing their own CPUs?
In your quest to avoid almost-religious overhyping of everything Apple, I'd contend that you're presenting the inverse: almost-religious anti-hyping of everything Apple. We can criticise them for their seemingly blinkered focus on casual consumers and end-users at the expense of developers and technologists without finding ways to rationalise away anything good Apple does.
Attributing the positive aspects of M1 to the ARM micro-architecture is contemporary revisionism. If micro-architecture was the key differentiator, Microsoft Surface RT would have been a runaway success and Qualcomm would have taken over the notebook CPU market by now. Little of what makes the M1 so impressive has anything to do with the ARM instruction set. The impressive parts are things like the entirely in-house GPU and the mild hardware acceleration of Rosetta 2.
It's also worth noting that Apple has been deeply involved in ARM since the late 1980s. They were a co-founder of Advanced RISC Machines Ltd. along with Acorn and VLSI. Apple were deeply involved in development of the ARM6 architecture and the Apple Newton was one of the first major commercial applications. So even if you limit your amazement to the ARM architecture, a portion of that credit goes to Apple anyway.
> In your quest to avoid almost-religious overhyping of everything Apple, I'd contend that you're presenting the inverse
I must point out that I believe you misinterpreted my intent. I'm largely indifferent to Apple. I'm not entirely sure if you'd find that more offensive or less since I'm interpreting this response as one of annoyance (I'd appreciate a clarification to this end).
I don't use Apple products. And you're absolutely right, the M1 is an impressive feat.
Understood. Intent is often difficult to robustly convey without being neurotically verbose.
I promise my response was not borne of annoyance. I just got the impression that you wanted to focus credit on the one thing about the A1 SOC which could plausibly attributed to anyone other than Apple. That seemed a stretch, is all.
Unfortunately, I know what you mean. I have a tendency toward verbosity that makes it feel that every time I write a post I'm crafting a treatise that might make the great statesmen of old blush (with the exception that I'm far too stupid to make it as interesting or entertaining).
And yes, intent is difficult to convey through text even if you're afforded a great deal of time and effort.
> you wanted to focus credit on the one thing about the A1 SOC
No, not at all. The ML cores and other, err, accessory CPUs they've added are incredibly fascinating. Mostly I was thinking of the ARM portion of the hardware.
Truthfully, I confess that part of my excitement lending itself to ARM is because I genuinely want to see more vendors follow Apple's footsteps and either a) craft their own designs or b) build out ARM-based platforms that will lead toward greater competition in the desktop/laptop markets (or both?). Fortunately, Apple's tendency toward setting market trends gives me some hope that we'll see noteworthy contributions that embarrass x86 and upend the idiotic convention that I think M1 has finally broken. Namely that you can have a chip that works on an incredibly tiny power budget or you can have performance. Apple's proven we can have both, and I think that's a good step forward. It excites me that we may actually see viable competition this decade, and if we do, I'll absolutely credit Apple with upending conventional wisdom and proving that our fixation on x86 for "performance" was myopic and unnecessary.
Maybe that explains my reasoning a bit more. At the expense of being unnecessarily verbose (though you may appreciate that). :)
Also, I owe you an apology. I wrote my reply out of some frustration last night. Regrettably I allowed myself to get sucked into a "discussion" with someone I didn't know was an antivaxxer until I realized their refusal to listen to an explanation of microbiology wasn't born out of lack of interest but rather suspicion. When I saw your post, I genuinely believe I unfairly took some of that frustration out on you. I'm truly sorry about that.
The explanation is appreciated. We’re all human and passions drive us in weird, occasionally unpredictable ways. I’m undoubtedly as guilty of it as anyone.
> We’re all human and passions drive us in weird, occasionally unpredictable ways.
Very true. It's just our nature, and we're all fallible beings in some way or another. Lord knows I'm at the top of the fallibility list.
> I’m undoubtedly as guilty of it as anyone.
I am as well; more than most, in fact.
> Also, fuck those anti-vaxxers.
Agreed!
I didn't realize that was their opinion at the time, so I shared a bunch of medical literature in the hopes of supporting my argument. What I didn't realize was their opinion of any supporting evidence was dismissed as clearly "paid for" by, uh, $LARGE_FACELESS_ENTITY.
While I'm sure that's true in some cases, the mRNA COVID vaccines are fascinating because the principle has been around since 1989. I'm hoping it works, not specifically because of SARS-CoV-2, but because it shows some possibility as an anti-cancer vaccine. We just need to prove the delivery mechanism is functional.
I really want it to work so that we can manage or eliminate other diseases as well. I think that's why I was so frustrated/disappointed.
> I just don't like the almost-religious overhyping of everything Apple. They're doing a great job. They're not perfect.
That may be so, but again that doesn't really apply to the M1's performance as hardware. If the article was about Apple as an investment, or Apple as the designer of the next computer for me, I'd agree that taking into account their software would make sense. But the whole point of the article is that equating Apple's being Apple with the M1 being a badly designed chip is wrong.
M1 is really great but I don't like to see the presentation style like "n% faster than top selling laptop!" (What model, workload, situation, etc...?). Please just show benchmark numbers and models.
