Hacker News Reader

Comment by RandomWorker | Hacker News Reader

I have a M1, got it because I wanted to try something new. Now I love my M1, but it’s a special time. They capitalized on Intel/AMD sleeping on RISC architecture and going to smaller nanometers. While apple is strategically buying up time on a limited production capacity like no one else can. Innovative design integrating all kind of slow components. I think it was an opportunity that won’t last, intel and amd are waking up to this fact. I see AMD and Intel coming up with some RISC and hybrid type solutions. Where they might not be able to do better is the PC integration. Memory and processing have always been separated and it’s going to be hard to integrate those just because of the cooperative structures. I think Apple can stay ahead if they keep going. If there is a M2 this year that is better, I’m buying it again with more memory, because M1 has been workable it’s not optimal for all my workflows.

View on HN

AMD does ARM stuff and has had plenty of time to get it right.

Intel has had, what, three, four bites of the RISC cherry? They've owned big chunks of the ecosystem Apple now exists in. Evidently one of their pitches involves RISC-V and specifically SiFive.

I think it's a mistake to see the future as Apple vs Intel or AMD; that is the battle of the first -- nearly over -- quarter of this century.

The future is Apple vs Samsung and some Chinese megacorps we don't necessarily recognise. Because the battle is now moving on from large-scale off-the-shelf products and chips for motherboards to SoCs and large-scale bespoke design.

It will be the speed and performance of native apps on Apple and Samsung ARM silicon vs the freedom of choice and universal portability of WASM on RISC-V and the-long-tail-of-every-other-architecture.

I think Intel is too late. They can play catch up with the ARM market, but it will take time because there are are enough other players that they're not relevant enough to be on top again. That is unless they can out-innovate the competition, which I doubt, but who knows.

Over the past few years I've read articles and watched videos on Intel making some kind of comeback, yet the journalists creating those pieces seem either oblivious to RISC or are ignoring it. I don't know which one. That tells me all I need to know, and so far it isn't looking like Intel is the future for consumer electronics.

Well. I think x86 is dead and Intel just doesn't know it yet. Might take a decade for them to realise.

Intel as a business? Maybe as a design house they will struggle to catch up. As a foundry? If they do not catch up with the amount of government support they are going to be able to claim, then there is something very wrong.

Most chip designers have at least one way forward, which is to produce chips that run WASM well on Linux or a BSD. This is actually a good time for minority architectures, because a truly portable architecture is on the horizon, and everyone has an incentive to see it work; that incentive being not losing all their business to Apple, Samsung and a few Chinese firms whose names we don't recognise yet!

>Well. I think x86 is dead and Intel just doesn't know it yet.

Let me toss that in my list of sentences that will poorly age, alongside "This is the year of gaming on the linux desktop" and "surely nothing wrong will happen if I plug in this SCADA on the Internet"

There was a small measure of satire/hyperbole in what I said. But you have separated that one sentence from the second sentence that provided context, so I don't think I particularly need to defend it on its own.

Perhaps I should have used a semicolon to defend against you.

The point was that in ten years time, I think people will look back and say: 2022 is the year x86 really ended as a forward-looking architecture. Ergo: dead, they just don't know it yet.

But to expand: one of the things that is very striking about tech stuff is how quickly the energy in any one system can deflate.

Looking at the Apple competition, do you think buyers will think:

1) Intel simply need to make x86 a bit better to compete, or

2) It is about time we started looking at architectures that can scale up to compete with Apple Silicon on all levels (absolute CPU power, power consumption, mobile to server)

Architecturally, x86 is in serious trouble and this isn't just something I -- a random non-chip-designer -- think. It's what Intel think, or their entire recent push would not be to become a fabricator for other chip designs, built on an unhealthy amount of anti-asian supply chain FUD.

Once again, no. For the price of a Mac Pro, you get x86 hardware that performs simply better. Despite HN having a hardon for "b-b-but it's only 85W", it doesn't matter to a very large portion of users. I don't give a single crap if my workstation uses up 600W to run a 5900X and an RTX3070 (or, well, if I could find one, but shhh). I merely care about raw power. My electricity bill is a non-factor. And that's just with something of "equivalent power" (not even taking into account the fact that people don't give a shit about OSX if it doesn't run their software/games/etc). I can get some x86 hardware that runs an M1 into the ground easily, and given that all the current gen is on old processes, upcoming x86 hardware will, once again, be more than just competitive, but better. Then Apple will come out with the M2, reap all the benefits that come out of being a non-upgradable SoC with soldered RAM and soldered CPU, maybe be on top of very select benchmarks for a while, etc. As with every time that Apple came out with "revolutionary new hardware", what it mostly meant is "we slapped TSMC with a load of cash and got their latest and greatest process".

So, to respond, buyers will:

1/ Not give a crap about whether it's x86 or Apple Flavored ARM (read: an ARM with undocumented extensions that is pretty much CISC), but merely look at price, in which case Apple buyers are an extreme minority. A $500 laptop with a mobile Ryzen is more than enough for the average user, and mostly, most people cannot afford a $1200+ macbook air.

2/ Not care about power consumption because it does not matter to a boatload of people. Get out of your bubble, noone outside of software developer nerds care if they have to plug in their laptop once every four hours. Most people have lunch breaks. Is it a great thing to have in a laptop ? Absolutely, and Apple has done great in that regard. Does it matter anywhere else ? Absolutely not. Once again, it's HN jerking itself off about pErFoRmaNcE pEr WaTt when Apple put out a CPU that can't even draw the 105W it's rated for, and has no proof that it can improve in single threaded performance with more power. But sure, toss out more cores, it's what x86 has been doing for years, but when Apple comes out with the M1 Ultra it's a stroke of genius to slap two of these bad boys together.

3/ People actually upgrade their hardware and don't rebuy a macbook pro every two years like half of this forum does.

ARM is not a forward looking architecture, and neither is x86. They both are architectures, with their advantages and flaws (and the Apple flaws are fucking massive, unupgradable-except-if-you-buy-next-year's-version is the very definition of non-forward thinking, especially for end users).

Not going to respond to the unnecessarily rude tone here generally, except to say that any argument you want to have taken seriously is better not couched in terms of "hardons" and "jerking itself off". If you are a teenager, consider trying not to sound like one. If you're not a teenager, consider trying not to sound like one. Either way, this kind of communication is why you're likely still not taken as seriously as you could be.

Also you're taking it all off on an angry tangent about the HN audience when I am talking about buyers (I meant including institutional buyers, not end users, and I didn't clarify that enough but I would have thought the focus on architecture would have made it clearer) but:

> I don't give a single crap if my workstation uses up 600W to run a 5900X and an RTX3070 (or, well, if I could find one, but shhh).

I would. Between this April and next April my energy bill will double, and it is unlikely to ever fall back to where it is now. Power consumption matters A LOT in Europe right now. Corporate buyers will have to care about that, for the long term.

> 3/ People actually upgrade their hardware and don't rebuy a macbook pro every two years like half of this forum does.

No, they really don't upgrade their hardware, in fact. Almost no computer buyer upgrades through anything other than replacement.

But I'm using a seven year old, secondhand MacBook Pro, so I'm not the target of your comment.

For what it's worth, I recently went and traded in a pile of old Macs because I was travelling with a newer Macbook Pro (16" with M1 Pro) and needed to be able to effectively replace not only my first early M1 laptop, but also the iMac Pro I'd got as a flagship machine before the M1s came out.

Between those and an intel Mac Mini and an older iMac, I got over $3000 in trade-in value. Made me feel quite foolish for not having tried that before. The oldest iMac still got me a couple hundred bucks, as would your seven year old MacBook Pro most likely. A two year old high-specced computer?

I get that the list price for any of these is still way higher than putting together a PC from parts… but again, I belatedly traded a bunch of stuff in, and got more than $3000 in gift card value from Apple doing it. That makes 'rebuying a macbook pro every two years' a whole other state of affairs. I'm kicking myself for not having tried that earlier: I'm sure I left over $1000 on the table simply by letting recent computers sit around unused rather than immediately trading them in for Apple credit.

> Power consumption matters A LOT in Europe right now.

Should have gone with nuclear years ago, our prices are stable.

Well, I was going to spend this money on upgrading my laptop, but I guess I'll just invest in a nuclear power plant instead... /s

At the cheap end, it's ARM too. Chromebooks if you need a keyboard, tablets if not. It's only enterprise corporate office drones that are tied to windows. I've got a gaming rig too, but we are also a minority. Most people don't need a $1200 Macbook air, but also most people don't need a $800 RTX3070. PC Gamers are already a niche. Mobile gaming is where the money is, then console, then PC games, and mobile is 10x the both of them.

So to the point at discussion: Apple uses their common tech across their entire Market. The M1 chips are the result of making the fastest mobile phones. The M1 is going into tablets. They are just printing money and pumping that money into chip design. If your DIY PC Building is what is going to keep Intel/x86 alive, then I agree with GP: Intel is the walking dead. And it really is just us gamers.

Who needs raw CPU/GPU at their desk? I had one of the first nehalems under my desk (NDA and everything). Then I heard about AWS and started running my shit on that. Back then, that was intel. Now it's Graviton ARM: more processing per $.

In fairness, are you not merely implying the opposite prediction? I'm sure there were people who thought pagers/beepers wouldn't instantaneously disappear from use, and there were plenty of developers who thought Java was going to be the language for the web. I'm not sure how pointing to the year of the Linux desktop is a good argument.

Though I think x86/x86_64 inevitably will outlive its usefulness for general purpose computing, but it's not literally "dead". There's too many Intel architecture devices out there still. To me, it seems clear that it's more of a dead architecture walking, because I really can't think of anything that x86 does better for most use cases. I'd love to be shown wrong, though.

You're rewriting history there. Intel has tried, tried, TRIED! to compete against ARM since it's been clear all the market growth was happening there around 2010.

Their most credible strategy was to produce modems to steal Qualcomm's insane margins on those chips and break the forced mariage of their modems with their not-so-good processors. It was such a resounding failure that Apple bought the whole division for scraps.

I don't see how Intel suddenly gets its shit together now that the ARM chips are in computers. The fact that the chips ended up in computers is a direct consequence of their decade of failure. I expect Apple to stay ahead of Intel and AMD in terms of power per watt for a decade at least.

It's kind of the same situation with Tesla vs. the big car companies. They were always going to catch up, and they made huge progress in the last few years. AMD and Intel have spent decades trading places, sometimes with unethical practices that would make 2000s Microsoft blush. I'm not sure Apple is prepared for what happens when AMD and Intel both come for them.

The personal computing market is a nothingburger in the grand scheme of things. Apple alone sells more ARM based mobile devices than the entire worldwide PC market not including servers.

Even with the server market, AWS is moving toward ARM where ever it can (disclaimer I work at AWS). Even if Intel regains momentum in the PC market, it’s a Pyrrhic victory if they can’t get into the mobile market.

Even if the PC market (including laptops) is quite a bit more than a nothingburger, the segment of the PC market that really cares about the top decile or two of performance for whatever reason (real or imagined) is tiny. A lot of people are fine with a 5 year old system that's still supported, or even a Chromebook at this point.

> They capitalized on Intel/AMD sleeping on RISC architecture and going to smaller nanometers.

Apple is singlehandedly using up almost all of TSMC's 5nm processes. There are no smaller processes available anywhere in the world, and Samsung's is disappointing. Processes that have largely been driven by demand from AMD and Apple (EPYCs and Ryzens from AMD, and Apple's mobile chips). They capitalized on nothing except having fat wads of cash to slap people around with.

Additionally, it's not "being a RISC architecture" that makes the M1 a beast. Slap a 5950X on a SoC with directly available memory and GPU, and you get similar (better, even, considering current gen CPUs do get better performance already) results. "Deep integration of an entire stack makes things more performant" is not a new thing, it's merely that Intel and AMD have been upholding the contract of x86/64 that is interoperability and replaceability (well, Intel has been doing some socket fuckery for upgrades, but upgrades are still entirely possible), and Apple can afford to throw it away.

EDIT: bUt ThIs oVeRlOoKs PoWeR CoNsUmPtIoN

I swear I'm going to come out with a souped up raspberry pi and tell HN "look, it uses 1W of power for 10GFlops" and you would buy it because tHe PoWeR pEr wATt iS BeTTeR !11!!

This is the upside of Apple's distressing tendency to burn backward compatibility. As a Mac dev who chooses to support really REALLY old machines, I've experienced the downside of this, needing to keep dedicated retro machines just for building retro software on, and developing everything repeatedly on different generations of computer instead of having it just work the first time.

This is the upside: if they want to make a quantum leap, they will absolutely do so if it benefits them, and everybody's expected to just re-buy, at Apple prices (mind you, trade-ins are a thing). Apple absolutely can afford to throw away anything that impedes them from doing what they want, and in the case of the M1 processors it's serving them really well. Turns out the fat wads of cash can be used to buy useful things, and when Apple wants to, they can CHANGE THE WORRRLLLD ;)

I'm sure the PC (not necessarily 'x86') world will find ways to more or less keep up: there are plenty of people who really, really want to keep up. It's just that complacency won't cut it: PC-land has to really seriously work to not fall way behind. Good. That's best for everybody. As a lifelong Mac guy since the 68040, I'm delighted the PC world has driven Apple to these extremes, because now I get to enjoy the results.

The 5nm process is only one part of what makes the M-series so fast and efficient. Yes, RISC is another but much of Intel’s chips are RISC at some levels.

A big part of what makes Apple’s chips so fast is that they optimize them for the needs of the OS. In interviews they have given examples of how to profiled the common operations in the OS and then made sure that those operations were as fast as possible on the chips. that integration is a big part of it. Other hardware and software vendors are not as tightly coupled and don’t do this kind of optimization.

Do you own a laptop? Is the ability to burn your lap when compiling a feature for you?

> when compiling

When opening a second chrome tab*

That might compete on performance but I wonder how “easy” it would be to compete on the power-performance curve.

It's possible, yet noone is doing that. I wish someone did!

The Udoo Bolt v8 is about the tiniest x86 "SoC" that I know that runs a Ryzen V1605B, 16GB of RAM and a Vega 8 that costs $500 and scores about half of the M1 Max in Passmark. Not too shabby for something that still has to deal with DDR4-2400 and storage far away!

This overlooks power consumption

Reply on news.ycombinator.com