Live data from Hacker News

Ode to the M1

fabiensanglard.net

251–260 of 286 posts

Re: Ode to the M1

#251
post #210

Earlier quoted context omitted.

The demand for small phone is completely overblown online.

Because people who don't go online need BIGGER phones? For what?

Objectively he’s right and there’s a reason the iPhone mini was discontinued. The demand for mini phones is niche but very loud online. It isn’t economical to support a mainstream product for the handful of people who are actually willing to pay for that specific niche feature point.

Kinda like headphone ports and microSD slots on android phones. Are there models that do this? Yes! Is anyone willing to pay more for it? No! But they are very very loud online about it.

There do exist smaller android phones as well, just like a headphone port nothing is stopping you from buying one instead of complaining on the internet!

Re: Ode to the M1

#252

I wish I could upgrade from my 13" mbp that's the last of the intel i5 series. My 16" M1 Max that I was issued at my last job was a dream, and now that I'm playing with SwiftUI and dealing with xcode, it's feeling slow and noisy. Unfortunately I was laid off, so no M1, and the ram upgrade to the 64gb or 96gb is insultingly expensive still. As nice as it is, it isn't a $4k boost in productivity.

B+H has been periodically running deals on the loaded M1 Max 10C/32C 64GB/4TB for $3000-3200.

https://slickdeals.net/f/16807523-apple-macbook-pro-16-2-lap...

M1 Max doesn’t get the HDMI 2.1 port (although you can coerce some cablematters adapters to do it), or the Wi-Fi 6E, and the M2 Max is generally faster especially in gpu (at the cost of higher power - perf/w doesn’t change much) but in general it’s not worth paying $2k more for.

Re: Ode to the M1

#253

The M1 is great, but I really dislike the limitation to 2 external displays unless you get the ultra model CPU. I've just gotten really used to my portrait landscape landscape setup! And of course I'm an outlier so corporate IT won't buy me an ultra CPU version. I guess that it's to force people to pay more, but it feels like such an arbitrary decision. Sure, restrict it to 2x6K displays, but I'm using 3x4K displays…

The 15” MBA not getting a Pro variant like Mac mini is a bummer. I hope they do that at some point because I agree, a 1-monitor limit makes it a toy.

The base M1/M2 only getting 1 display does make sense from apples perspective though. M1 is fundamentally a tablet chip, just a fast enough one that it crosses the line to ultraportable laptop as well (and it’s quite fast for a tablet chip as a result). Tablets don’t need 2-3 external displays. Nor does the average user of a MacBook air. Remember this is a product that’s essentially coming from a netbook heritage, 1 display is fine in that segment.

What doesn’t make sense is not offering a M2 Pro option especially on the 15” Air which could obviously handle it. Because other people do need it, and the MacBook Pro isn’t the same product, it’s got fans and a different display and chassis etc. There is a segment who is just using the MBA as a fanless ultrabook and the Pro cpu is necessary to support the needs of this segment. I really hope 15” MBA gets an M3 Pro option at some point.

And please for the love of god put the 13” touchbar model out of its misery. It looks like it’s getting refreshed again with the M3 line. The new chassis on the 14” and the 16” is just lightyears better.

Re: Ode to the M1

#254
post #118

I believed fanboyism was left at the 2010s

You were wrong. Ever hear people talking about AMD gpus online?

It’s crazy how normalized it is for certain fanboys to just call everyone who disagrees with them “brainless”, “ruled by mindshare”, “buying it for the blue bubble”, etc. I see it all the time even on HN, it is crazy how normalized the android and AMD fans have made it to just insult a group of people to their faces. Even in a tightly moderated community!

But put out one article saying “yo M1 is a good laptop” and watch the “I thought we left fanboyism in the 90s!!!” replies flood in.

Re: Ode to the M1

#255
post #127

Earlier quoted context omitted.

What am I supposed to see there?

> Atomic was founded by Sam Zeloof and Jim Keller. i assumed you were referring to jim

I was not.

M1 was led by Wei-han Lien.

Ascalon is led by Wei-han Lien.

Jim Keller is (and continues to be) CEO of Tenstorrent.

Re: Ode to the M1

#256

Earlier quoted context omitted.

From the fine article: "Remember when Photoshop opened in 1s on SSD?" I do remember. And it now takes A LOT longer on insanely fast CPUs and SSDs. And I hate it every time I have to wait for it.

So it’s not Electron that’s slow, but modern software?

I have a liquid cooled 14 core processor that runs at 5.7GHz. 96Gb of DDR5. Top of the line Samsung m.2 drive.

Photoshop still takes a good 30 seconds+ to open.

Re: Ode to the M1

#257

Earlier quoted context omitted.

> macOS is appallingly bad at backend development That's news to me. In what way are they bad compared to Linux? Both can run docker, Intellij, Insomnia and Dbeaver.

The main thing is Docker requires emulation

True although if you develop for arm then you can use virtualisation. Still, performance wise I've not noticed any impact.

Re: Ode to the M1

#258

Earlier quoted context omitted.

> I hope Electron/CEF die soon, and people get back to building applications that don't consume hundreds of megabytes of RAM to render a hello world. Web technologies are fine, but what we really need is some kind of lightweight browser which allows you to use HTML/CSS/JS, but with far lower memory usage. I found https://ultralig.ht/ which seems to be exactly what I am looking for, but the license is a major turn off…

What I'd really like to see with CEF et al, is JS being dropped, in favor of directly controlling the DOM from the host language. Then we could, for example, write a Rust (or Kotlin, Zig, Haskell, etc) desktop application that simply directly manipulated the DOM, and had it rendered by a HTML+CSS layout engine. Folks could then write a React-like framework for that language (to help render & re-render the DOM in an e…

> directly controlling the DOM from the host language

I think your path to this future is getting a DOM API specced for WASM (as in, not going via a Javascript bridge). WASI might help.

If and when that happens, then you can repurpose that same API without necessarily needing to compile to WASM. The biggest hurdle is that the current DOM API is a Javascript API, not an API describing e.g. memory layouts.

Re: Ode to the M1

#259
I have an m1 mbp and a asus zephyrus g14 of similar vintage, and I have to say day-to-day, stock, the m1 mbp can get better battery life idle, and is less prone to getting into a state in which its fans need to activate. I'd say they can both get over 10 hours of battery life easily which gets into the point where who cares anymore, but it's easier to get the asus to drop to only 5.

But, speed-wise, I haven't noticed a difference. They're both snappy. In fact the asus would be faster for all practical purposes of mine because: * macos has comically bad UX, it wishes for me to use the mouse far more often than it should, and fights me on keybindings. * programs like blender and stable diffusion just aren't optimized for apple. they're optimized for nvidia. i can't say if the m1 gpu is actually good, because nobody seems to care to support it, which brings me to my final point:

it doesn't much matter if m1 is or isn't good as long as it's exclusive to apple's walled garden. if I can't put it in my PC, it may as well not exist, both from a user point of view but also support of developers. The best outcome of the m-line of CPUs would be for hardware competitors to make another great ARM chip, that I can actually use elsewhere.

Re: Ode to the M1

#260

Earlier quoted context omitted.

The RAM usage is from JavaScript heaps, not from multiple copies of the binaries. (There basically aren't memory metrics that can show the effect of binary sizes on macOS. It'd show up as disk reads/page ins/general slowdowns instead, but SSDs are so fast that it's very hard to see one.)

Multiple copies of the same binary can't share a memory page for the code. They do take more RAM.

And there are no metrics you can look at to see the impact of this, so it's not contributing to any of the metrics you are seeing, as I said.
Post reply on HN