Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

M3 Macs: there's more to performance than counting cores

eclecticlight.co

Post reply on HN
Print view View original

311–320 of 419 posts

Re: M3 Macs: there's more to performance than counting cores

#311

Post by scarface_74 »

lilyball wrote: post #289
musha68k wrote: post #282

As a user of many Apple Silicon generations in different / mostly top configurations by now I’m obviously a big fan. One thing that I have “observed” though (with no data to back it up): it seems to me as if under heavy load / resource over-provisioning the Intel systems from before seemed to recover more gracefully? I wonder if it’s just me, the particular OS version I had been using at the time or some other thing…

Right now I have an intel laptop for work and an apple silicon laptop for personal use. The workloads I do on these machines is different so comparisons are a bit hard to do, but I've seen the intel laptop exhibit poor behavior under load that I've never seen from the apple silicon machine.

By “load” you could easily mean “running Microsoft Teams”.

Re: M3 Macs: there's more to performance than counting cores

#312

Post by spacedcowboy »

rfoo wrote: post #268
spacedcowboy wrote: post #204

Earlier quoted context omitted.

Because the newer tools changed the licensing terms, and the corporate lawyers won’t let anything under GPL3 anywhere near anything if they can help it. GPL2 was viral, but the terms were easier to stomach. Apple is allergic to GPLv3 code because there is a clause in the license requiring you provide a way to run modified version of the software which would require Apple to let users self sign executables. This is a…

> Apple is allergic to GPLv3 code because there is a clause in the license requiring you provide a way to run modified version of the software which would require Apple to let users self sign executables. Wow, I don't know this before. Good job FSF! This, should, be, a, basic, right.

You can of course sign your own binaries. You can’t alter them sign a system binary. I’m good with that.

Re: M3 Macs: there's more to performance than counting cores

#313

Post by shepherdjerred »

baz00 wrote: post #217
shepherdjerred wrote: post #126

Earlier quoted context omitted.

From what I know, this isn't some artificial limitation that Apple imposes. Adding support has a cost that most customers don't need to pay. Regardless, most people aren't going to be plugging in three monitors into their laptops, so most people aren't going to care about this.

Intel support 4 displays on a bottom end i5-1235U. Even my crappy little Lenovo M600 with an N-series Celeron supports 3 displays. Two 4k ones fine. I didn't have a third to test it with. It's either an architectural fuck up or intentional segmentation.

> It's either an architectural fuck up or intentional segmentation.

Do you think there is any chance that Apple would have had to make some engineering compromise (cost, performance, efficiency) for a feature that very few people would use?

Re: M3 Macs: there's more to performance than counting cores

#314

Post by user_7832 »

lukas099 wrote: post #159
user_7832 wrote: post #84

Earlier quoted context omitted.

> Lots of compromises are made to get the Air to the size, weight, and most importantly, price point (ie sub-$1k) that it is. > IMO the $999 one (or $899 with education discount) is literally the best price:perf ratio of any computer available today or at any time in my recent memory. It’s astounding how good it is for that price. It is possible for your comment to be correct while the poster above is also correct. T…

> knowingly gimp the hardware to upsell the more expensive versions. I'm skeptical. A ton of money is dumped into optimizing everything and then they just throw it down the drain? That would drive more people to the competition than to higher-priced Macs.

I meant that bit in general and not specifically for the mac, but for eg they sold iPads with 32gb storage as recently as 2020 iirc. Even now the base level macs have much less ram and storage than similarly priced windows counterparts. I think part of it is just so they can say "macbooks start at $xx!". There are more examples I can't think of right now.

Re: M3 Macs: there's more to performance than counting cores

#315

Post by buildbot »

throwaway49594 wrote: post #260
GeekyBear wrote: post #179

Earlier quoted context omitted.

Given that Intel and AMD tend to choke the memory bandwidth on anything below their server chips, this seems short sighted. The M3 Pro's "reduced memory bandwidth" is still double the memory bandwidth you see on Intel and AMD's HEDT chips. Step up to the M3 Max and you're looking at five times the memory bandwidth of Intel and AMD's chips.

You're comparing strictly CPU memory bandwidth to CPU + GPU. If you add CPU + GPU bandwidth for a PC you'll get similar numbers.

Your PC can't use the GPU memory bandwidth for the CPU whatsoever. So why would you add the bandwidth?

Re: M3 Macs: there's more to performance than counting cores

#316

Post by creativenolo »

fsociety wrote: post #41
whywhywhywhy wrote: post #8

Feels like we’ve gone from “it’s just better, it’s obviously better you can see with your eyes and feel it as you use it” when M1/2 dropped. To trying to justify why the M3 is 20% faster than the M1 and the M2 was also 20% faster than the M1 and weirdly Apple is only comparing it to their older processor. Like people, maybe it’s just an underwhelming update… no need to pretend it’s not.

Why would you market to people who just bought an M2 Pro this year? Those who are going to upgrade, will anyways. Seems silly to market to people telling them to buy new multi-thousand dollar laptops every year.

Yeah and in addition, people know Intel to Apple silicon was a massive leap, so why market how much fast it is than the intel version. That play has been played.

The stats compare to m2 are there. They are just not the headline/summary.

Re: M3 Macs: there's more to performance than counting cores

#317

Post by user_7832 »

marmaduke wrote: post #306
musha68k wrote: post #290

Earlier quoted context omitted.

OOM kills are basically the sole reason for me to upgrade iPhones. Nothing more annoying than force-reloaded/lost state in/between apps while on-the-go and under tight time constraints. Also somewhat true: new lenses are amazing but “computational photography” defaults sometimes less compelling than my old iPhone X’s more “honest” output. I’m going to play around with RAW some more if I get the time for it though.

Just curious, I had a 2020 SE daily and never had a OOM, how do you do it?

I still use a 2020SE and experience apps reloading all the time. I cannot keep more than 2 apps reliably in memory at any point. This includes Safari, youtube, reddit (honestly a poorly made app), spotify etc - fairly "common" ones.

Re: M3 Macs: there's more to performance than counting cores

#318

Post by AnthonyMouse »

Nullabillity wrote: post #273
jlokier wrote: post #209

Earlier quoted context omitted.

> Why? Why do I need to use 3rd party package managers, or manual installations from 3rd party sources to install common tools that aren't a decade+ outdated? There actually is a reason. It's not lack of maintenance. It's because Bash 3.2 is the last version licensed as GPLv2. Bash 4.0 and later changed license to GPLv3. Apple decided it's not safe for them to ship any software with GPLv3 with macOS, because of stron…

> So from a certain point of view, Bash 3.2 is the most recent version they can ship. Apple could choose to be GPLv3 compliant tomorrow, if they wanted to.

They also seem to be about the only ones who seem to think this is a problem. Even Microsoft has the GPLv3 bash in WSL.

Re: M3 Macs: there's more to performance than counting cores

#319

Post by forrestthewoods »

pier25 wrote: post #50
scrlk wrote: post #9

Earlier quoted context omitted.

It's a mixture of several problems: * TSMC N3B being a bit of a flop (yield issues, too expensive) * Brain drain from Apple's chip design teams over the last few years * Tim Cook trying to push the average selling price up to keep revenue growth going in the face of sales declines (e.g. hobbling memory bandwidth, reducing the number of performance cores for M3 Pro) I don't expect there to be a M1 style generational l…

The GPU is increasing a lot though. Another point is the CPU improvements are constant. In about two years the base M3 is now on par with the M1 Max.

> The GPU is increasing a lot though.

Is it? Does it matter? Unified RAM is cool. But I’d rather have an Nvidia 4090.

Re: M3 Macs: there's more to performance than counting cores

#320

Post by wubrr »

tcmart14 wrote: post #259
jamespo wrote: post #236

Earlier quoted context omitted.

I don't understand this pretence that the vast majority of the time laptops are plugged into the power anyway.

Perhaps because until recently, battery life has sucked? I used to plug in my laptops all the time. Now with the M1, it only gets plugged in to charge, then I don't plug it in until I charge it again. I have been enjoying sitting on my couch when I want to, to get work done rather than at my desk.

You can sit on the couch and have your laptop plugged in. I also don't keep my lenovo plugged in all the time and don't notice any battery issues. Realistically you're not sitting on a couch, working for 6 hours straight either.
Post reply on HN