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

111–120 of 419 posts

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

#111

Post by GeekyBear »

pohl wrote: post #20
jmull wrote: post #15

It seems like the M3 favors efficiency and thermals over raw power. This makes sense for the devices they've put it in. I wonder if we're going to see different processors (M3X or M4) for the next release of the pro desktops, that favor power over thermals and efficiency? Maybe there will be a kind of tick-tock, with odd numbered processors favoring efficiency and even numbered processors for power? M3, M5, M7 for th…

Doesn't prioritizing efficiency and thermals make sense for any device? Even a desktop has moments when it's doing almost nothing but has tasks that can adequately be handled by an E core where powering up a P core is overkill. Doesn't being efficient mean that more of the thermal budget can be held in reserve for when it's actually needed? Don't desktop users desire for their powerful machine to also be quiet?

> Doesn't prioritizing efficiency and thermals make sense for any device?

Intel did just ship 14th Gen Core i9 chips with a 400 watt power draw under load and very little in the way of a performance gain.

> In some areas, the extra clock speeds available on the Core i9-14900K show some benefit, but generally speaking, it won't make much difference in most areas.

The most significant win performance for the Core i9-14900K came in CineBench R23 MT... the Core i9-14900K sits 6% ahead in this benchmark

https://www.anandtech.com/show/21084/intel-core-i9-14900k-co...

Having your best benchmark result only see a 6% gain in return for a more than 400 watt power draw makes it pretty clear that just as with the Pentium IV of old, Intel isn't on a sustainable path.

It also makes the M3's 20% performance gain along side efficiency gains look pretty darn good in comparison.

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

#112

Post by tootyskooty »

bloopernova wrote: post #98

I want a better system load metric now that we've got heterogenous cores in CPUs. The Pixel 8 Pro has 3 types, efficiency, performance, and one single ultra performance core. If your efficiency cores are always close to max usage, but your performance cores are idle, is your system being heavily or barely used? I understand that system load isn't really a useful metric between systems, but it's useful to compare on a…

Sounds like power usage might fit the bill.

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

#113

Post by blktiger »

scrlk wrote: post #9
cubefox wrote: post #3

This seems to be a lot of effort to rationalize the surprisingly small performance increase from M2 to M3. Initially the assumption was that M2 to M3 would be a bigger step than M1 to M2, not a smaller one. Perhaps TSMC 3nm is showing the limits of scaling?

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…

Weren't all of the new M3 Macs announced the same price as they previously were or lower? Same with the recently announced iPhones? Or am I mis-remembering? Seems like prices not increasing given all the recent inflation are actually a price decrease pretty much across the board not an increase on the average selling price?

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

#114

Post by nomel »

masklinn wrote: post #103
_ph_ wrote: post #95

Earlier quoted context omitted.

They are not artificially limiting the number of external screens. To support more screens, they would have to reserve die area for an additional controller. That could mean a gpu less or something. This is about engineering tradeoffs.

Considering the M3 has 55% more transistors than the M1, I'm sure they could have found some room for a third display controller, or a more capable one (one with MST support for instance), or just a more flexible one. They're selling a $1600 machine which can't drive two external displays, it's sad. Intel's HD Graphics have been doing better since 2012 and they're trash.

> considering the M3 has 55% more transistors than the M1

And that's why everyone loves node shrinks. You can stuff more transistors for the same cost.

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

#115

Post by masklinn »

user_7832 wrote: post #100
masklinn wrote: post #87

Earlier quoted context omitted.

More importantly VFR is not useful for just high refresh rates. I would argue that for most people on a phone it's the least useful part of the thing. Instead VFR allows saving battery by slowing down the refresh rates, down to as low as 1FPS for static or high latency content. That's part of why always on display is only available on the 15 Pro: at 60Hz it's a battery killer. > Sort of like going from a slowly accel…

> So... functionally useless. (Disclaimer, I haven't driven a sports car but I have driven cars between "nice" and "struggles to reach the speed limit") While it might appear to be very similar, having a sports car that you're capable of driving well/defensively can literally be a life saver. There was this video from a few years back on reddit of this guy in a 911. Vehicle in front suddenly stops/crashes, he needs t…

> While it might appear to be very similar, having a sports car can literally be a life saver.

Or it can be a life ender by planting you into the side of the road, something mustangs and BMWs are well known for, amongst others.

> that you're capable of driving well/defensively

If you need to be a skilled pilot to face the situation, it's less the car and more the pilot which does the life saving. And even skilled pilots who know their cars well can fuck it up. Rowan Atkinson famously spun his F1 into a tree, after 14 years of ownership, having been racing for 2 if not 3 decades.

> There was this video from a few years back on reddit of this guy in a 911. Vehicle in front suddenly stops/crashes, he needs to rapidly change lanes at highway speeds.

Something which ends up in a collision as often as not.

> The little 911 unsurprisingly handled it excellently. A slow/heavy car would've very likely spun out and crashed.

Or not. Or maybe with a less capable machine they'd have driven more prudently and would have kept more space.

> Having more power if you don't need it doesn't hurt (if you're trained well).

So having more power can literally hurt. And routinely does.

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

#116

Post by cduzz »

bloopernova wrote: post #98

I want a better system load metric now that we've got heterogenous cores in CPUs. The Pixel 8 Pro has 3 types, efficiency, performance, and one single ultra performance core. If your efficiency cores are always close to max usage, but your performance cores are idle, is your system being heavily or barely used? I understand that system load isn't really a useful metric between systems, but it's useful to compare on a…

Isn't this just a subset of the existing capacity monitoring problem where I want to know how loaded a multi-core system is under the following scenario:

Some critical path, single-threaded task is at 100% of capacity of a core; the system has 10 cores -- is my system at 10% utilization or 100% utilization?

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

#117

Post by stetrain »

speedgoose wrote: post #70
stetrain wrote: post #46

Earlier quoted context omitted.

14" MBP with the highest M2 Max configuration, 64GB RAM, 1TB SSD was $3899. A new model with the same configuration is also $3899.

Wow, that’s overpriced for such low specs.

For 16 core CPU, a dedicated GPU equivalent, 400GB/s LPDDR5 memory, 3024x1964 Mini-LED, and 18 hours of battery?

The closest I can configure a Dell XPS 15 is $3099, and that's a 14 core CPU and likely lower memory bandwidth and lower performance SSD. They claim 18 hours of battery but only with the base screen, the upgraded screen is presumably less.

And from personal experience using an XPS 15 is a significantly worse experience in stability, heat, fan noise, and real-world battery life.

And here are benchmarks for the XPS vs the M3 Max:

https://browser.geekbench.com/v6/cpu/3367184

https://browser.geekbench.com/v6/cpu/3372431

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

#118

Post by jwells89 »

turtlebits wrote: post #93
user_7832 wrote: post #81

Earlier quoted context omitted.

High refresh rates aren't just for games, they make the entire interface smoother. Things like frame rate or latency are quality of life things that you don't need , but once you upgrade, the difference becomes clear. Sort of like going from a slowly accelerating station wagon to a Porsche/sports car even when (capped by the speed limit) if you've experienced that. Hopefully someone else has a better analogy.

IME, the difference is negligible. I went from a iPhone 12 to a Galaxy S21 Ultra and yes, there is a difference, but it's not something I notice, even when occasionally using my wife's iPhone 12. The bigger difference I notice is in loading times for apps.

For me the difference is visible, but the added smoothness is more of a cherry on top than anything. I switch between 60hz, 120hz, and 240hz panels every day and the 60hz panels don’t bother me a bit. On desktop and laptop displays I’d take integer-scaling-friendly pixel density over high refresh rates any day.

Variable refresh rates should be standard across the board however, not restricted to high refresh panels. There’s no more reason for a static screen to be redrawing at 60hz than there is for it to be redrawing at 120hz or 240hz.

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

#119

Post by wubrr »

varispeed wrote: post #85
user_7832 wrote: post #72

Earlier quoted context omitted.

While some engineering compromises I can understand, most appear to simply be anti-consumer/cash-grabs. Need more than 8gb ram? Hope you've got sufficient money. Want a high refresh rate screen present in cheap android phones? Sorry not for our base models.

Even the lowest tier models probably beat any PC laptop when it comes to usability and comfort though.

Strong disagree.

UI is very buggy and inconsistent.

CLI tools/shell is attempting to follow more-or-less standardish unix/linux setup but fails pretty hard, with many of the same commands existing but behaving just differently enough to be annoying.

Uses it's own set of shortcuts, different from what's standard on windows/linux and most other operating systems.

Not sure what people mean when they say completely vague and ambiguous things like 'usability' and 'comfort' - are you referring to the fact that it has rounded buttons and corners and pretty ads?

It basically falls in between a standard windows and highly modified ubuntu setup, but does far worse at both. The UI, apps, etc is far more buggy and unintuitive than windows, and the os is far less 'unix' and far less customizable than ubuntu.

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

#120

Post by dnissley »

rahoulb wrote: post #22
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.

I bought my M1 MacBook Pro not long after they were released. It's approaching 3 years old which is a pretty standard age for people to start looking at a replacement - I guess that's why they're comparing it to the M1. Slight tangent - even though the MBP has been struggling to handle my dev work (partly because the codebase I'm working on has grown significantly in the last two years) I won't be upgrading. Work bou…

Oldest MBPs with an M1 are from late 2021, so only 2 years old.
Post reply on HN