Apple MacBook Pro 14 2023 M3 Pro review - Improved runtimes and better performance

[Redaktion]

Apple has released the MacBook Pro 14 with a slimmed-down M3 Pro SoC and compared to its M2 predecessor, it costs about US$100 more. In turn, you get more CPU performance, a brighter display in SDR use and considerably longer runtimes.

https://www.notebookcheck.net/Apple-MacBook-Pro-14-2023-M3-Pro-review-Improved-runtimes-and-better-performance.779538.0.html

[Neenyah]

Horrible power consumption, literally worse than Intel laptops, lol. My X1 Carbon with Intel is about 30% more efficient; currently not idling but watching Twitch, reading here etc. and using just 10.7W total system power with Razer Core X connected to it (but the RX 6800 XT is at 3W right now).

1h 17min of battery life at max load of this MBP is a comedy because more powerful gaming laptops are around the same figure. And also funny is idle average of 13.1, load average 47.4, load maximum 73.6...

vs

...a completely random Zen 4 laptop I took out for comparison where idle average is 6.5 (-6.6), load average 41.9 (-5.5) and load maximum 50.8 (-22.8).

[NikoB]

Is this the way it is now in NB? Hide screen data and draw only final conclusions? Should we take their word for it?

Graphics performance is at the level of the GTX1650 4 years ago, which is not surprising given such low performance of the memory controller (relative to Apple's declarations, of course, regarding the x86 camp with a video chip built into the SoC, they are still 1.5-2 times faster).

The price clearly doesn't match the features. If there were 36GB of RAM and a 4TB SSD, it would be more adequate, of course, without taking into account the flickering screen that is harmful to the eyes...

Otherwise, a boring, grossly overpriced laptop that doesn't represent anything cutting-edge for late 2023...but it's Apple, right?

[LL]

It is unexplainable how with those issues it is still a recommended laptop.

[ArsLoginName]

Andreas... please explain "however, the U and P models from AMD and Intel lost out." when in fact the 8 core 7840U at 24 W (long) power scored an average of 1811 CB R15 points (Lenovo T14s G4)compared to the 1911 CB R15 of the M3 at 27 W (25 W ?) and on 3 nm. 7840U = 95% of the performance with less cores at 89-97% the power consumption and on an older node. There are probably larger manufacturing distributions of +/- 4-5% in performance due to processor/laptop assembly variations than what you are observing.

Please keep things in the proper context and report accurately.

[ArsLoginName]

Further, when you compare the wi-fi run times of the new MacBook to the Lenovo T14s G14 you will find the MacBook gets 16.3167 hrs/72.6 Whr battery = 0.2247 hr of wi-if per Whr of battery. The Lenovo T14s G4 gets 13.2167 hrs (793 min) of wi-fi time from a 57 Whr battery for 0.2319 hr of wi-fi time per Whr of battery. So the AMD system is even more efficient. The HP 845 G10 gets even an higher efficiency of 12.983 hrs (779 min) from an 51 Whr battery for an efficiency of 0.2546 hr of wi-if per Whr of battery. Apple gets long wi-fi times by using very large batteries.

[NikoB]

The overall performance rating of cores, normalized not only by consumption, but by technical process, still turns out to be in favor of Intel:
Intel->AMD->Apple. But thanks to AMD's access to a better process technology than Intel (still worse than Apple, but very close), they lead the absolute performance of the three.

[Bizarro_NikoB]

Mac's heyday was OSX 10.8.5 Mountain Lion and the machines that ran it. Everything since then have been overpriced disposable electronics to include this garbage iteration, but keep drinking the Apple Kool Aid and shilling for a company that glues their batteries in devices and don't believe in right to repair yet still claim they care about their userbase.

[A]

1h 17min of battery life at max load of this MBP is a comedy

You still have to push it to max load first.

...a completely random Zen 4 laptop

A completely random Zen 4 laptop which is 2K/8K vs 3K/15K M3 geekbench. You have to compare your "completely random Zen4 laptop" to base prev gen M2 Macbook Air and see how's power consumption there.

[A]

scored an average of 1811 CB R15 points (Lenovo T14s G4)compared to the 1911 CB R15

Maybe because Cinebench R15 is back from 2013 and not only was optimized for ARM, not even running native on ARM, but M chips never even existed when it was created?

They've added some M chips support only in Cinebench 2024, it's said right there on main page of their website.

[A]

Further, when you compare the wi-fi run times of the new MacBook to the Lenovo T14s G14 you will find the MacBook gets 16.3167 hrs/72.6 Whr battery = 0.2247 hr of wi-if per Whr of battery. The Lenovo T14s G4 gets 13.2167 hrs (793 min) of wi-fi time from a 57 Whr battery for 0.2319 hr of wi-fi time per Whr of battery. So the AMD system is even more efficient.

Can't compare these numbers directly except more/less, displays/browsers take more energy than CPUs in these tests.

[A]

and don't believe in right to repair yet still claim they care about their userbase

They are directly selling you official spare parts online for your laptop and are buying your laptop back.

[Neenyah]

1h 17min of battery life at max load of this MBP is a comedy

You still have to push it to max load first.

My point is that the battery life is a**. There, more power hungry laptop with larger OLED screen (consumes way more battery), 4080 140W (self-explanatory), i9H CPU - it's pretty clear that it's overall more powerful laptop than this MBP and despite just 23.4 Wh larger battery it pulls literally double the runtimes at max load:

"Load (maximum brightness) 2h 24min"

So much about M3's "efficiency" 🙄 I mean everything is efficient when it's not being used...

...a completely random Zen 4 laptop

A completely random Zen 4 laptop which is 2K/8K vs 3K/15K M3 geekbench. You have to compare your "completely random Zen4 laptop" to base prev gen M2 Macbook Air and see how's power consumption there.

Benchmarks are useless (unless all you do for a living is running benchmarks 8-12 hours a day).

[A]

"Load (maximum brightness) 2h 24min"

From this data you can deduce it cuts down CPU/GPU power more than in half, unless you believe there's "Iiiiintel magic" that makes 140W GPU work for 2.5 hours on 100Wh battery under full load. It's a pale shadow of the same laptop on AC. Math kinda tells in full performance mode it'll work for 35-45ish mins on battery.

Macs do not enter any "power saving mode", they are running full speed on battery, so math adds up and is honest. You can toggle Low Power mode manually though if you want it, it about halves power consumption.

Benchmarks are useless (unless all you do for a living is running benchmarks 8-12 hours a day).

Well at least get a competitor of the same level of performance. Or compare it properly, to base M3, which will probably mop the floor with it on power consumption.

[Neenyah]

"Load (maximum brightness) 2h 24min"

From this data you can deduce it cuts down CPU/GPU power more than in half, unless you believe there's "Iiiiintel magic" that makes 140W GPU work for 2.5 hours on 100Wh battery under full load. It's a pale shadow of the same laptop on AC. Math kinda tells in full performance mode it'll work for 35-45ish mins on battery.

Totally not saying anything about Intel, but about Apple's M3 not being that magical in efficiency as (some) reviewers and many people praise it to be.

Macs do not enter any "power saving mode", they are running full speed on battery, so math adds up and is honest. You can toggle Low Power mode manually though if you want it, it about halves power consumption.

With many laptops yes, with some not really. It's actually switchable in the BIOS for many users, default for most laptops is "max performance" on AC and "best battery" for, well, battery. You can switch both to "max performance" in the BIOS if you want (not with all OEMs but with most). For example, here is P1 G6 (default is "balanced").

Benchmarks are useless (unless all you do for a living is running benchmarks 8-12 hours a day).

Well at least get a competitor of the same level of performance. Or compare it properly, to base M3, which will probably mop the floor with it on power consumption.

i9 13905H + RTX 4080 (140W) + 32 GB RAM are comparable in performance to the base M3 with that joke of 8 GB RAM? Ok, lol.