Asus ProArt P16 laptop review - AMD Zen 5 meets RTX 4070 laptop and 4K OLED

[John Doe]

runtimes better than an arm-based M3 MacBook Pro 16" on a smaller battery

What? M3 Pro mops the floor with AMD runtime, even M3 Max (which is much more performant) runs longer.
AMD shills have no boundaries.

I'm referring to the battery runtimes for this comparison, where they test three parameters. The P16 got 21-hours for video playback and 2-hours for the heavy loads test, which is longer than the M3 Max got in this test comparison. While I'm not an 'AMD shill', as I admit that's cherry-picking data as Apple has longer Wi-Fi runtimes and it's a gap that can't easily be closed with future BIOS/Driver updates, my point was to show that this laptop can definitely be used on battery as a response to the earlier comment.

[Goberman]

I'm referring to the battery runtimes for this comparison, where they test three parameters. The P16 got 21-hours for video playback and 2-hours for the heavy loads test, which is longer than the M3 Max

Do you realize M3 Max is like 1.5x more powerful chip and even it stomped AMD in Wifi battery runtime. Compare this AMD to something of it's level - M3 Pro chip - and AMD gets completely destroyed.

[Goberman]

Also M3 Max outran this AMD in Wifi battery life test by x2 - and Wifi test is actually a better one because it shows a real workload, while video test compares the efficiency of onboard hardware codes, not CPUs.

If you don't know how to add M3 Pro to the table, here's numbers 1481 1204 83.
AMD can't reach those runtimes.

[ArsLoginName]

I'm referring to the battery runtimes for this comparison, where they test three parameters. The P16 got 21-hours for video playback and 2-hours for the heavy loads test, which is longer than the M3 Max

Do you realize M3 Max is like 1.5x more powerful chip and even it stomped AMD in Wifi battery runtime. Compare this AMD to something of it's level - M3 Pro chip - and AMD gets completely destroyed.

You and John Doe are both correct. For this model, the Macbooks beat this 370HX in wi-fi runtime but in video playback, the 370HX wins. I've posted on the wi-fi runtime elsewhere (reddit and here).

However, you need to understand this very poor showing in wi-fi run time is strictly due to the MediaTek Wi-Fi 7 chipset.

Compare
HP 845 G10 - 7840U - 779 min wi-fi time from a 51 Whr battery = 15.27 min/Whr (will get similar 14.5+ min wi-fi/Whr battery from Lenovo T14s G4 with same 7840U and similar 30 W/25 W power settings. Both of these laptops do not use a MediaTek wi-fi chipset).

Asus Zenbook S16 = 640 min wi-fi from a 53% larger 78 Whr battery = 8.205 min/Whr and the difference can't be due to the OLED screen in the Zenbook S16 because the video playback runtime increases by 5% more than the battery capacity increase.

ProArt P16 = 500 min wi-fi from 90 Whr battery. Even worse. But again. Over 1250 min in video playback test which is comparable to the Zenbook S 16's 1204 minutes. So slightly more power hungry OLED in this ProArt due to more pixels.

Your famous and illustrious M3 Pro based Macbook Pro 14 is 1213 min of video playback and  979 wi-fi minutes from a 72.6 Whr battery. Similar video playback times but only 13.48 minutes per Whr battery = worse because it is fabbed on TSMC 3 nm and not TSMC 4 nm. The M3 Pro 16 = 1204 wi-fi minutes from 99.6 Whr battery = 12.09 minutes per Whr. Even worse efficiency than the OLD AMD platform.

On the AMD systems, one should be able to sway out/Get rid of the MediaTek wi-fi and see how the systems compare for wi-fi. You can't do that on Apple.

[Goberman]

ArsLoginName, what a pile of bad data you got here. Different laptops, with different performance, different screens - and you really think you can just divide hour by Wh and get anything relevant? Where did you get this idea?

So 7840U, that is slower than base M2 chip, "won" in your "efficiency calculations" when compared to big kids? Why not compare it to base M2 Air, that does 884 minutes Wifi on 52.6Wh battery and thus wins? You HAVE TO consider only chips of a similar performance to compare efficiency. It's like weight categories for boxers.

strictly due to the MediaTek Wi-Fi 7 chipset

Top baseless claim-2024.

7840U loses to M2, same performance category give or take.
Asus ProArt P16 loses to M3Pro, same performance category give or take.
That's it.

[robert123456]

if we are not comparing cpu's inside their performance categories anymore, my Radxa pi zero 2w with aliexpress screen is the best laptop, it can run 48+ hrs on 100w power bank with a glorious solid 120 points of geekbench multicore brilliance.

so yeah, better stick to performance brackets when comparing or Radxa will smack yo a**.

[ArsLoginName]

Goberman... all data from notebookcheck so all had same video playback, wi-fi tests and screen brightness values. So there is no bad data as you suggest.

I brought up 2 major points. 1 - The wi-fi times for the 370HX platforms are not in accordance with all the other performance and efficiency improvements. The wi-fi run times regressed for this new 370HX platform. 2 - MacBooks on a more advanced processing node, do not have a higher wi-fi efficiency than Zen 4 based 7840U based business laptops.

For the 370HX platforms (both ProArt and Zenbook S16), the local video playback time increased in excess of battery capacity increase for these notebooks. This test does not use the wi-fi chipset. This means the OLED displays can't draw that much more power than the displays in the 7840's I was referencing else local video playback time would have decreased relative to the battery capacity increase. Thus AMD had to increase the efficiency of video playback in excess of any increase in total power consumption by the screen.

It's simple math. Run time = battery capacity/(display power + platform task power including wi-fi). Whr/W = run time in hours.

Macbooks have had impressive video and wi-fi run times because they have used the largest batteries in their notebooks. That is what one buys. A complete notebook regardless of pixel density, display technology (IPS, mini-LED, OLED), wi-fi chipset, processor, and battery. Your M3 Pro Mac Books are on a more advanced node than these 370HX so they *should* have been more efficient per Whr based upon processor/GPU/IO/RAM etc.

My analysis shows that Mac Books do not have the highest wi-fi efficiency per Whr of battery capacity and this new AMD platform has terrible wi-fi efficiency compared to several of their old Zen 4 regardless of whether this higher power ProArt or the same power consumption Zenbook S 16 compared to the Zen 4 based 7840U's. This new AMD system has a MediaTek wi-fi chipset. The HP 845 G10 and Lenovo T14s do not.  On the AMD systems, you can change the wi-fi. On the Mac Book Pros, you can't and they are not the most efficient platform in terms of wi-fi run times.

[Goberman]

I brought up 2 major points. 1 - The wi-fi times for the 370HX platforms are not in accordance with all the other performance and efficiency improvements. The wi-fi run times regressed for this new 370HX platform. 2 - MacBooks on a more advanced processing node, do not have a higher wi-fi efficiency than Zen 4 based 7840U based business laptops.

Ok this is just bullshit at this point. Feel free to stay ignorant.

My analysis shows

Your "analysis" is a comparison of laptop with 8000 geekbench and 15000 geekbench. And you've ignored the fact your 8000 geekbench laptop lost to 9000 geekbench macbook.

Bye. You are hopeless.

[John Doe]

I'm referring to the battery runtimes for this comparison, where they test three parameters. The P16 got 21-hours for video playback and 2-hours for the heavy loads test, which is longer than the M3 Max

Do you realize M3 Max is like 1.5x more powerful chip and even it stomped AMD in Wifi battery runtime. Compare this AMD to something of it's level - M3 Pro chip - and AMD gets completely destroyed.

Last comment I'll make on this as otherwise I will sound like an AMD shill but performance and efficiency are at opposing sides, that said the Wi-Fi test is not running the M3 Pro/Max or AI 9 370HX at full load, that's what the loads test is for. What I said still stands. I will explore your "more powerful" claim, so I'll look at the full load performance and endurance below.

Performance metrics also vary significantly between what, who, where and when a device is tested. Even if you use a site (NanoReview in this comment as it shows No. of samples in the dataset) that uses a large pool of benchmark data to establish a more accurate mean, performance STILL varies greatly on how the benchmark tests are designed like Cinebench(x86-bias) vs Geekbench (Linux + MacOS bias...? or rather Windows sucks). I also hear you about only comparing devices in a similar class, but I merely looked at this review. I'll go the extra step and try verifying your claims:

So M3 Pro = AI 9 370HX: 12-Core = 12-Core both are in 16" creator laptops, similar size and retail price. Using multicore benchmarks to ensure full-load, CinebenchR23 shows an average 14799 vs 24407 in favour of the AMD-chip (65% faster is expected as it's bias to x86). How about Geekbench 6 to avoid x86 platform bias? 14579 vs 15545 in favour of the AMD-chip. (So, 7% faster in what Geekbench consider "real-world" load). How long did they last in their respective loads test here? P16 = 116min vs MBP 16(M3 Pro) = 83min. (M3 Pro 28% less endurance on a 11% bigger battery)

So M3 Max vs AI 9 370HX: 16-core vs 12-core, again both are in 16" creator laptops, similar size but here Apple is almost twice the retail. CinebenchR23 'nT' averages 23507 vs 24407 in favour of AMD by 4% (again not Apples to Apples and rather Apples to Windows as x86-bias). Geekbench 6 nT shows 20930 vs 15545 in favour of Apple by 35%, this is more realistic. Now what are the runtimes where you actually use this 1.35x 'more powerful' advantage? P16 (90WHr) = 116min vs MBP 16 (M3 Max, 99.6Whr) = 71min.

So going back to your 1.5x more powerful claim.. Yes, the M3 Max is 1.35x more powerful but it's also going to last 0.61x the duration of the AI 9 370HX in the P16, even with an 11% battery size advantage. You correctly guessed the M3 Pro has the same performance as the AI 9 370HX but lasts 0.71x the duration under heavy load scenarios. Does that mean the AMD AI 9 370HX is more efficient than any of the Apple M3 variants as a whole?
No, but under heavy loads or just video playback... it's closer than you think.

[tyniyun]

CinebenchR23

I've stopped reading after I saw you referring to R23 when talking about Apple ARM chips. How many times in this forum people had to tell you that Cinebench only added Apple chips support/optimizations in Cinebench 2024. It It is literally second paragraph on their home page.

Cinebench 2024 is designed to accommodate a broad range of hardware configurations - while it seamlessly supports x86/64 architecture (Intel/AMD) on Windows and macOS, it also extends its reach to Apple Silicon on macOS and Arm64 CPUs on Windows

How about Geekbench 6 to avoid x86 platform bias? 14579 vs 15545 in favour of the AMD-chip.

And this is a lie, if you add M3 Pro in this review you will see 15480 macbook and 15367 this AMD laptop. Plus you've ignored a whopping 3100 single-core of macbook, because it's VERY inconvenient for AMD.

in their respective loads test here?

In their battery life test M3 Pro DESTROYED your AMD. And as for "Load" test, do you even remember at this point that Windows laptops switch to low power mode on battery and macs do not? AMD is DESTROYED in every normal productivity test, video and web browsing, and managed to come ahead in "Load" test just because it cut down it's power on battery for unknown amount.

AMD shills on this forum are funny. Complete brain rot.

[lol]

ROFL !!!!!!
I really hope these "AMD totally beats Apple" peeps are on AMD salary because otherwise it's mental, one just keeps "beating Apple" with hallucinated data, the other one compares budget CPU to tops ignoring performance bracket, TDPs and even screen sizes.

Also some info for folx so they don't ashame themselves more --- "Load" test is performed at max brightness, so whatever display is brighter usually uses more power, e.g. macbook uses 5 watts more than this ASUS even just idling at max brightness --- just because of display. And with display off macbook idles at 3 watts less than this Asus. It's all there in reviews if you are actually LITERATE AND READ+UNDERSTAND THEM.

[Toortle]

The same iFan(atic) with multiple names is always polluting every thread around in NikoB's fashion.

[Downhill]

Wow, this quickly went donwhill...
One word. Uneducated...
There is a graph in review. "Power consumption Cyberpunk / stress test"...
One of data points says, "Idle 150 nits", 7.36-16.9W...
Now go to Macbook Pro 16 M3 Pro review, same graph, Idle 150 nits, 7.2-8.44W...
Now look at graphs "Power consumption with an external monitor"...
Asus, idle, external monitor, 7.74-13.8W, avg 10.2W...
Macbook, Idle, external monitor, 5.58-6.35W...
What are you discussing if AMD idles at 4W+ more with screen off on AC power. What are you comparing exactly? Screens? OLED with 350 nits vs MiniLED with 550 nits? Or you are comparing which producer made Windows slow laptop down more on battery? And then extrapolating all this to chips?..
Uneducated...

The same iFan(atic) with multiple names is always polluting every thread around in NikoB's fashion.

Paranoia...

[John Doe]

CinebenchR23

I've stopped reading after I saw you referring to R23 when talking about Apple ARM chips. How many times in this forum people had to tell you that Cinebench only added Apple chips support/optimizations in Cinebench 2024. It It is literally second paragraph on their home page.

Cinebench 2024 is designed to accommodate a broad range of hardware configurations - while it seamlessly supports x86/64 architecture (Intel/AMD) on Windows and macOS, it also extends its reach to Apple Silicon on macOS and Arm64 CPUs on Windows

How about Geekbench 6 to avoid x86 platform bias? 14579 vs 15545 in favour of the AMD-chip.

And this is a lie, if you add M3 Pro in this review you will see 15480 macbook and 15367 this AMD laptop. Plus you've ignored a whopping 3100 single-core of macbook, because it's VERY inconvenient for AMD.

in their respective loads test here?

In their battery life test M3 Pro DESTROYED your AMD. And as for "Load" test, do you even remember at this point that Windows laptops switch to low power mode on battery and macs do not? AMD is DESTROYED in every normal productivity test, video and web browsing, and managed to come ahead in "Load" test just because it cut down it's power on battery for unknown amount.

AMD shills on this forum are funny. Complete brain rot.

To avoid being called an AMD shill again, I can clarify if needed. (I've only ever used Intel FYI)
1. Why use Cinebench R23? I use NanoReview which doesn't show CB2024 yet so finding 100+ benchmarks or reviews and averaging them would be a hassle, therefore I outright said, "test is like comparing apples to windows" if you didn't catch that. I included it to prove the fact that benchmarks are not all equal which mainly alluded to an earlier statement I made. In stats you never just take one sample and say that's representative of the whole group, that's why I didn't just use the benchmark numbers in this review and call it a day. I also didn't use 1T (Single Thread) benchmarks as that doesn't fully saturate/load the CPU, if you're that pedantic about single core, yes M3 Max is a whopping 5% faster than the AI 9 370HX according to a large pool of Geekbench 6 tests on those respective chips (3141 vs 2983).

2. If you're not sure about the performance cut on battery, it's the difference of 10W to the CPU (so from 80W PL1 to 70W PL1) which is a 7% reduction in performance according to the NotebookCheck review. So, if we go by that logic, Apple win's the reward hitting 0% quicker under full load cause Apple doesn't cut power to the CPU and ASUS cut's 10W and lasts considerably longer under full CPU load. It's not a question about efficiency it's a question of power draw vs battery capacity.

I hope this helped.

[ArsLoginName]

Ok this is just bullshit at this point. Feel free to stay ignorant.

Bye. You are hopeless.

Ignorant? LOL. You can't understand math apparently. Typical Mac fanatic saying longest battery life, etc but when shown the math, can't understand basic math and English. LOL. Then resort to name-calling and do nothing to present anything to the contrary but resort to moving the goalposts and bring up the most waste of a benchmark there is, Geekbench. Like another post brought up over their very low power device with their GB score, one can only compare items in similar power envelopes else we should all be using our phones for all computations based upon their GB6 scores of 6700-7200 for SD8G3 and A17 Pro at 10-ish W. M3 Pro is only 15k at 27 W. Are you saying a 6 big 6 little core M3 Pro MacBook is only 2x as good as a 2 big core 4 efficiency core iPhone while consuming almost 3x the power? Quick. You better tell Apple you know more than they do and they don't need the big, expensive M3 Pro.

Being informed is looking at all data and analyzing. You've said and shown nothing to disprove any of the information regarding wi-fi efficiency (minutes of use per Whr of battery) of Mac Books and this Zen 5 (370HX at 33 W/28 W Zenbook S16) compared to several Zen 4 (7840U 30 W/25 W) of a similar power envelope.

As for ignorant (i.e., lacking knowledge or awareness in general), you have to be closer to the definition of it to not even know about tthe information presented here. If you were aware, you'd have realized MacBooks only get the wi-fi and video playback times they do because of their very large battery capacities and always on leading edge node. But they get similar minutes/Whr of battery as several x86 platforms when put under the same testing conditions.