Asus Zenbook S 16 laptop review - The first Copilot+ laptop with AMD Zen 5 inside a 1.3-cm-thick case

[IT amateur]

How can they call it ryzen NINE if it gets completely raped by passively cooled M4 tablet in geekbench.

+1 Have to agree even being an anti-apple / apple hater.

Can't believe how we got here. Apple is light years ahead of BOTH arm and x86. Mentioning x86 vs arm wars no longer matters anymore and is irrelevant.

What have all the other companies (and the entire industry) been doing the last 4 years? Sleeping?

When the M4 MBP's release, I think we can finally stop and put to rest the whole arm vs x86 battle. It's going to completely destroy everything.

Conclusion: Just go Apple and if you can't everything else is bad so get whatever is cheap < $699. Spending > $1400 in the pc ecosystem no longer makes sense anymore.

For the price of these strix laptops you can get a well specc'd MBP + a dedicated gaming console and get a far better experience as well.

[Jedy]

How could you miss the touchscreen? I have the laptop in front of me here and it is touchscreen, along with all the media saying it's touchscreen...

[Jedy]

Also, the battery tests are fair, been using the laptop for two days, I get 10+ hours on light use, lol to the guy who says divide that by 2. Only things I don't love about it are mushy keyboard, trackpad could be clickier, and honestly, the bottom of the laptop does get quite hot with medium use, like lots of chrome tabs and videos running. Besides that, everything else good. Surprised display only got 85% while something like a macbook pro gets a 94%

[Andreas Osthoff]

Our particular review unit was not equipped with a touchscreen.

[rk]

When the M4 MBP's release, I think we can finally stop and put to rest the whole arm vs x86 battle. It's going to completely destroy everything.

It will release with 8/256 base spec and for 32/1T you will have to shell out $3000+. 8/256 will be way cheaper, it will be the fastest media consumption / typewriter device out there but for that you don't need so much "performance".

[haha]

It will release with 8/256 base spec and for 32/1T you will have to shell out $3000+. 8/256 will be way cheaper, it will be the fastest media consumption / typewriter device out there but for that you don't need so much "performance".

Then you'd better start working and saving for a proper new macbook instead of yapping on the internets.

Typical windows->mac immigrant:
"why mac is too expensive"
buys his first crappy windows laptop
"why mac is too expensive"
buys his second crappy windows laptop
"why mac is too expensive"
buys his third crappy windows laptop
tired of windows bullshit and finally buys mac
"ah now i get why"

[jdrch]

The lack of TB4/5 is a dealbreaker for me.

[NikoB]

The lack of TB4/5 is a dealbreaker for me.

In fact, TB4 vs USB40 V1 have one hidden advantage. Most monitor models with 144Hz+ and HDR do not work at these frequencies without DSC (lossy compression), because HDMI 2.1 on their input ports is cut exactly 2 times to 24Gbps, which the vile manufacturers keep silent about in the specifications.

As a result, today you can connect such a monitor only if the monitor has a TB4 input, as far as we could find out from the reviews of competent buyers. It is in this mode that you get a 40Gbps band, which is enough to output 4k@165Hz + HDR (30 bit in 4:4:4 in lossless mode).

Therefore, the lack of a TB4 port is a death sentence for any desktop and laptop, if you plan to buy current monitor models 4k@144Hz+.

But you still won't be able to play effectively on them for a simple reason - most often the TB4 port is switched to igpu, not dgpu, i.e. the frame buffer is rendered in the system memory, which dramatically reduces fps in games and 3D software.

That's why DP2.0+ with UHBR10 at least is so important, which officially appeared in Zen4 Phoenix igpu, but it is still not there, to the shame of NVidia/AMD, in their discrete video chips. AMD has DP2.0+/UHBR20 only in professional video cards since last year.

As soon as you have a DP2.0+/UHBR20 interface, the topic of extremely clear (almost like a laser printer) fonts on screens up to 32" will be closed forever, because 8k monitors will immediately give 250ppi+ even on 32".

But even UHBR20 is already morally obsolete because it cannot provide smooth rendering in dynamics even in 2D, because 60Hz+ is not available there. Formally, TB5 is cooler than DP2.0+/UHBR20, but the nuance is that 120Gbps there is made for a different reason - because there is support for 2xDP2.0+, even in 120Gbps mode, one monitor will only have 60Gbps (as now on DP1.4b in TB4 there are shameful 20Gbps in 2-monitor mode, although DP1.4b, the full version, requires 32Gbps), which is 20Gbps less than the full version of DP2.0+/UHBR20 with 80Gbps.

Thus, today there is simply nothing to connect potential 8k@120Hz models to - even TB5 cannot deliver 160Gbps.

Despite the fact that DP2.0/UHBR20 was released in another era, in 2019(!), i.e. exactly 5 years ago, to the shame of the IT industry, it has not yet been put into mass operation. And at the same time, it was already morally obsolete at the start for the reason stated above...

That is why the days of copper wires are numbered - at distances of up to 2m and above, the only solution can be optics with a bandwidth of 160Gbps+.

[NikoB]

And that's why I've been writing for a long time about the need to sharply increase the bandwidth of the x86 memory system - to service such heavy configurations with 8k monitors and other equipment, the RAM bandwidth has long been above 200 GB / s since 2019, but only now, to the shame of AMD / Intel, we saw boring and slow 100 GB / s with LPDDR5 7500 in Zen5 Strix. And we need 2-3 times more! That's why Zen5 Strix does NOT support DP2.0 / UHBR20 - for this, AMD will release Zen5 Strix Halo with a 256-bit memory controller, although the question is whether the controller efficiency will be around 80% ... Apple, to its shame, failed to ensure the memory controller efficiency even in the region of 40%, with top M3 / M4 with theoretical 400 GB / s at a 512-bit bus.

[NikoB]

And of course igpu suffers greatly compared to dgpu from very slow RAM even in Zen5. You need much more bandwidth or dedicated HBM3/GDDR6 with 256+ bit bus on the SoC crystal for VRAM.

[batury]

Also, the battery tests are fair, been using the laptop for two days, I get 10+ hours on light use, lol to the guy who says divide that by 2. Only things I don't love about it are mushy keyboard, trackpad could be clickier, and honestly, the bottom of the laptop does get quite hot with medium use, like lots of chrome tabs and videos running. Besides that, everything else good. Surprised display only got 85% while something like a macbook pro gets a 94%

I think by "divide by two" he meant that intel traditionally reduces performance on battery, because according to review on AC this laptop uses avg. 8.5-10W in idle, so on AC power it mathematically can't even idle for 10hrs on 78Wh battery, and yet it gets 20hrs of video playback on battery, which is 3.9W consumption.

Macbooks' displays were rated on "v7" rating and this one is "v8", I think PWM now weights display score down more.

Please note: We have recently updated our rating system and the results of version 8 are not comparable with the results of version 7. More information is available

[batury]

errr, not "intel" but "x86"

[Real NikoB (old name bl)]

, and yet it gets 20hrs of video playback on battery, which is 3.9W consumption.

There is in reality a pulse load on the SoC, which is extremely dependent on the complexity of the information being processed.

In general, when the SoC consumption is less than 10W, other system components become significant factors - the screen, the motherboard with other devices.

The problem with x86, to their complete shame, is that they cannot even decode 4k@60fps on YouTube in VP9 with a total consumption of the SoC and SoC power supply of less than 5W. This is simply ridiculous against the background of smartphone consumption under the same load...

[Steve1]

In general, when the SoC consumption is less than 10W, other system components become significant factors - the screen, the motherboard with other devices.

NBC measures consumption on wall socket.

[damndarn]

It looks like there's a driver issue with the wifi card.
ASUS just released an update for the driver.
Can you update the battery runtime after the update?
The WIFI runtime seems to be too low.