Why is my new M5 Max MacBook Pro getting worse WiFi performance than my M3 MacBook Air?

[Gourlish]

I got a 16 inch M5 Max MacBook Pro this week, and overall it's a very powerful machine. At the same time I'm a bit concerned about its WiFi performance, which seems noticeably worse than on my M3 MacBook Air, at least on my home network. Both are connected to Channel 44/5GHz/80MHz, and my M3 MacBook Air is getting a consistent TX rate of 585-650mbps. By comparison my MacBook Pro seems to be getting 390mbps most of the time with occasional drops to 195mbps, 144mbps or 85mbps. When it comes to downloading large files my M3 Air is able to download them via my home router at around 20-30MB/s while my new MacBook Pro is struggling to download at more than 15MB/s, sometimes dropping to 4-5MB/s, although it occasionally reaches 20-30MB/s. What is the likely problem causing worse network performance on a better machine?


My home router only supports WiFi 5, and I don't know if Apple's new N1 chip is perhaps less capable with older router technology than the WiFi chip in the M3 MacBook Air.

Here's a comparison between them (my M3 Air at the top, and my M5 Max Pro at the bottom):

 

[Bigwaff]

Post the make/model/year/specs of your router.

 

[chown33]

Builtin Wireless Diagnostics.

Use Wireless Diagnostics on your Mac

Use Wireless Diagnostics on your Mac to help resolve Wi-Fi issues, and learn about Wi-Fi best practices.

support.apple.com

 

[Gourlish]

Post the make/model/year/specs of your router.

It's a Plusnet Hub One (Plusnet are a British ISP). I think the unit dates from only 2020 or 2021, but it's basically a rebadged BT home router from the early to mid 2010s as far as hardware goes.


Wireless Diagnostics claims the WiFi is working normally. I doubt there's anything wrong with the N1 chip, but it's probably something to do with my home router not being the best suited for it. I still live at home with my parents and my dad maintains it, but it might be worth seeing if he can set the router to have split 2.4GHz and 5GHz bands, which I've heard does improve the performance of some Apple devices (even though I've had no problems with my M3 MacBook Air) on this router model.

 

[Bigwaff]

The MCS Index of 0 and NSS value of 0 for M5 Max MBP is really bad. This probably means router's WiFi channel selection is heavily congested and the MBP can't establish a reliable connection to the router, which make sense given the behavior you described. Try changing the WiFi channel on the router to one not so heavily congested with noise. The M3 MBA got lucky or was closer to the router.

 

[chown33]

Does your Wireless Diagnostics have a menubar?
If so, look for a Window menu.
Choose Performance from the Window menu.
It should present a window with ongoing graphs for speed, quality, and signal.
What range does speed, quality, and signal show over time?

I don't have any Mac running Tahoe, so can't check what Wireless Diagnostics can do in Tahoe.

In my past experience troubleshooting wireless networks, the individual windows are far more useful for monitoring the ongoing connection than any summary report file.

Example screenshots from my Sierra MB Pro

 

[Gourlish]

Does your Wireless Diagnostics have a menubar?
If so, look for a Window menu.
Choose Performance from the Window menu.
It should present a window with ongoing graphs for speed, quality, and signal.
What range does speed, quality, and signal show over time?

I don't have any Mac running Tahoe, so can't check what Wireless Diagnostics can do in Tahoe.

In my past experience troubleshooting wireless networks, the individual windows are far more useful for monitoring the ongoing connection than any summary report file.

Example screenshots from my Sierra MB Pro

View attachment 2630337

View attachment 2630338

Here's a comparison between my M3 Air (first image) and my M5 Max MacBook Pro:

 

[chown33]

Does anything change in the graphs when you move the M5 closer to the wifi AP (router)?

You might be able to evaluate channel congestion using a Scan. Mine brings up a window with all the APs it can see. It's presented in a table, and clicking on a column header (Channel) sorts by that value. A simple scroll shows how many APs are on that channel. For example, chan 157 on the 5 GHz band is by far the most populated channel for me. Fortunately, my router uses a different channel.

 

[Harddrive]

Do you have the ability to change the router to WPA3 ?

 

[Gourlish]

Does anything change in the graphs when you move the M5 closer to the wifi AP (router)?

You might be able to evaluate channel congestion using a Scan. Mine brings up a window with all the APs it can see. It's presented in a table, and clicking on a column header (Channel) sorts by that value. A simple scroll shows how many APs are on that channel. For example, chan 157 on the 5 GHz band is by far the most populated channel for me. Fortunately, my router uses a different channel.

I might try testing its performance near my home router tomorrow (my M5 Max MacBook Pro is currently plugged in and in use as a desktop machine with my monitor, a keyboard, a mouse and some peripherals). My bedroom (where I'm using it) is directly above the half of my living room where the router is, and this is also where I tested my M3 MacBook Air from.

I used it on the other side of the upper floor of my house on Thursday evening, and its WiFi still performed fine despite being about 8 metres further away than what it is right now (and 2 rooms away). By comparison my M3 Air's WiFi is often weaker there. I suspect its N1 chip simply doesn't like the current router configuration in my house (it's giving MCS Index and NSS scores of 0).

 

[Gourlish]

The MCS Index of 0 and NSS value of 0 for M5 Max MBP is really bad. This probably means router's WiFi channel selection is heavily congested and the MBP can't establish a reliable connection to the router, which make sense given the behavior you described. Try changing the WiFi channel on the router to one not so heavily congested with noise. The M3 MBA got lucky or was closer to the router.

This is still the case even if I disconnect my M3 MacBook Air from my router.

It's one of these things (a rebadged WiFi 5 router that uses early to mid 2010s technology):

https://m.media-amazon.com/images/I/61Jwrv4XC-L._AC_UF1000,1000_QL80_.jpg

My Android smartphone, a Nokia X10, also has problems with it, and often has random DNS timeout issues while connected to it (my M5 Max MacBook Pro doesn't have this issue but simply has a slower connection than my M3 MacBook Air which seems to have no problems).


I've known that this router has seemed faulty for about 2-3 years when I got my Nokia X10, but my dad (who I still live with) insists that it is the only one he is happy to use "because it works with the solar panels", which didn't seem to work with a newer setup, allegedly because the solar panels need manually configuring to a new router in our loft, and the address of a newer router my dad tried to sort out was 1 character too long for the solar panel configuration interface to support.

I am probably going to get my dad to check if "Smart Setup" (a captive portal feature for dumb people) is turned off on our router, and also see if splitting the 2.4GHz/5GHz bands (which are currently auto-assigned) into two different networks fixes the problem, because it sounds like some newer WiFi chips struggle with both of these.

 

[chown33]

I might try testing its performance near my home router tomorrow (my M5 Max MacBook Pro is currently plugged in and in use as a desktop machine with my monitor, a keyboard, a mouse and some peripherals). My bedroom (where I'm using it) is directly above the half of my living room where the router is, and this is also where I tested my M3 MacBook Air from.

Maybe you could use a wired Ethernet adapter to connect the M5 to the router. Even temporarily, it would let you know whether an increased LAN link speed would give any net gain in internet speed.

How many different devices are connected to this WLAN? How much traffic is normal on the WLAN itself, as distinct from traffic to/from the internet?

 

[Mike Boreham]

Is Private Relay setting the same on both? This will affect Tx speed if not the same.

 

[GMShadow]

There's an app called Wi-Fi Explorer in the Mac App Store that's good for seeing if you've got a congestion issue with your wireless channels. My 2020 iMac was *really* picky about any sort of noise.

 

[Gourlish]

Maybe you could use a wired Ethernet adapter to connect the M5 to the router. Even temporarily, it would let you know whether an increased LAN link speed would give any net gain in internet speed.

How many different devices are connected to this WLAN? How much traffic is normal on the WLAN itself, as distinct from traffic to/from the internet?

There is no ethernet outlet in my room. The current router downstairs is in a place where it is impossible to connect a computer without being sat on the floor. The old ADSL outlet in my house was in my hall and was quite central, but when my home was upgraded to FTTP in 2023 the new outlet was installed in an awkward spot in the corner of my living room under the TV, probably with the assumption that the WiFi would be great with the new router the guys provided (my dad then switched back to the old FTTC era ADSL router because he couldn't configure the solar panels on the new one). Both routers were only WiFi 5, although the newer one (which my dad put into storage) did seem to perform better during the brief period I used it.


I'm now using my M5 Max MacBook Pro on my bed in exactly the same spot as the M3 MacBook Air was in when I gave those previous test scores. Interestingly my M5 Max MacBook Pro is now getting a fairly consistent TX score of 650mbps (up from 390mbps in its desk spot), which is higher than the 585mbps my MacBook Air was showing most of the time, yet its MCS Index/NSS values are still both 0 and it is still getting poorer download speeds than my MacBook Air was in the same spot.

This isn't related to WiFi, but I was also wondering if it is normal for the both the trackpad and underside (above the "MacBook Pro" engraving) of a new 16 inch Apple Silicon MacBook Pro to have a slightly hollow/rattly feel when tapped on? From what I can find I think this is within Apple's tolerances, but I'm not sure if all are like this.