2021 16" MBP is 13% slower than 10 year older 17" MBP

[SimonTT]

I have a 2011 17" i7 MBP with 16GB Ram a 1TB SSD and the infamous AMD dGPU.
The dGPU is dead and after a long time I revived it by deactivating the dGPU deep in the system and installing Catalina via DosDude1s Patcher.
It is surprisingly fast, perfect for light browsing, that's why I wanted to document how I got it to this state from constant boot looping because of the dead dGPU.
I did the Speedtest by Ookla and was shocked when I saw that the 17" was not just as fast as my 2021 16" M1Pro but actually faster by 14% (55,39 vs 48,52 Mbps Down). Thinking this must have been a fluke I repeated the test thrice with the same result each time.
I couldn't believe that the 16" with Wifi 6 is slower than the 17" with Wifi 4, so I gathered a couple apple devices an ran the test 10 times on each one. I made sure to run nothing but safari (apart form my iPad Air which ran Apple Music and the 13" which ran Numbers and wasn't representative anyway since its headless). I started each test once the last device was finished in this order : "16", 17", 13", Air 4, mini 13, 2020 12.9, 2015 12.9" and repeated this 10 times. All devices were on the same Wifi (Wifi 6 Router) and desk.

The first test was no fluke the 17" was 13% faster at download and 6% at Upload than the 16".
I have no idea how a device that's 10 years older, has a broken GPU and is running an unsupported OS can be faster, especially with Apple silicon. A maxed out M1 Pro is slower than a second gen i7. If you have any possible explanations please write them down but I really have no idea how this works.

It would be great if others could repeat the test in order to be a little more scientific.

If anyone wants to see my detailed results here you should be able to access them "https://www.icloud.com/numbers/074FYZ8hty0e32-I30HNt4Mag#Speedtest"

 

[Sheepish-Lord]

Lots of problems when trying to do a test like this. Here’s a few you have;

1. What is your internet plan? Getting 50Mbps is pretty weak and unless your plan supports higher bandwidth both of those machines will easily cap out their WiFi.

2. 55Mbps vs 48Mbps isn’t that different. Additionally, the way Ookla works is through multi connections unless you specify a single point connection. Almost a non issue honestly and could vary given routine traffic or where you live.

3. If you truly have a WiFi 6 router then you need to do an internal data transfer test. The external speed test you’re doing is being capped by your internet plan.

4. Your oldest device is a 802.11n and has a max theoretical data rate of 450Mbps so unless you have an ISP plan exceeding approx 300Mbps then you won’t see the benefits of the new laptops WiFi 6 because you’re externally data capped as I stated earlier.

Think of it like owning a Ferrari and a Prius. Both can easily do the public speed limit on the street (your external ISP data) but take them on the race track (internal network speed) and the Ferrari wins. Not to be rude but you seem to be missing basic network infrastructure concepts as your testing process is incorrect for what you’re attempting to convey.

 

[Donald.jung]

Its probably your network, I'm getting 500mbps on newer MacBook

 

[throAU]

I did the Speedtest by Ookla and was shocked when I saw that the 17" was not just as fast as my 2021 16" M1Pro but actually faster by 14% (55,39 vs 48,52 Mbps Down). Thinking this must have been a fluke I repeated the test thrice with the same result each time.

You may find that the new MacBook is preferring IPv6 which may have different routing through your ISP which could impact the speed test. This was a change made in one of the OS updates that I do not believe your 2011 MacBook supports. Even worse, if you do not have native IPV6 through your ISP, it could be tunnelling). Could also be a feature like private browsing on the new machine changing the routing as well.

I have several old Macs (including a 2011 15"), and am not getting the same experience as yourself.


Your best bet is to test transfer between the two machines and a device on your LOCAL NETWORK (and measure the speeds doing that) to eliminate internet routing related variables.

 

[Fishrrman]

Here's something to try.
It won't hurt anything, and might help.

On the new MBP:
(I may have some things "unclear" here because I'm composing this reply on an older OS)
Open the Network settings
You want to be able to see the pane that shows all available connections
Normally wifi will be "at the top" for a Mac laptop
Click on the current wifi entry to select it
Then, at the bottom, click the "-" icon to DELETE it (that's right, delete it)
Once deleted, REBOOT THE COMPUTER

Now, go back to the Network pane
There should be NO entry for wifi
We need to create a new one
Go to the bottom and click the "+" sign to ADD a new one

Once done, try the speedtest again.
I suggest "Speedtest.net" -- do it right online.

Seems that I recall when the m-series Mac laptops first started hitting the market, numerous users were complaining about slow wifi connection speeds.

Doing the above seemed to help.

I had a similar experience when I first got my MacBook Pro 14" (m1pro).
I did the "delete, then re-create" routine on it, and that seemed to fix it.

Again, it hurts nothing and is worth a try.
Let us know how you make out.