Abysmal USB speeds on Mac Mini M2

[saldin]

I've connected USB2 and USB3 external hard drives to my Mac Mini M2, by themselves, with no other USB device connected to any of the remaining ports, and the speed is just atrocious. Screenshots attached.

This is an external USB2 WD hard drive:

This is an external USB3 WD hard drive:

When I connect the same USB3 drive to a Windows PC, file transfers normally go at around 100 MB/s.

What's going on?

 

[AppliedMicro]

Same cable?

 

[saldin]

Same cable?

Same drive, same cable.

 

[chabig]

My guess is that it's a low quality cable problem. Even though it's working better on the Windows machine, they probably have different USB controllers and negotiate the data rates differently. Try a premium cable.

 

[Bigwaff]

Just for kicks.. disconnect your Mac mini M2 from the network... either pull the ethernet or turn off WiFi adapter. Run your speed test again.

 

[bradman83]

The speed of the USB-2 drive is in line with USB-2 top speeds (especially if its a spinning drive).

When you go into System Report does it show USB-3 drive connecting at the full 5 Gbps? Or at reduced 480 Mbps?

 

[saldin]

The speed of the USB-2 drive is in line with USB-2 top speeds (especially if its a spinning drive).

When you go into System Report does it show USB-3 drive connecting at the full 5 Gbps? Or at reduced 480 Mbps?

The speed of the USB2 drive is not the problem. I put it to hammer the point that the USB3 drive is operating at nearly USB2 speeds.

Yes, the negotiated speed reported by System Info is "Up to 5 Gb/s".

 

[saldin]

My guess is that it's a low quality cable problem. Even though it's working better on the Windows machine, they probably have different USB controllers and negotiate the data rates differently. Try a premium cable.

It's the cable that came with the drive, and I don't have a premium cable lying around or will go out of my way spending for one. I can't be making excuses for premium Apple hardware when a regular Windows machine performs the way I expect it to.

 

[saldin]

Just for kicks.. disconnect your Mac mini M2 from the network... either pull the ethernet or turn off WiFi adapter. Run your speed test again.

I disconnected the network cable and ran the test, getting even slower speeds (36.0 MB/s write and 40.1 MB/s read). I had to stop the test midway to check I had selected the USB3 drive (it was) and ran it twice more while checking every time it didn't switch targets to the USB2 drive by itself (it didn't) to verify it wasn't an error. I made sure to try again with wifi completely disabled (not just disconnected) and the results remained the same.

I then ran test with the network cable connected and the transfer speed improved.

I disconnected the cable once again and repeated the tests and the transfer speed remained unchanged from the immediately previous test.

I have no idea why doing this changed things, or why there was even a change in the first place, to be honest, and I'm more than a bit annoyed at the inconsistency of it all.

 

[arw]

The results are indeed weird.
Only thing I can even remotely think of is Spotlight or any other indexing/antivirus/backup software.
Have you tried putting the USB3 drive on Spotlight's ignore list?
On a side note, what is the file system of the external drive?

Prevent Spotlight searches in files on Mac

On your Mac, prevent Spotlight from searching and showing results from files in specific folders or disks.

support.apple.com

 

[bradman83]

On a side note, what is the file system of the external drive?

He mentioned plugging it into a Windows PC which I would take as being exFAT (or maybe FAT32).

In that case it’s unfortunately a macOS problem; Apple’s implementation of exFAT is notoriously slow (FAT32 isn’t much better).

 

[macfacts]

Even a Chromebook can read/write to an external ntfs drive. Apple isn't a 2 man garage operation but acts like it.

 

[FreakinEurekan]

I disconnected the network cable and ran the test, getting even slower speeds (36.0 MB/s write and 40.1 MB/s read). I had to stop the test midway to check I had selected the USB3 drive (it was) and ran it twice more while checking every time it didn't switch targets to the USB2 drive by itself (it didn't) to verify it wasn't an error. I made sure to try again with wifi completely disabled (not just disconnected) and the results remained the same.

I then ran test with the network cable connected and the transfer speed improved.

I disconnected the cable once again and repeated the tests and the transfer speed remained unchanged from the immediately previous test.

I have no idea why doing this changed things, or why there was even a change in the first place, to be honest, and I'm more than a bit annoyed at the inconsistency of it all.

This isn’t a Mac mini issue; it’s a macOS issue. I’d suggest a post in the appropriate macOS forum.

 

[dumastudetto]

I bet a premium Apple cable would solve the problem.

 

[saldin]

The results are indeed weird.
Only thing I can even remotely think of is Spotlight or any other indexing/antivirus/backup software.
Have you tried putting the USB3 drive on Spotlight's ignore list?
On a side note, what is the file system of the external drive?

Prevent Spotlight searches in files on Mac

On your Mac, prevent Spotlight from searching and showing results from files in specific folders or disks.

support.apple.com

They're formatted as encrypted APFS. I know about enumeration being an issue on rotational hard drives, but the problem I have concerns write and read speed, and what Blackmagic measures is plain read/write speed to media.

I have iStat Menus and waited to start the tests when I saw the R and W activity lights for the device were off to avoid interference.

I haven't disabled Spotlight on the drive because I rely on it to sift thru stuff.

 

[saldin]

I had an old machine that stopped working and found myself with a free hard drive. I put it in a generic StarTech USB3 enclosure I had lying around, connected it to the Mac Mini and ran some tests to see if there was any performance difference with different disk formats.

Finding #1: It's how fast the brand name WD USB3 drive should have been. It's baffling to me why it performed on a Windows PC and underperformed on Mac.

Finding #2: Encrypted APFS is indeed slower writing, but read performance is on par with the others.

StarTech enclosure (ExFAT):

StarTech enclosure (HFS+) (couldn't test the encrypted format because it's not available anymore):

StarTech enclosure (Encrypted APFS):

PS: I also tested the WD drive with the StarTech cable and the result was a hair higher, but about the same as last time, still far from the StarTech performance:

 

[saldin]

This isn’t a Mac mini issue; it’s a macOS issue. I’d suggest a post in the appropriate macOS forum.

These are physical devices with cables. Thank you for your concern, but I think it's fine here.

 

[carylee2002]

Perhaps look into getting a mac mini hub to suppliment your setup and also have your usb output distributed via external ports. Even on my 2018 Mac Mini, I had to do this so the usb didn't share the same output thru two ports. Besides it also gives you easy access when dealing with zipdrives or Sdcards.

 

[bzgnyc2]

They're formatted as encrypted APFS. I know about enumeration being an issue on rotational hard drives, but the problem I have concerns write and read speed, and what Blackmagic measures is plain read/write speed to media.

I have iStat Menus and waited to start the tests when I saw the R and W activity lights for the device were off to avoid interference.

I haven't disabled Spotlight on the drive because I rely on it to sift thru stuff.

I'm piecing together from your various postings but it appears you are testing a (rotating) HDD with APFS (Encrypted).

Few notes:

• Don't use APFS with (rotating) HDD -- for better or worse APFS was forward-designed for SSD and is suboptimized for HDD. HFS/+ was designed for HDD and still works fine on APFS.
• Though Black Magic measures plain read/write to the drive, it isn't bypassing the OS in general and the filesystem in particular -- if the filesystem on that hardware is performing slowly, Black Magic won't make it work better. If APFS isn't writing data contiguously to the HDD, Black Magic read/writes performance will be reduced.
• That drive probably doesn't provide hardware encryption and unclear to me if Encrypted APFS is being handled in hardware or software on that setup
• Understand your reluctance to turn off Spotlight, etc but for the purposes of testing and isolating the issue it would help if you try enabling and disabling certain functions even if you will eventually need them. If nothing else that helps rule something out as a source. If the cable consistently performs well with a high-performance SSD on all computers, it probably isn't the cable. If turning off encryption or Spotlight makes things go faster that's clue. If it performs much better with HFS versus APFS or exFAT that's a clue.