BUT... don't worry about it.
TRIM will make next-to-no difference at all. It's one of those issues that is "lots of smoke but no fire".
Produce documented info if you have it.
https://forums.macrumors.com/threads/slow-ssd-write-speeds-on-older-imac.2053041/#post-24734082
Here is a recent thread you should read. This person has a Samsung 840 running without TRIM and write speeds had dropped to 25MBps. Just turning TRIM on brought the write speeds up to 150MBps. Then running fsck to TRIM unused blocks fully restored write speeds to 460MBps.
So there is your evidence. There are other threads showing exactly the same results.
Why/How does Windows support trim on this particular SSD external drive (Samsung T5) over usb compared to Mac OS?
You are correct. USB cannot pass along the SATA TRIM command.
What some people have read and it has caused confusion, is some UASP (USB-attached SCSI protocol) enclosure chipsets can convert the SCSI Unmap command (similar to TRIM) to TRIM commands at the drive. Windows can run this SCSI Unmap command with the Powershell command Optimize-Volume and the -ReTrim option. OS X cannot run the SCSI Unmap command (at least not that I have been able to find). So at this point, no matter what chipset you use, you are not going to get TRIM over USB on a Mac.
There is some discussion at the bottom of this article.
It doesn't. Here is an older post of mine that explains a bit.
No... you would have to run it manually.Thanks for the info. Quick question. Is that Powershell Command run every time the SSD Drive is connected via USB to the computer or once only? Does it subsequently send the correct commands everytime when the SSD Drive is attached to that computer?
Thanks
Garbage collection is an additional feature, while TRIM is a mandatory feature, especially for OS and applications disks (virtual memory).TRIM is supported in HS but you can't send SATA commands like trim over USB as far as I know. its also not really needed anymore as modern SSDs have garbage collection built in so you can have that thing packed all the way full without issue. thats what I've done for years now without a single issue.
No. TRIM DOES NOT matter that much any more.
It's Sushi's "advice" that is misinformation.
I stand by this post.
An open request to anyone reading this that is currently booting and running an iMac or Mini via USB:
Has the lack of TRIM mattered much for you?
If so, have you been able to document the information?
I say to Sushi:
Put up, or shut up.
Produce documented info if you have it.
For someone with an older Mac (2012-2015), who has only a platter-based hard drive inside, and who wants to "squeeze more years of use" from it, adding an external USB3 SSD is a cost-effective solution. No breaking open the machine, no risk of breaking something inside.
It will greatly speed up the computer, and so long as the user leaves a good amount of free space on the SSD, it will probably continue to perform well long enough to "outlive" the computer itself.
Here was my reply to your demand almost a year ago. If you like, I can find other posts with more examples.
1700MB/second is 13.6 gigabits/second, and well above the maximum throughput for USB 3.1 Gen 2. This "benchmark" is obviously caching data and is not accurate.Hi all, Im new here.
Mac TRIM over USB possible new info. So I got an Samsung Evo 860 into a USB 3.1 Gen2 enclosure. Driver formatted to APFS
- Using BlackMagic Speed
- Used the 5GB test, initial bursts speed were even higher than below (probably drive cache)
- Result: Write: around 480MB/s Read around 500MB/s
- Turned TRIM on via terminal (even though TRIM is not suppose to be supported over USB)
- Same Test = Write: around 520MB/s Read around 1600 -1700 MB/s
1700MB/second is 13.6 gigabits/second, and well above the maximum throughput for USB 3.1 Gen 2. This "benchmark" is obviously caching data and is not accurate.
Trim is not supported on USB in macOS.
The only thing that can be made of it is that the benchmark application is not working properly. It's simply impossible to get 13Gb/sec out of an interface that's only rated to a maximum theoretical 10Gb/sec.Yes but, how do you explain that the cache on this is far lower (1GB) and the test is set to 5GB? Thats why I set the test at 5GB to avoid any cache 'misreadings'. The 1600-1700/MB/sec was sustained, it lasted the whole 5GB write test cycle and I did the test several times with the same result. If this was purely cache, would it not simply burst into 1600MB/sec for about 1/4 of a second then hover to it's advertised 550MB/sec thereon.
What do you make of this?
Hi all, Im new here.
Mac TRIM over USB possible new info. So I got an Samsung Evo 860 into a USB 3.1 Gen2 enclosure. Driver formatted to APFS
So people, what do you make of this test? Is TRIM now supported over USB? The Mac system profiler doesn't show this however, how do you explain the above test. What's going on? Anyone?
- Using BlackMagic Speed
- Used the 5GB test, initial bursts speed were even higher than below (probably drive cache)
- Result: Write: around 480MB/s Read around 500MB/s
- Turned TRIM on via terminal (even though TRIM is not suppose to be supported over USB)
- Same Test = Write: around 520MB/s Read around 1600 -1700 MB/s
- Then turned TRIM off again
- Speeds went back to 480MB/s Read around 500MB/s
- Then turned TRIM ON again
- Speeds went back up to 520/1600+MB/s
Meanwhile, I did some research and apparently there are some external USB drive enclosures that support the UASP protocol and translate the SCSI UNMAP command to the SATA TRIM command. So now it is possible to have TRIM over USB although I don't know if MacOS sends these UNMAP commands to UASP drives. Maybe somebody else knows the answer to that.
You are correct. USB cannot pass along the SATA TRIM command.
What some people have read and it has caused confusion, is some UASP (USB-attached SCSI protocol) enclosure chipsets can convert the SCSI Unmap command (similar to TRIM) to TRIM commands at the drive. Windows can run this SCSI Unmap command with the Powershell command Optimize-Volume and the -ReTrim option. OS X cannot run the SCSI Unmap command (at least not that I have been able to find). So at this point, no matter what chipset you use, you are not going to get TRIM over USB on a Mac.
There is some discussion at the bottom of this article.
Here is an older post of mine that discusses this.
You can run the UNMAP command from Windows, but macOS does not support it.
The only thing that can be made of it is that the benchmark application is not working properly. It's simply impossible to get 13Gb/sec out of an interface that's only rated to a maximum theoretical 10Gb/sec.
Ah yes, since USB3.1Gen2 can't exceed 10Gbps Im very curious why 1700MB read happens only with TRIM enabled and drops to about 480MB/sec with TRIM disabled. It' certainly weird, especially over USB mac.
PS: The Samsung 1TB Evo 860 Im testing on is brand new. My BMD benchmark tool is version 3.2
...
PS: The Samsung 1TB Evo 860 Im testing on is brand new. My BMD benchmark tool is version 3.2
Well the other weird thing is that TRIM should have literally no effect on read performance, it's only an optimization for writing data to an SSD.
So TRIM being enabled or disabled is really irrelevant here. Also, I don't even know what it means for TRIM to be enabled or disabled in this case, because TRIM is enabled/disabled on a per-drive basis. How are you enabling TRIM on a USB drive? There isn't a feature in MacOS to do this.
[doublepost=1551382556][/doublepost]
Not only will TRIM have literally no effect on read performance in any situation, TRIM will have no effect on write performance for a brand-new drive, until it has been completely filled with data.
Something has gone wrong with your benchmark software. Whatever it is, it isn't related to TRIM.
Im turning on TRIM via the terminal: sudo trimforce enable, command
Oh well, we will never know then, maybe BMD Speed-test was faulty, will never know now, I returned the USB enclosure and and now have a bare SSD -> now looking to buy Thunderbolt enclosure.Oh okay, right, yes, that will only affect SATA drives. There is no way to enable TRIM for USB drives, as has been discussed, because there is no way to issue SATA commands over USB. There is a way to issue SCSI commands, via UASP, hence the discussion of UNMAP being translated to TRIM. But I doubt that a command called "trimforce" would enable the SCSI UNMAP command.
But this is all a moot issue with regard to your benchmark results, for the reasons I explained earlier.