Thanks. The system report shows the TB3 port as in use to my StarTech dock at 20Gbps. Beyond that, I can't find anywhere in the report to show the connection speed of what's hanging off the StarTech dock. I can see the SSD in storage, but it doesn't show how much bandwidth.
The connection is via USB-C to my StarTech TB3 port, so I do I know the speed is throttled from that point to the SSD. I'm okay with that. Eventually, I'll move to TB3 / NVMe.
My biggest concern is the performance numbers above. Same SSD, same port, same cable - writes are -90% and reads are -59% with encryption on APFS. The drop-off is huge.
(I've tried different file systems (MAC OS Journaled, Ex-Fat). They all give lower performance, so not an option.)
I'm now resigned to using the SSD unencrypted for photos. It's only a temporary solution until the NVMe products become more mature.
I was just checking here to see if there are any glaringly obvious mistakes with my encryption set-up.
Sidestepping a few issues which we can come back to later, little more information is needed:
-Did you recently update to Sequoia or your version of Sequoia?
Slightly different issue but if this sounds familiar this all just might be general issues with Sequoia and some external drives.
-Can you track down the actual drive model number? Seems unlikely your drive is beating its original specs so that example someone else found above may not be the same model as yours. That's like a 20 year old drive and maybe you have like a 10 year old drive...
Easiest way to get the model number at this point may be to connect drive directly to Mac and then go to System Report/Information, then to USB, and then share the info (you can scrub the serial number and related). Here is from a USB flash drive I just bought:
SanDisk 3.2Gen1:
Product ID: 0x55ab
Vendor ID: 0x0781 (SanDisk Corporation)
Version: 1.00
Speed: Up to 5 Gb/s
Manufacturer: USB
Location ID: 0x0d200000 / 1
Current Available (mA): 900
Current Required (mA): 896
Extra Operating Current (mA): 0
Media:
SanDisk 3.2Gen1:
Capacity: 250.15 GB (250,148,290,560 bytes)
Removable Media: Yes
BSD Name: disk4
Logical Unit: 0
Partition Map Type: GPT (GUID Partition Table)
S.M.A.R.T. status: Verified
USB Interface: 0
-Can you run an alternate disk benchmark tester? Something like AmorphousDiskMark would provide some more info. Also from above it looks like you posted an Excel summary of the FPS estimates from BlackMagic Disk Tester. This highlights the issue but it would easier to work from the raw MB/second numbers rather than estimated FPS of different video formats.
Side notes:
*Today's technology is giving us a lot options and therefore permutations to work with but unfortuantely this provides some ambiguity as to what you have and how it is connected. In this case:
-USB-C is a connector and now used by USB3.x, USB4, USB 4 v2 (WTF is this naming?), and Thunderbolt 3-5
-Cables can have USB-C connectors on both sides and work for all of those or just work for USB 3.x, etc.
-SSD are typically native NVMe or SATA (note these aren't exactly equivalent technically)
-Over USB, the host may connect to the drive using USB Attached SCSI Protocol (UASP) or USB Mass Storage Bulk-Only Transport (BOT)
-Over Thunderbolt, an NVMe drive basically appears to the host as if it was plugged directly into its internal PCIe bus
I bring all this up in that when you say you connect the drive via USB-C, that leaves some details ambiguous. There are certain combinations that clump together but it's hard to know for sure and therefore diagnose.
As an example, my Samsung 970 Pro NVMe drive can sit in my Plugable USB3 enclosure and connect to my Mac Mini via USB-C. But it can also sit in my Acasis Thunderbolt enclosure and connect to my Mac Mini via USB-C. But in the latter it will go >2x faster (plus gain some additional features). And since that Acasis enclosure can autonegotiate to USB or TB3 hosts and the Mac Mini autoadapts to USB or TB3 peripherals while some other Acasis enclosures are USB-only or TB3/USB4/etc only, it's hard to know what you got just by looking...
Similarly, StarTech has a lot of docks. I am surprised yours is only getting 20 Gbps and guessing yours is an older model with an older Thunderbolt chipset (probably Link Width: 0x2).
However, since you already eliminated the dock as the source of this issue by connecting the drive directly to the Mac, we can ignore the dock and whether a monitor is connected to it. If you were getting 500MB/sec when you expected 1000MB/sec then those details might be needed (monitors get priority over data over shared TB3 connections and the 1/2 speed TB3 to your dock is already less to work with).
-This is essentially a holding pattern activity. I'm trying to utilise some existing drives for additional storage until I'm happy to go down the NVMe route (concerns about current thermals and power management with an NVMe enclosure).
I wouldn't think of NVMe as hotter but rather the heat as a function of the efficiency of the particular drive and the enclosure's heat dissipation ability / quality. Where modern drives get hotter is when you run them faster. Then Thunderbolt->NVMe lets a drive run faster than through USB3->NVMe so it has the potential to run hotter. However, USB3->NVMe should run no hotter than USB3->SATA all else being equal. On the flip side, Thunderbolt->NVMe throttled to 10Gb/s it should run no hotter than USB3->NVMe. There's a even a Mac tool, Disk Mount Conditioner (from Terminal, "dmc"), to make a faster disk appear to be a slower disk (for testing/debugging purposes) but I don't recommend going that way.
For your next purchase I would look for TB3/4 -> NVMe enclosure/drive combination in which someone considered the heat dissipation quality of the enclosure and the heat efficiency of the NVMe. Either packaged by a vendor or researching the components if you plan to go the DIY route.
I can't advise on the former but can note that Arcasis sells very aluminum enclosures with fans. When I activated the fan on my new one, it dropped the actual SSD's temperature > 20 deg C. Additionally there are review sites that test NVMe drives for their MB/sec/watt. My process for selecting an SSD is 1. durability (TBW rating/TB capacity), 2. sustained sequential write speed (MB/sec after 3-5 minutes), and 3. heat efficiency (MB/sec/watt). This recently led me to purchase a Seagate Firecuda 530R versus other SSD that are cheaper and/or measure better in other dimensions. Time will tell how this theory translates into practice...
