A slow external SSD is not automatically faulty. Compare the computer port, cable, hub and copy conditions to narrow down the bottleneck.
What to check first
For example, Samsung specifies the T7 as USB 3.2 Gen2 10 Gbps, with sequential read/write speeds up to 1,050/1,000 MB/s. These manufacturer figures come from internal testing with the supplied cable, not a promise for every setup. A USB-C shape alone does not identify transfer capability.
A worksheet you can use
Compare copies of the same files. Do not move your only originals, and there is no need for an unofficial full-drive write test.
| Item | What to record |
|---|---|
| Computer port | Exact computer and port transfer specification |
| Cable | Model and supported data specification |
| Hub | Model, port specification and other attached devices |
| Direction | Which device is read from and which is written to |
| Files | Count, total size and large files versus many small ones |
| Time | Completion time and other tasks running |
Check in this order
- Connect the SSD directly to a supported computer port with its undamaged supplied cable
- Copy the same files in the same direction and compare with the hub route
- Change only one port or cable at a time and record where the result changes
If the problem remains
If performance remains poor, send the connection record and copy conditions to manufacturer support. The source drive’s reading speed can also limit the copy. Prioritize backup when disconnections or errors occur. Do not format the drive or disable security features just to chase speed.
Related guide
Does a Highly Rated Product Fit You? A Review Evidence Worksheet
A worked example
Hypothetical example: copies are slow through a hub but faster when connected directly. That comparison justifies checking the hub path further, but does not prove the SSD is faulty. Keep the files and transfer direction consistent.
Separate steady slowness, a mid-copy drop and errors
A symptom alone cannot prove a faulty cable or SSD. Pair observations with a next check before buying replacement parts.
| Observed result | Next check | What it does not establish |
|---|---|---|
| Slow from the start, but copying completes | Check exact port, cable and hub specifications, the source drive and file mix; compare a direct connection. | A speed number alone does not identify the cable standard or prove SSD failure. |
| Fast initially, then slower | Record the total completion time, and whether large files were followed by small ones. Do not combine initial and final rates as though they were one sustained rate. | A drop alone does not prove overheating or failure; cache and file mix can also matter. |
| Errors, disconnections or unreadable files | Stop speed comparisons and preserve originals and existing backups. With persistent read errors, consider manufacturer or recovery support before repeated copying. | Timings from a failed transfer do not establish normal performance. |
Change one connection route at a time
Copy an already backed-up set of files into a separate destination folder; do not move your only originals. Keep file count, total bytes, source, destination and direction the same. Record completion time and other running tasks for each route. Do not test if the destination lacks space. One comparison of a direct connection and your usual hub route is enough to start.
A hypothetical average from total completion time
These are teaching numbers, not an SSD test. Assume 12,000,000,000 bytes copied, with one MB defined as 1,000,000 bytes.
| Route | Completion time | Whole-copy average |
|---|---|---|
| Direct connection | 40 seconds | 300 MB/s |
| Same files and direction, through the hub | 80 seconds | 150 MB/s |
Average MB/s = total bytes ÷ elapsed seconds ÷ 1,000,000. This example gives a reason to investigate the hub route; it does not prove a defective hub or the SSD’s maximum performance.
The second copy may encounter a different cache or background workload. If the difference is small or conditions changed, record that uncertainty instead of declaring a winner. Do not compare a peak-rate screenshot with an average or repeatedly fill the drive to chase a higher number.
Choose a stopping point from the record
- You found a route that finishes the work without errors: keep its configuration and file conditions on record and use it. You do not need to replace parts until an advertised maximum is matched.
- Copies complete but an unexplained difference remains: send support the computer, SSD, cable and hub models, port specifications, direction, total bytes, elapsed time and any errors.
- Errors repeat: end the performance comparison and focus on preserving data and obtaining support.