How to Troubleshoot a Slow External SSD Without Replacing It

Your external SSD is transferring data at a fraction of its advertised speed. Before you blame the drive, consider the more likely culprit: the path it's connected through.

Advertised speeds are interface ceilings, not guarantees. A drive rated for 2000 MB/s will top out near 1000 MB/s on a USB 3.2 Gen 2 host port—and that's expected behavior, not a defect. The full performance path runs from the SSD module through the enclosure controller, cable, and host port, then through the OS, driver, and filesystem. Any one link in that chain can cap your throughput.

This guide walks through a diagnostic sequence that isolates which link is the real bottleneck. You'll measure the drive under controlled conditions, test the connection path, rule out software interference, and watch for thermal throttling. By the end, you'll know whether to swap a cable, change a setting, add cooling, replace the enclosure, or—only after everything else fails—replace the drive.

First, Determine Which Problem You Actually Have

The most useful question isn't "why is my USB SSD slow?" It's whether the drive is slower than advertised or slower than it used to be. Those point to different causes.

Slower than advertised usually means a connection-path mismatch. USB naming is a mess, and the port on your laptop may not support the same USB mode as the drive. SanDisk's own support documentation lists different maximum read speeds for the same drive depending on the host interface: a drive rated for 2000 MB/s on a USB 3.2 Gen 2x2 host delivers about 1000 MB/s on a Gen 2 host. That's the drive working as designed on a slower path.

Slower than it used to be suggests something changed: a cable degraded, an OS update altered driver behavior, background software started interfering, or the drive is now thermal-throttling during sustained transfers. Users on Apple's support forums have reported external SSDs dropping from roughly 800–900 MB/s to 5–30 MB/s after a macOS update—a software regression, not flash failure.

Before you spend money, you need to know which category you're in.

Rule Out the Obvious Before You Benchmark

Start with the cheap fixes. Most slow-drive cases resolve before you ever run a benchmark.

Swap the cable. Use a known-good cable rated for the interface you expect. USB-C cables are not all equal: some carry power only, some support USB 2.0 speeds, and some are marginal at Gen 2 rates. A defective or under-spec cable can silently negotiate a slower mode.

Bypass hubs and docks. Plug the drive directly into the host port. Hubs and docks share bandwidth across connected devices and can negotiate a slower USB mode. If the drive speeds up when connected directly, your hub or dock is the bottleneck.

Confirm the negotiated link speed. Don't assume the port is running at its label. Check what the OS actually negotiated. On Windows, look at the bus speed reported in Device Manager. On macOS, hold Option and click the Apple menu for System Information, then check the USB section. On Linux, lsusb -t shows the negotiated speed. If the drive is running at USB 2.0 speeds (480 Mbps), you've found your problem.

Check for competing devices. Other devices on the same USB bus can consume bandwidth. Disconnect anything you don't need during testing.

Restart and reconnect. OS-level USB enumeration glitches can leave a drive stuck at a slower mode until it's replugged. Sometimes the fix is genuinely that simple.

Measure the Drive Under Controlled Conditions

Once the obvious checks are done, you need a trustworthy number. File-copy estimates in the OS are unreliable because write caching inflates apparent speed. Use a real benchmark or a large-file copy with direct I/O flags that bypass the OS cache.

Test with a file larger than the drive's cache. Many SSDs use a portion of their NAND as a pseudo-SLC cache that absorbs writes at high speed until it fills. A small test file measures burst speed. A file large enough to exhaust the cache measures sustained behavior—which is what matters for backups, VM images, and large exports.

Measure read and write separately. A drive can be fast at one and slow at the other. Write performance is more likely to expose cache and thermal limits.

Record your test conditions. Note the port, cable, filesystem, drive fill level, and whether the drive was idle or warm. These variables change results, and you'll need them for comparison later.

Compare against the right ceiling. Your measured number should be compared against the maximum for the interface you actually negotiated, not the drive's headline spec. A 2000 MB/s drive on a Gen 2 host should hit roughly 1000 MB/s. If it does, the drive is healthy and the host port is the limit.

Isolate the Connection Path: Port, Cable, and Enclosure

Now you'll A/B test to separate host limits from drive limits.

Test on a different host port. If your computer has multiple USB ports, try them all. Some ports may support different USB modes, especially on laptops where one port is wired to a different controller.

Test on a different computer. This is the most informative test you can run. If the drive performs normally on another machine, the drive is fine and the problem is your host. If it's slow everywhere, the problem is the drive, the enclosure, or the cable.

Swap the enclosure. For DIY NVMe enclosures, the enclosure controller is a potential bottleneck. A USB 3.2 Gen 2x2 or USB4 drive drops to roughly Gen 2 speeds on a host that only supports Gen 2—but it can also drop to Gen 2 speeds if the enclosure controller only supports Gen 2. Confirm the enclosure controller and the SSD module are compatible, and that the module isn't the limiting factor. An older SATA-based enclosure will cap out around 500 MB/s regardless of the SSD inside.

The decision rule: if the drive performs normally in another compatible path, the drive is fine. The bottleneck is the enclosure, cable, or port. Replace the cheapest link first.

Check the Filesystem and Background Software

OS-level interference can slow transfers dramatically without indicating any hardware fault.

Indexing. Windows Search and macOS Spotlight both index external drives when connected. On a large drive, indexing can hammer the drive for minutes or hours after connection, making every transfer feel slow. SanDisk's support documentation explicitly lists Windows and macOS indexing as a common cause of slow external SSD performance.

Antivirus and cloud sync. Real-time antivirus scanning serializes reads and writes on large transfers. Cloud-sync tools do the same when they detect a newly connected drive. Pause both during testing to see whether throughput recovers.

Encryption overhead. BitLocker, FileVault, and hardware encryption all add overhead that varies by implementation. If your drive is encrypted, test with encryption disabled only if you can do so safely—this is a diagnostic step, not a recommendation to run unencrypted.

Filesystem choice. Filesystem and allocation behavior affect small-file and random-I/O workloads more than large sequential transfers. If your slowdown appears with many small files rather than one large file, the filesystem may be the culprit.

Test with indexing and scanning paused. If throughput recovers, you've found your problem—and it wasn't the drive.

Watch for Thermal Throttling on Long Transfers

A drive that looks fast in a short test but slows after minutes of sustained writes is often thermal-throttling. Portable drives and compact enclosures have limited cooling, and sustained writes generate significant heat.

Log temperature alongside throughput. Many SSDs report temperature through SMART. Use a monitoring tool that can log temperature over time, and run a sustained write while logging. If throughput drops as temperature rises, you've confirmed thermal throttling.

Repeat the test after cooling. Let the drive cool completely, then run the same test again. If it performs normally when cool, heat is the cause. If it's still slow, you're looking at cache exhaustion or a drive limit instead.

Cooling fixes are cheaper than replacements. Better airflow, a heatsink where the enclosure allows, or simply giving the drive space to breathe can resolve the issue. Note that adding a heatsink to a compact enclosure may affect fit, portability, and warranty—check before you modify.

Sustained Writes vs. Burst Speed: Know What You're Buying

Short sequential benchmarks often measure burst speed while the drive's cache absorbs the write. Once the cache fills, sustained throughput drops to the drive's native NAND write speed. This is by design, but it matters enormously for certain workloads.

A drive that hits excellent burst numbers may still be a poor fit for long backups, VM images, build caches, or scratch workloads. If your transfers are short and infrequent, a lower sustained figure may be irrelevant to your outcome. If you regularly move 100 GB of VM images, sustained write behavior is the spec that matters.

This is also where you should match the drive to the workload. A drive that cannot sustain large writes is a poor fit for backups regardless of its burst rating. Conversely, a drive with modest sustained performance may be perfectly adequate for ordinary file storage.

When to Reconfigure, Cool, Replace the Enclosure, or Replace the Drive

Here's the decision rule that ties each diagnostic finding to a concrete action:

If the drive is healthy on another path, the bottleneck is the cable, port, or enclosure. Replace the cheapest link first: cable, then port (if your host has a faster one), then enclosure.

If heat correlates with slowdown, try cooling before spending on hardware. Improve airflow, add a heatsink where the enclosure allows, or reposition the drive during long transfers.

If the drive is slow on every compatible path and stays slow when cool, the drive itself is the limit. Replacement is justified—but match the replacement to the workload. A drive that cannot sustain large writes is a poor fit for backups and VM images regardless of its burst rating.

If the drive is slow only during certain software conditions, the fix is configuration, not hardware. Pause indexing, adjust antivirus settings, or change how you connect the drive.

And one more thing worth remembering: a slower but stable connection may be perfectly adequate for backups or ordinary file storage. Not every slow drive needs a fix. If your transfers complete in acceptable time and the drive is reliable, the headline number on the box may simply not matter for what you do.

The goal of slow external SSD troubleshooting is not to hit the advertised speed. It's to find out whether the drive is actually failing, whether the path is the limit, and whether a cheap fix solves the problem before you spend money on a replacement.