SD Card Speed Test: How to Check SD Card Performance on Windows 11/10?

SD cards are everywhere. Your camera. Your dash cam. Your kid's Nintendo Switch. That Raspberry Pi project sitting on your desk. Maybe your Steam Deck is loaded with one too. They're small, cheap, and easy to overlook, until something goes wrong.

Here's the thing most people don't realize: the speed gap between a good SD card and a bad one isn't 10% or 20%. It can be 5x. Sometimes more. And that gap shows up in places you'd notice immediately, a camera that buffers between shots, a game that takes two minutes to load, video footage that stutters or drops frames right when it matters.

The frustrating part? You usually don't find out until you're already using the card. By then you've shot the photos, recorded the video, or copied the files. And if the card turns out to be a counterfeit? Those files might not even be safe.

The fix is simple: test SD card speed before you trust the card with anything important. It takes five minutes, it's free, and it tells you exactly what you're working with.

Why Test SD Card Speed?

Four reasons. Some are obvious, some less so.

• Verify what you actually bought. The number on the SD card packaging represents the best-case scenario. A card labeled "up to 100 MB/s" might deliver 30 MB/s write speed in practice, and that's if it's genuine. Manufacturers print burst read speeds under ideal lab conditions, not the sustained write performance you'll get when recording 4K video or copying a batch of RAW photos. The only way to close the gap between marketing and reality is to test it yourself.
• Figure out why your device is slow. Not every sluggish device has a slow SD card inside. But plenty do. Camera buffering between burst shots. Drone footage that corrupts mid-flight. A Nintendo Switch with load screens that feel endless. An Android phone that stutters because the SD card is being used as internal storage. Before you blame the device, check the card. A quick speed test either confirms or eliminates it as the problem.
• Spot fake SD cards. Counterfeit SD cards are a real, widespread issue, especially from third-party sellers on large marketplaces. Some fakes have modified firmware that reports a 256GB capacity when the actual flash storage is only 32GB. Everything works fine until you pass 32GB, at which point the card silently overwrites your older files. Other fakes are real cards with fake speed ratings: a cheap Class 4 chip inside a body that says V30. Speed testing catches both types. If a U3-rated card tests at 8 MB/s write speed, you know something's off.
• Your card might be dying. SD cards wear out. Flash memory has a limited number of write cycles, and as cells age, the controller works harder to keep everything running, remapping bad blocks, retrying failed writes, running error correction behind the scenes. The result? Speed drops. A card that benchmarked at 90 MB/s when new and now tests at 35 MB/s is sending you a signal. Bad blocks accumulating, aging NAND cells, a struggling controller, any of these can cause it, and speed is usually the first symptom you'll notice.

Read More: How to Check SD Card Health in Windows 11/10/8/7?

Understanding SD Card Speed

Most guides skip this part. They tell you to run a benchmark and stare at the big number. That's not very useful if you don't know what the big number actually represents. So let's fix that.

Read Speed vs. Write Speed

Read speed is how fast data comes off the card. Opening photos, playing back video, loading a game, transferring files to your computer, that's all read operations.

Write speed is how fast data goes onto the card. Recording 4K video, shooting burst photos in RAW, copying files from your PC. Write operations.

These two numbers are almost never the same on an SD card. Write is almost always lower. Sometimes dramatically so, a card might read at 100 MB/s and write at 20 MB/s. And here's what matters: for most real-world use cases, write speed is the one that causes problems. A slow read means your photos load a bit slower. A slow write means your camera buffers, your video drops frames, your file transfer takes forever.

When you run a speed test, look at both. But give extra attention to the write result.

Sequential vs. Random Speed

This one separates "good for copying files" from "good for running applications."

Sequential speed measures how fast the card handles one continuous stream of data, like copying a movie file from point A to point B. It's the number on the box. It's always the highest number.

Random speed, measured in IOPS (Input/Output Operations Per Second), measures how fast the card reads and writes small chunks of data scattered across different locations. That's what happens when you launch an app, query a database, or boot an operating system from the card.

Here's why this matters: a card can have excellent sequential speed and terrible random speed. Manufacturers almost never advertise the random number. So you end up with a "fast" card that makes your Raspberry Pi feel like it's running through mud, or your Steam Deck games take three times longer to load than they should.

A proper disk benchmark tool measures both. When you see the results, don't just look at the sequential numbers. Check the 4K random read and write too, especially if you're using the card for anything beyond storing photos and videos.

Sustained vs. Burst Speed

This is the one that tricks people.

Modern SD cards have a small SLC cache, a fast write buffer that absorbs incoming data at high speed. While data is hitting the cache, everything looks great. 90 MB/s, 100 MB/s, sometimes more.

Then the cache fills up. And write speed drops off a cliff.

You start recording 4K video. First minute looks perfect. Then the camera stutters, drops frames, or stops recording entirely. The card's sustained write speed can't keep up with the bitrate.

A 5-second speed test only shows you the burst number. To get the sustained number, you need a test that runs long enough to fill and exhaust the cache. That means configuring a larger test file, at least 1GB, ideally 2GB or more. Tools like DiskGenius let you set the test data size and block size precisely, so you can push past the cache and see what the card actually delivers when it matters.

Method 1: Test SD Card Speed with DiskGenius Free

DiskGenius is free disk management software for Windows that includes a built-in speed benchmark. It measures sequential read/write, random read/write, and IOPS. Beyond speed, it can verify your card's real capacity and scan for bad sectors, which is handy if your test results raise red flags.

Free Download DiskGenius

Step 1: Use a quality card reader.

This is the single most common source of inaccurate results. A USB 2.0 reader maxes out around 35 MB/s. If your SD card can do 100 MB/s but your reader can only push 35, you'll think the card is slow when it's actually the reader holding it back. Use a USB 3.0 or USB 3.1 reader. If your test results seem capped at 30 to 35 MB/s no matter what card you test, the reader is almost certainly the bottleneck.

Step 2: Connect directly to your computer.

Plug the reader straight into a USB port on your motherboard. On a desktop, that usually means the back panel ports, not the front. Avoid USB hubs. They split bandwidth across devices and add latency. Same goes for extension cables. You want the cleanest possible connection between the card and your system.

Step 3: Close background programs.

Cloud sync services, antivirus real-time protection, file indexing, all of these can read from or write to the drive during your test and skew the numbers. Close them. You want that SD card doing nothing except running your benchmark.

Run the Speed Test

Step 4: Download and install DiskGenius.

Grab it from the official website (https://www.diskgenius.com/download.php). The free version supports full speed benchmarking, no need for a paid license.

Step 5: Select your SD card and open the speed test.

With the card inserted via your USB 3.0+ reader, launch DiskGenius. Find the SD card in the drive list on the left. Be careful here, make sure you're selecting the SD card, not your system drive or an external hard drive. Navigate to "Disk" → "Disk Speed Test" from the menu bar.

Step 6: Configure the test parameters.

This is where DiskGenius gives you more control than most simple benchmark tools.

You'll see two testing modes:

• File Test: runs the benchmark through the file system. This gives you results closer to what you'd experience in real-world usage.
• Benchmark Test: tests raw disk performance at the block level. This number will be higher because it bypasses file system overhead.

You can also adjust:

Block size: smaller blocks (like 4K) test random performance. Larger blocks (512KB, 1MB) test sequential. Testing with multiple block sizes gives you a complete picture.

File size: set this high enough to push past the SLC cache. 1GB minimum, 2 to 4GB if you want the sustained speed number. If you only test with a tiny data size, you'll get the burst number, which can be misleadingly optimistic.

Test type: read-only, write-only, or mixed read/write. Run at least one read test and one write test separately so you get clear numbers for each.

Step 7: Run the test and review your results.

Hit "Start Test" and let it finish.

DiskGenius will display sequential read and write speeds in MB/s, random read/write at various block sizes, and IOPS. Pay attention to the speed-over-time graph too, if speed drops sharply partway through the test, that's the SLC cache being exhausted. The lower number is your sustained speed.

Method 2: Check SD Card Speed Using Built-in Windows Tools

Windows has a couple of built-in options that give you a rough speed estimate. They're less precise than a dedicated benchmark, but they work in a pinch.

1. File Explorer Copy Test

The simplest approach. Find a large file on your computer, a 1 to 2 GB video works well. Copy it to the SD card and watch the transfer dialog; Windows shows a real-time speed estimate. Then copy it back to test read speed.

It's not precise. Windows includes caching effects in its estimate, and you're only testing one type of sequential operation at whatever block size the file system happens to use. But if a brand-new U3 card shows 12 MB/s during a file copy? You already know something's wrong, no fancy tools needed.

One tip: disable write caching in Device Manager before testing. Otherwise Windows may buffer data in RAM first, making the card look faster than it really is.

2. winsat Command

Open Command Prompt or PowerShell as Administrator and run:

winsat disk -drive H

Replace H with your SD card's drive letter. Windows runs a series of sequential read and write tests and prints the results in the terminal. It's actual numbers, not just a visual estimate, which is an improvement over the file copy method.

The downsides: output is raw text with no graph, no random speed data, no configurable parameters. Run it twice because the first pass benefits from caching. Also worth noting that winsat is officially deprecated, it still works on current Windows versions, but Microsoft could remove it in a future update.

Where these methods fall short: neither one shows you random performance, sustained vs. burst speed, or IOPS. For a quick post-purchase sanity check, they're fine. For a full SD card performance test with actionable detail, a tool like DiskGenius Free is the better route.

How to Interpret SD Card Speed Test Results?

You've got numbers. Now what? Start with this reference table. It shows the minimum speeds you need for common tasks:

Now match your results to your use case.

Sequential write meets or exceeds the threshold? You're in good shape.

Sequential write is far below the card's advertised rating? Three possibilities: the card is fake, the card is failing, or your card reader is the bottleneck. Re-test with a different reader before drawing conclusions.

Random 4K numbers are low? That's expected for most SD cards. If you need strong random performance, look for cards with an A2 rating, they're specifically designed for high IOPS.

Speed drops partway through the test? That's the burst-to-sustained transition. The lower number is what you'll actually get during long video recordings or large file copies. If it falls below your use case's minimum, the card isn't suitable for that workload even though it "should" be based on the box.

Can Speed Tests Damage SD Card?

This is a question that comes up a lot, and the concern makes sense on the surface. A write-mode speed test is deliberately writing data to every corner of the card. Doesn't that eat into the card's lifespan?

Technically, yes. Every write operation consumes a tiny fraction of the card's total write endurance. But the amount a single speed test writes is so small relative to that total that it's practically meaningless.

To put it in perspective: a typical SD card is rated for somewhere between 1,000 and 10,000 write cycles per cell, depending on the NAND type. That translates to terabytes of total data written over the card's lifetime. A single benchmark run that writes 2GB of test data is a rounding error. You'd need to run the test hundreds or even thousands of times back to back before seeing any measurable impact on the card's remaining lifespan.

So no, running a speed test once, or even once a month, won't damage your SD card.

What can damage an SD card is something else entirely. Repeatedly yanking the card out mid-write. Using it in extreme heat or cold beyond its rated operating temperature. Writing to it constantly without rest (like in a 24/7 security camera that never stops recording). A manufacturing defect. Those are the things that kill SD cards. Not a five-minute benchmark.

There is one caveat worth mentioning, though. If the card is already failing, if the NAND cells are degraded and the controller is barely holding things together, a heavy write test could theoretically push it over the edge. Not because the test itself is harmful, but because the card was on its last legs and any significant write operation would have done the same thing. If your card is showing signs of failure (speed drops, intermittent recognition, read-only lock), get your data off it first. Then test.

As for your data: a read-only speed test doesn't modify anything on the card. It just reads. Your files stay exactly where they are. If you're running a write-mode test, the tool may overwrite some data depending on its configuration. DiskGenius lets you choose read-only mode so you can benchmark without touching existing files. When in doubt, back up first, not because the test is dangerous, but because backups are always a good idea.

How to Check SD Card Health (Is Your Card Failing)?

Speed tests catch performance problems. But sometimes the card has deeper issues that haven't shown up as a speed drop yet, or the speed drop is severe enough that you need to know what's going on underneath.

Warning signs:

• Benchmark results dramatically lower than when the card was new

• Files becoming corrupted or disappearing without explanation

• The card switches to read-only mode (the controller does this as a self-preservation measure when it detects critical failure)

• The SD card keeps asking to format

• Write operations that used to be fast now take much longer

• Intermittent recognition failures, sometimes the SD card not showing up

What to do about it:

DiskGenius includes a surface scan feature that checks the entire card sector by sector for bad blocks. If the scan turns up a significant number of bad sectors, the NAND flash is physically degrading. At that point, the card is on borrowed time. Copy your data off it and replace it.

You can also run a capacity verification to rule out a fake card. This writes data across the reported capacity and reads it back, confirming whether the card is actually as big as it claims. A 256GB card that starts corrupting data past 32GB is really a 32GB card with tampered firmware.

Keep a record of your speed test results over time. A downward trend across multiple tests is an early warning, you can replace the card before you actually lose data, instead of discovering the problem when your vacation photos vanish.

When to replace immediately:

Bad sectors found during a surface scan. Write speed dropped 40% or more from the original benchmark. The controller has locked the card to read-only. Any sign of data corruption. Don't wait for it to get worse. Flash memory doesn't heal.

How to Improve SD Card Performance?

If your card is slow but not failing, maybe it's just a budget card, or it's been formatted poorly, there are things you can try before buying a new one.

1. Upgrade your card reader. Swapping from a USB 2.0 to a USB 3.0+ reader is the single biggest improvement most people can make. If your card is UHS-II (two rows of pins on the back), make sure your reader supports UHS-II too. A UHS-II card in a UHS-I reader runs at UHS-I speeds.

2. Reformat the SD card. File system fragmentation builds up over time, especially on cards that have been used, deleted, reformatted, and reused across different devices. Reformatting with the right file system and allocation unit size clears this out. Use exFAT for cards 64GB and larger, FAT32 for 32GB and below.

3. Pick the right card for the job. Video recording demands sustained write speed, look for V30, V60, or V90 ratings. Gaming and app storage need high random IOPS, look for A2 ratings. General purpose use is fine with a U3 / A1 card. Buying the most expensive card doesn't help if it's optimized for the wrong workload.

4. Don't fill it to the brim. Leave 10 to 15% of the card's capacity free. The controller needs empty space for wear leveling and garbage collection. A nearly full card forces the controller to work harder, which slows everything down.

5. Always eject safely. Pulling an SD card out mid-write can corrupt the file system. It takes two seconds to click "Eject" in Windows. Use that time.

Frequently Asked Questions

How often should I test my SD card speed?

Immediately after buying it, verify the card is genuine and performing at its rated speed. After that, every 3 to 6 months for cards used in critical roles like dashcams, security cameras, or professional photography. For casual use (storing photos, occasional file transfers), once after purchase is probably enough.

Why are my results much lower than what's printed on the card?

Most likely reasons: your card reader. USB 2.0 caps out around 35 MB/s and will make any card look slow. Other possibilities: you're comparing the write speed result to the advertised read speed (they're different numbers), or the card is counterfeit or degraded. Re-test with a USB 3.0+ reader and check that you're comparing like-for-like.

Can I test SD card speed on Mac or Linux?

Yes. On Linux and Mac, you can use the dd command for a basic sequential test, or a tool like f3 (Fight Flash Fraud) to check both speed and capacity. The speed concepts covered in this guide apply regardless of operating system.

What's the difference between a speed test and a health check?

A speed test measures how fast the card performs right now, and the disk health check examines the physical integrity of the storage cells underneath. A card can pass a speed test but still have bad sectors forming. Ideally, run both, speed test to verify performance, surface scan to verify the card isn't developing physical problems.

My card is fast for reads but slow for writes. Is that normal?

Completely normal. Write speeds are almost always lower than read speeds on SD cards. The important question is whether the write speed meets the minimum for what you're doing with it. Check the reference table in the results interpretation section above.

Will a speed test damage files?

It depends on what you mean by "files."

If you're worried about the photos, videos, and documents currently on the card, the speed test won't touch them. If you've recently accidentally deleted files or formatted the card and you're hoping to recover that data, do not run a speed test. Not even a read-only one.

The rule is simple: if the card contains data you need to recover, don't test it. Don't benchmark it. Don't write to it at all. Use SD data recovery software to recover everything you need first.

Conclusion

Testing SD card speed takes less time than brewing a cup of coffee. Five minutes with the right setup, a USB 3.0 reader, a direct connection, and a benchmark configured to push past the SLC cache, and you know exactly what that card can do.

You also know what it can't do. And if the results look suspicious, you can immediately verify the card's real capacity and scan for bad sectors without switching tools.