Micro SD Card Speed Test: How to Check & Read Results

To test a microSD card’s speed, you use a benchmarking tool on a computer or phone that writes and reads a large file, then reports the transfer rate in MB/s. The two numbers that matter are the sequential read speed (how fast you can copy files off the card) and the sequential write speed (how fast you can copy files onto it). This article explains what those numbers mean, how to run a reliable test, and how to pick a card that actually meets your device’s needs.

What the Speed Test Actually Measures

A microSD card speed test measures two distinct operations:

• Sequential read: reading one large, contiguous block of data (e.g., a 2 GB video file). This is the speed you get when copying files from the card to a computer.
• Sequential write: writing one large block of data to the card. This is the speed when copying files onto the card or recording video directly to it.
• Random read/write (4K): accessing many small files scattered across the card. This is more relevant for running apps or using the card as internal storage on a phone, but it is rarely listed on the packaging.

Most cards advertise only the sequential read speed, because that is the highest number and the easiest to market. The write speed is often much lower and is the one that causes problems—a card that reads at 95 MB/s might write at only 20 MB/s, which is too slow for 4K video recording.

How to Run a Speed Test (Step by Step)

You need a card reader that supports the card’s bus speed. A cheap USB 2.0 reader will cap the test at about 30 MB/s, making a fast card look slow. Use a USB 3.0 or USB-C reader, or the built-in slot on a laptop if it supports UHS-I or UHS-II.

1. Insert the microSD card into the reader and connect it to your computer.
2. Download a benchmarking tool. Free options for Windows include CrystalDiskMark and H2testw. On macOS, Blackmagic Disk Speed Test (free on the App Store) is common. On Linux, you can use dd or GNOME Disks.
3. In CrystalDiskMark, set the test size to 1 GiB (the default) and the number of runs to 3–5. Click All to run all tests.
4. For a real-world file copy test, copy a single large file (e.g., a 4 GB movie) from the card to the computer and note the speed shown by the OS. Then copy it back to the card. This gives you a practical read and write speed.
5. Run the test twice with the card empty and again when it is 50% full. Speeds often drop as the card fills because the controller has to manage more data.

If you suspect a counterfeit card, use H2testw (Windows) or F3 (macOS/Linux). These tools write data to every sector and verify it, which catches fake cards that report a larger capacity than they actually have.

Understanding the Ratings on the Card

Speed classes are printed on the card itself and tell you the minimum sustained write speed, not the maximum.

• Speed Class (C2, C4, C6, C10): C10 means a minimum of 10 MB/s write. This is the minimum for Full HD video.
• UHS Speed Class (U1, U3): U1 = 10 MB/s, U3 = 30 MB/s. U3 is required for 4K video.
• Video Speed Class (V6, V10, V30, V60, V90): The number is the minimum write speed in MB/s. V30 is common for 4K, V60 and V90 for 8K and high-bitrate recording.
• Application Performance Class (A1, A2): A1 guarantees 1500 random read IOPS and 500 write IOPS; A2 doubles that. This matters for running apps from the card (e.g., on a Nintendo Switch or Android phone).

The bus speed (UHS-I, UHS-II, UHS-III) determines the theoretical maximum. UHS-I caps at 104 MB/s, UHS-II at 312 MB/s, and UHS-III at 624 MB/s. A card marked UHS-II has two rows of pins, but it will work in a UHS-I slot at UHS-I speeds.

Real-World Speeds vs. Advertised Speeds

Advertised speeds are usually the maximum read speed, measured under ideal conditions. In practice, you will see 70–90% of that. For example, a card rated at 100 MB/s read might deliver 85 MB/s in a benchmark. Write speeds are typically 30–60% of the read speed on budget cards. A high-end card rated at 160 MB/s read might write at 90 MB/s.

Also, the file system matters. Cards come formatted as exFAT, which is fine for most uses. If you reformat to FAT32 (needed for some older cameras), you lose the ability to store files over 4 GB. The file system itself does not change the raw speed, but the way the OS writes can add overhead.

Common Mistakes When Testing

• Using a USB 2.0 reader: Caps at ~30 MB/s, so a fast card looks slow.
• Testing on a phone with a slow internal reader: Many phones have UHS-I slots, but some budget models only support UHS-104 or even slower. The card’s speed is irrelevant if the phone’s reader is the bottleneck.
• Not checking the card’s temperature: After a long write, the card may throttle to avoid overheating. Run the test after the card has cooled down.
• Ignoring the difference between MB/s and Mb/s: A card that says 100 MB/s is 100 megabytes per second, not 100 megabits. 1 MB/s = 8 Mb/s. Network speeds are usually in Mb/s, so don’t compare them directly.

How to Choose the Right Card for Your Use

Match the card to the device and the task:

• For 4K video recording: Look for V30 or U3. If you shoot 8K or high bitrate (200+ Mbps), go for V60 or V90.
• For a Nintendo Switch or Android phone with adoptable storage: A1 or A2 is essential for app loading times.
• For a Raspberry Pi or security camera: Choose a card rated for high endurance (e.g., MLC or TLC with a high TBW rating). These cards are designed for constant writes.
• For a DSLR or mirrorless camera: Check the camera manual for the required speed class. Many modern cameras need U3 for continuous shooting.

Frequently Asked Questions

Why is my card slower than the number printed on it?

The printed number is the maximum read speed under ideal conditions. Real-world speeds are lower due to the card reader, the file system, the card’s controller, and thermal throttling. Also, many budget cards list the read speed but have a much lower write speed, so if you are testing write, you will see a big difference.

Does the brand of the card matter?

Yes, but not in the way you might think. Major brands like SanDisk, Samsung, and Kingston use reliable controllers and NAND flash, but counterfeit cards with those brand names are common. Buy from an authorized seller and test with H2testw to verify authenticity. No-name cards often fail to meet even their own specs.

Can I use a microSD card in a full-size SD adapter for testing?

Yes, but the adapter can affect speed. A cheap adapter may not have the contacts aligned perfectly, reducing throughput. If you are testing a card’s true speed, use a direct microSD reader or a high-quality adapter.

How long does a speed test take?

A typical CrystalDiskMark run with 1 GiB and 3 passes takes about 2–5 minutes, depending on the card. H2testw with a 32 GB card can take over an hour because it writes and verifies the entire capacity.

What is a good speed for a microSD card in 2024?

For general use (photos, documents), a U1 card with 30 MB/s read is fine. For 4K video, look for U3/V30 with at least 60 MB/s read. For professional use, UHS-II cards with 250+ MB/s read are available, but they require a UHS-II reader to achieve those speeds.

Does formatting the card improve speed?

Formatting can restore performance if the card is heavily fragmented or has corrupted file tables, but it will not make a slow card faster. If a card is consistently slow, it may be counterfeit, damaged, or just a low-end model.

Related

• Split Numbers from Units of Measure?
• Speed Up Linux System Calls on Power
• Enshrouded: All Beacon Dungeons for Gate to the Stars