Just Got a New Drive? Here's How to Test Speed of your SSD or HDD

You unbox a brand new NVMe SSD. The product page says 3,500 MB/s. You plug it in, run a quick copy, and... something feels off. It's not slow, exactly, but it doesn't feel as fast as the numbers promised.

Before you assume something's wrong with the drive, here's the thing: the speed printed on the box is a best-case number. What you actually get depends on which slot you plugged it into, what cable you used, whether your BIOS is configured correctly, and a handful of other factors most people never think about. The drive might be perfectly fine.

The only way to know for sure? Test disk speed before you start putting data on the drive. It takes a few minutes, and it tells you everything you need to know: is this drive performing as expected, or is there something that needs fixing?

Running a proper speed test before filling a new drive with data is something experienced users always recommend. You paid for a specific level of performance. Better to confirm you're actually getting it now than discover a problem later.

Why Do Drives of the Same Type Show Different Speeds?

This comes up all the time. Two people buy the same NVMe SSD model. One gets 3,400 MB/s, the other barely breaks 900 MB/s. If you test the speed of an SSD and compare it against someone else's results for the same model, the gap can be surprisingly large. Most of the time, the drive itself isn't the problem.

The slot protocol. Not all M.2 slots are created equal. They look the same physically, but the underlying wiring can be very different. A slot connected through SATA tops out around 600 MB/s. One wired to PCIe 3.0 x4 can do 3,500 MB/s. PCIe 4.0 x4 goes up past 7,000 MB/s. Put a PCIe 4.0 drive in a SATA slot and you've just left most of its performance on the table. The same logic applies to USB devices. Plug a USB 3.0 flash drive into a USB 2.0 port and you'll be stuck at roughly 30 MB/s. That's a tenth of what the drive is capable of.

Controller and NAND differences exist within the same model. SSD manufacturers occasionally swap out controller chips or change flash suppliers between production runs. The packaging looks identical. The model number hasn't changed. But the actual performance can differ. It's not common enough to worry about, but it's worth knowing.

Heat. Laptop SSDs and portable USB drives are especially prone to this. During sustained writes, temperatures climb fast. Once the controller decides things are getting too hot, it deliberately slows the drive down to protect itself. You'll notice this in a speed test as a curve that starts strong and then falls off a cliff halfway through.

What Those Speed Test Items Mean?

Open any benchmarking tool and you'll see a grid of abbreviations: SEQ, RND, Q1, Q8, 4K, 1M. If you don't know what these mean, the results are just noise. Let's break them down.

Sequential (SEQ) means reading or writing data in one continuous stream. Copying a large video file from one folder to another. Installing a game. Importing a project into your editing software. This is the number manufacturers love to put on the box because it's the biggest one.

Random (RANDOM) means jumping around the disk, reading and writing small chunks from scattered locations. Booting your system. Opening an app. Browsing through a folder of photos. Random performance matters more for how fast a computer feels in daily use. Manufacturers don't talk about it much. The numbers aren't as impressive for marketing.

Queue Depth (Q) is how many read/write requests are stacked up waiting to be processed. Q1 means one at a time. Q8 means eight in the queue at once. Drives perform better at higher queue depths because the controller can optimize the order of operations. For normal desktop use, Q1 is what you care about most.

Thread count (T) is how many threads are hitting the drive at the same time. More threads mean more pressure on the drive, which means higher numbers on screen. It's useful for understanding the drive's ceiling, but it doesn't reflect how most people actually use their computers.

Block size changes what the test is measuring. A 4 KB block mimics fragmented, tiny file operations. A 1 MB block mimics large sequential transfers. The difference between the two tells you a lot about what kind of workload a drive handles well.

Units matter too, slightly. MB/s is decimal (1 MB = 1,000,000 bytes). MiB/s is binary (1 MiB = 1,048,576 bytes). The difference is about 4.86%. IOPS measures how many individual I/O operations happen per second. For server and database workloads, IOPS is the more relevant metric. For desktop users, MB/s tells you what you need to know.

Here's the part most people miss. That eye-catching speed number on the manufacturer's spec sheet? It comes from sequential reads, maximum queue depth, biggest block size. Best possible scenario. What actually determines whether your system boots quickly and apps load without lag is something called "random 4K read at Q1, single thread." Knowing this one distinction changes how you look at every benchmark result you'll ever see.

File Test VS. Benchmark Test: Which One Do You Need?

When you test disk speed, most tools give you two options, and they're not interchangeable.

A Benchmark Test talks directly to the disk's physical sectors. No file system in between. This tells you what the hardware itself can do, raw and unfiltered.

A File Test works through the operating system's file system. It creates real files, reads them, writes them. The numbers you get are closer to what you'd experience copying a folder of documents or launching Photoshop. But those numbers are also influenced by whether your partitions are properly aligned, what file system you're using, and how your OS handles caching.

If you're verifying a new drive, run both. Here's why.

A benchmark test that looks great paired with a disappointing file test usually means the hardware is fine. Something in your software configuration is getting in the way. Could be a 4K alignment issue. Could be a BIOS setting. Could be a driver problem. These are all fixable.

If the benchmark test itself looks shaky, with big dips or erratic movement in the speed curve, that's a hardware-level concern. Bad sectors, thermal throttling, or a controller malfunction. Different problem, different solution.

DiskGenius Free Edition has a built-in disk speed test feature that covers both modes. No need to install anything extra. Let me walk you through using it.

How to Test Disk Speed of your SSD or HDD with DiskGenius?

Launch the speed test feature

In the DiskGenius main interface, select a partition on the drive you want to test from the disk list on the left side.

Click "Disk" – "Disk Speed Test" from the menu bar. The speed test dialog will appear.

At the top of the dialog, you can switch between partitions using the "Source Partition" button. The current disk temperature is displayed in real time in the upper-right corner.

Once the target disk is confirmed, you can proceed with the file test and the benchmark test.

Run File Test

Click the "File Test" tab. DiskGenius will automatically detect the storage type (NVMe SSD, SATA SSD, HDD, or USB storage device) and configure default test parameters and test items based on the detected disk type.

In most cases, you don't need to change anything, just click "Start Test" and the software will begin testing immediately.

Want to stress-test the drive and see how it handles heavy workloads? Click "Settings" and crank up the queue depth and thread count.

The results are presented in a table. The most important metric to focus on is RND4KQ1T1 (random 4K read at queue depth 1, thread 1), this figure directly reflects the drive's performance during system startup, application loading, and other high-frequency daily operations.

Run Benchmark Test

Switch to the Benchmark Test tab.

You can choose from three test modes: Read Only, Write Only, or Mixed Read/Write. Select the mode that suits your needs, then click "Start Test". DiskGenius will begin the benchmark test.

During the test, DiskGenius generates a read/write speed curve. The blue curve represents read speed; the red curve represents write speed.

What should you look for?

For a mechanical hard drive, a gradual slope from high to low is expected. The outer tracks of the platter are physically longer and pass under the head faster. That's just physics.

For an SSD, the curve should be mostly flat. If you see it drop sharply in certain spots or wobble up and down throughout, that's a red flag. Could be thermal throttling. Could be degraded NAND cells. Worth investigating.

Check the average speed against what the drive is supposed to do. An NVMe drive reading at 500 MB/s? Almost certainly installed in the wrong slot. Check your motherboard manual.

When the test finishes, save the report. You might not need it today. But six months from now, if things start feeling sluggish, you'll have a baseline to compare against. That's genuinely useful.

Evaluate the Results

Benchmark curve is smooth. File test random 4K looks good. Everything's working properly. Use the drive with confidence.

Benchmark is fine but file test numbers are low. The hardware is healthy. Something in your software setup needs attention. Check 4K alignment, AHCI mode in BIOS, or interface compatibility.

Benchmark curve is all over the place, with big drops. Possible hardware issue. Further diagnostics are needed.

Notes:

Your results will vary depending on the controller, flash type, motherboard, and OS environment.

If your new drive tests a bit below the number on the product page, that's normal. It doesn't mean something is broken.

Your SSD Speed Is Too Low? Here's How to Diagnose It

Things Worth Knowing Before Testing Disk Speed

Use Read Only mode for your first test on a new drive. You get accurate hardware data and there's zero risk of overwriting anything.

If the drive contains lost files that haven't been recovered successfully, please run hard drive data recovery first. That's because disk speed testing involves writing data, which will overwrite lost data. Once that happens, recovery becomes extremely difficult.

Close background applications that access the disk (such as antivirus software, cloud sync clients, and download managers) before testing, to minimize interference with results.

Laptops and USB devices run hot during extended tests. Keep an eye on the temperature reading that DiskGenius shows in the dialog. If it climbs fast, stop the test, let things cool down, and try again with a shorter duration.

Wrapping Up

Whether you want to test disk speed on a freshly purchased NVMe drive or check the speed of an older SSD that's been feeling sluggish lately, the process takes just a few minutes. You're not just collecting numbers for the sake of it. You're confirming that the slot, the protocol, the BIOS settings, and the alignment are all where they should be. If something is off, better to catch it now than after you've moved a terabyte of data onto the drive.