You just bought a brand-new NVMe drive. It's supposed to be screaming fast. You fire up the black magic speed test, hit that big start button, and watch the needles jump. But then? The numbers don't match the box. Or worse, the test says you can’t handle 8K video even though you’ve been editing it all morning.
What gives?
Honestly, most people treat this tool like a simple thermometer. Stick it in, get a number, move on. But Blackmagic Design didn't build this to be a general-purpose benchmark like CrystalDiskMark. They built it for filmmakers. If you don't understand how it’s actually hammering your hardware, you're going to make some expensive mistakes when buying storage for your next project.
The Stress File Secret
Most benchmarks write tiny, random bits of data. It’s a "burst" test. The black magic speed test does the opposite. It creates a massive "stress" file—usually 1GB to 5GB—and writes it in a continuous stream.
This is basically a torture test for your drive’s controller.
Cheap SSDs use a trick called SLC caching. They have a small portion of fast memory that acts like a waiting room. When you copy a 2GB file, it flies. But once that cache fills up, the drive's true, slower nature is revealed. Since Blackmagic keeps looping the test, it eventually exhausts that cache. That’s why you might see your speed start at 2500 MB/s and suddenly crater to 600 MB/s after thirty seconds.
It’s not a bug. It’s the truth of how that drive will behave during a three-hour color grading session in DaVinci Resolve.
Decoding the "Will It Work" Grid
Below those two iconic speedometers, there’s a massive checklist. It’s filled with formats like ProRes 422 HQ and Cinema DNG RAW.
People obsess over the MB/s numbers, but the grid is where the real info lives. It’s a binary reality check. If the drive can't sustain the data rate required for 12K 60fps, you get a "X" instead of a checkmark.
Why bit depth matters more than resolution
You might see green checks for 4K 60fps but "X" marks for 1080p 10-bit. That feels broken, right? It's not. 10-bit or 12-bit uncompressed video is a data firehose. A 4K H.265 file (the stuff your iPhone shoots) is actually quite small because it's heavily compressed. Professional formats like Blackmagic RAW or ProRes are massive because they preserve more data.
If you're seeing failures in the "How Fast" panel for 8K, check your connection.
- USB-C (3.1 Gen 2): Maxes out around 900-1000 MB/s.
- Thunderbolt 3/4: Can hit 2800 MB/s but often throttles due to heat.
- WiFi/NAS: If you’re testing a network drive over 5GHz WiFi, don't be shocked when you see 30 MB/s. That’s just physics.
Common Mistakes That Kill Your Scores
Don't test your boot drive while you have Chrome, Spotify, and a video render running in the background. The black magic speed test needs a clear lane.
One weird quirk? On modern macOS versions (like Sonoma or Sequoia), the app sometimes defaults to "Read Only" on your system drive. You’ll hit start and... nothing. You have to go into the settings (the gear icon), click "Select Target Drive," and choose a specific folder like Downloads to give it write permissions.
Also, watch the heat.
External SSDs like the Samsung T7 or SanDisk Extreme Pro get incredibly hot during a sustained black magic speed test. When they get hot, they throttle. If you leave the test looping for ten minutes, you aren't testing the drive's peak speed; you're testing its cooling efficiency. For real-world editing, that's actually a more useful number anyway.
💡 You might also like: Apple Store Genius Bar: Why Your Tech Support Experience Is Changing
Blackmagic vs. The Others
Is it better than AJA System Test? Sorta.
AJA lets you customize the file size up to 64GB, which is better for testing massive RAID arrays. CrystalDiskMark (Windows) gives you "Random 4K" scores, which tell you how fast your computer will feel when opening apps.
But for video? Stick with Blackmagic. It’s the industry standard because it mimics the specific way a camera or an editor writes data. It’s "Sequential" testing on steroids.
Actionable Steps for Better Storage
Stop looking at the peak speeds on the box. They are marketing fluff based on "ideal conditions" that don't exist in a hot edit suite.
- Run the 5GB Stress Test: Don't settle for the 1GB default. Larger files force the drive to work harder and show its true sustained speed.
- Loop for 5 Minutes: If the speed stays consistent, you’ve got a "pro" grade drive. If it drops by 50%, that drive is for backups, not active editing.
- Format Corrects: Ensure your drive is formatted as APFS (for Mac) or exFAT (if sharing with Windows). Using an old HFS+ format on a modern NVMe can actually shave a few percentage points off your throughput.
- Check the Cable: Seriously. Half the "slow drive" complaints on forums are just people using a charging cable instead of a high-speed data cable. Look for the lightning bolt logo or the "20/40Gbps" mark.
If the black magic speed test shows your drive is failing the 4K ProRes 422 HQ checkmark, it’s time to stop editing off that old mechanical "spinning" drive. Your CPU might be fast, but it’s starving for data. Upgrade to an NVMe enclosure with a Thunderbolt 4 bridge chip if you want those needles to actually stay in the green.