Speed is deceptive. You buy a shiny new NVMe SSD, the box promises 7,000 MB/s, and you feel like a god of productivity until you actually drop a 400GB 8K ProRes raw folder onto the timeline. Then, everything chugs. This is exactly why AJA System Test Lite exists. It isn't just another benchmark tool designed to give you a "high score" to brag about on Reddit. It’s a reality check for video professionals, DITs, and anyone who realizes that "burst speed" is basically a marketing scam when it comes to sustained video workflows.
Most people download generic disk utilities. They see a big number and walk away happy. But video isn't a single file transfer; it's a constant, unrelenting stream of data that demands consistency over raw peak performance. If your drive pauses for even a millisecond to catch its breath (thermal throttling, anyone?), your frames drop. Your render fails. You lose money.
What AJA System Test Lite Actually Does (And Why It’s Free)
AJA Video Systems is a heavy hitter in the broadcast world. They make the Kona cards and the Cion cameras. They built AJA System Test Lite because they got tired of support calls from editors wondering why their expensive RAID arrays were dropping frames. It’s a stripped-down, lightweight version of their internal diagnostic tools, designed to simulate specific video resolutions and codecs.
The "Lite" version, which you usually find on the Mac App Store or as part of their larger driver packages, is remarkably simple. You pick a file size—maybe 1GB if you’re testing a thumb drive, or 64GB if you’re testing a serious work disk—and you hit start.
It’s about simulation.
Unlike CrystalDiskMark, which feels like a lab experiment, AJA lets you select a "Video Frame Size" like 4K-UHD and a "Codec" like 10-bit RGB. This is huge. It changes how the data is packaged during the test. Writing a single massive chunk of data is easy for a computer; writing thousands of individual frames in a specific format is what actually kills your performance.
The Secret Language of the Sweep Graph
The most important part of the app isn't the big round dial showing MB/s. It’s the graph tab. Honestly, if you aren't looking at the graph, you’re missing the point of the software.
A "fast" drive that shows a jagged, saw-tooth pattern on the graph is actually a terrible drive for video editing. You want a flat line. If you see massive dips in the write speed halfway through a 16GB test, that’s your drive's SLC cache filling up. Once that cache is full, the drive drops to its "real" speed, which is often a fraction of what was advertised.
I’ve seen "Gen 4" drives that claim 5,000 MB/s sustained, but after ten seconds of the AJA System Test Lite stress, they crater to 600 MB/s. That’s the difference between smooth 8K playback and a stuttering mess that makes you want to throw your monitor out the window.
Codecs Matter More Than You Think
When you toggle between ProRes 422 HQ and 10-bit YUV in the settings, you aren't just changing a label. You’re changing the workload.
- Lower Bit Depth: Easier on the controller, focuses on pure throughput.
- High Bit Depth (12-bit RGB): Forces the system to handle much denser data packets.
- Resolution Scaling: Testing at SD (Standard Definition) vs 8K Full reveals if your bottleneck is the drive's connection (Thunderbolt/USB-C) or the flash memory chips themselves.
Most people leave it on the default settings. Don't do that. If you’re a YouTuber shooting 4K 10-bit, set the tool to 4K and 10-bit. Test the reality you live in.
Common Misconceptions About the Results
A lot of users get frustrated because AJA System Test Lite shows lower numbers than other benchmarks. They think the app is broken. It’s not.
Most benchmarks use "compressible data." This is basically cheating. It sends strings of zeros that the drive controller can compress instantly, making the speed look way higher than it is. AJA uses non-compressible data because video files—especially once they've been compressed by a camera—don't have much "room" left to be shrunk. What you see in AJA is the "honest" speed.
Also, check your "Disk Write Cache" settings in your OS. If your OS is caching the test, AJA might report speeds of 12,000 MB/s, which is physically impossible for the hardware. This usually means you’re just measuring your RAM speed, not your disk. Turn off "system caching" in the AJA settings to get the truth.
🔗 Read more: Sex bots for women: Why the conversation is finally changing
Thunderbolt 3 vs. USB-C: The Great Bottleneck
I get asked all the time why an external SSD isn't hitting the numbers on the box. Usually, it's the cable. Or the port.
If you run AJA System Test Lite on a drive plugged into a standard USB 3.0 port, you'll cap out around 450 MB/s. Jump to USB 3.2 Gen 2, and you might see 900 MB/s. But even with Thunderbolt 4, you'll rarely see the "theoretical max" because of overhead. AJA accounts for this overhead in a way that feels much more "real world" than a command-line interface test.
How to Run a Proper Stress Test
If you want to know if your system is actually ready for a feature-length project, don't just run a 1GB test. That’s like sprinting five feet and saying you’re ready for a marathon.
- Set the file size to at least 16GB, or ideally 64GB.
- Let the test run through multiple cycles.
- Watch the temperature of your drive.
- Look for the "Disk Full" performance drop.
Some drives perform great when they're empty but lose 50% of their speed once they're half full. You can test this by filling your drive with junk files and then running AJA System Test Lite again. You might be surprised—and horrified—at the results.
Precision vs. Accuracy
There is a nuance here that tech reviewers often skip. Precision is getting the same result every time. Accuracy is getting the "correct" result. AJA System Test Lite is highly precise, but its accuracy depends entirely on your configuration. If you select a network drive (NAS) over a 10GbE connection, you have to account for network latency. AJA handles this better than most, but it’s still a variable.
For those using Blackmagic Disk Speed Test (the main competitor), you'll notice AJA gives you more granular control over the type of data. Blackmagic is a "one-click" wonder. AJA is a tool for people who want to tweak the parameters until they find exactly where their system breaks.
Why "Lite" is Often Better Than the Pro Version
The full AJA System Test is part of a larger install, but the Lite version is standalone. It’s clean. It doesn't leave junk in your library folders. For 99% of editors, the Lite version provides every single metric needed to diagnose a slow timeline. You get the MB/s, you get the Frames Per Second (FPS) calculation, and you get the reliability of a company that actually builds the hardware used in Hollywood.
Actionable Steps for Your Next Benchmark
To get the most out of your testing, follow this sequence:
👉 See also: East Coast Doppler Radar: Why the Forecast Still Misses Those Massive Coastal Storms
- Restart your machine first. This clears out any active RAM caching that might skew the initial write bursts.
- Disconnect other peripherals. If you have three webcams and an audio interface on the same USB bus, your bandwidth will be split.
- Run the "System Report." In AJA, you can export your results. Do this when the drive is new. Save it. Six months later, run the test again. If the speed has dropped significantly, your drive might be failing or needs a "TRIM" command to clean up deleted data blocks.
- Test the target resolution. If you plan on moving to 6K next year, test 6K now. Don't assume your current setup will handle the jump just because 4K works.
- Focus on Write speeds for Capture, Read speeds for Playback. If you’re recording directly to a drive (like from a Blackmagic 6K or a Ninja V), the Write speed is the only number that matters. If you're just editing, the Read speed determines how many streams of video you can have in your multicam clip.
Stop guessing. Stop trusting the sticker on the box. Use the tool to find your actual ceiling before you’re on a deadline and the "Dropped Frames" warning starts flashing red.
The reality of your hardware is usually a bit slower than the dream, but knowing exactly how much slower allows you to build a workflow that actually works. Over-provisioning your storage based on these tests is the cheapest insurance policy you can buy for your creative sanity.