AS SSD Copy-Benchmark Explained: ISO, Program and Game

The main AS SSD Benchmark window measures synthetic speed: sequential, random 4K, and access time. The Copy-Benchmark answers a more everyday question: how fast does this drive copy real collections of files? It lives under the Tools menu, it is not part of the main Start button score, and it tests three familiar patterns named ISO, Program, and Game. This guide explains what each pattern contains, how to run the test, and how to read the two numbers it gives you for each scenario without confusing them with the main score.

You need version 2.0.7316.34247 from the download page extracted as a complete portable folder. If you have not run the main benchmark yet, do that first using how to use AS SSD Benchmark, because the copy results make more sense next to your Seq and 4K rows than they do on their own.

How to open and run the Copy-Benchmark

Open AS SSD Benchmark, select the drive you want to test at the top of the main window, then open Tools and choose Copy-Benchmark. A separate window appears with the three scenarios and a Start button. The test creates folders of test files on the selected drive and copies them using operating system cache functions, timing how long each copy takes and reporting both duration and speed.

Make sure the selected drive has a few gigabytes of free space and that no other heavy copy or download is running. Because this test copies within the same drive, it exercises reading and writing at the same time on that one device. That is different from copying from one SSD to another in daily use, where the slower of the two drives and the connection between them can limit the result. Keep that in mind when you compare copy numbers with real file transfers between drives.

The test takes longer than a glance at the window suggests, especially the Program scenario with its many small files. Let it finish all three rows. Stopping halfway leaves you with the least interesting part of the picture, because the contrast between the three patterns is the whole point of the tool.

ISO: two large files

The ISO scenario copies two large image files. Large, continuous files are the friendliest workload for any storage drive: the data flows in long sequential streams, there is little per-file overhead, and the drive can work close to its sequential capability. Expect ISO to show the highest copy speed of the three scenarios on almost every drive.

Read the ISO row as your best-case copy speed on this drive. It is relevant for moving disk images, large archives, virtual machine files, and big video files around on the same drive. If even the ISO copy is slow, look for a general problem: a nearly full drive, a slow connection for an external drive, heavy background activity, or a drive that slows sharply during sustained writes. Compare it with the Seq row from the main window; they will not match exactly, because one is a synthetic measurement and the other is a timed copy through the operating system, but they should tell a similar story.

Program: many small files

The Program scenario copies a typical program folder: a mix that includes many small files alongside larger ones. Every file carries overhead: it must be opened, created at the destination, written, and closed, and the file system must update its records for each one. Thousands of small operations turn a copy into a random, stop-start workload rather than a smooth stream.

Expect Program to be the slowest of the three, often dramatically slower than ISO on the same drive. That drop is normal and it is valuable information. It mirrors installing software, copying a development project with thousands of small files, or moving a photo library of small files. A drive with strong 4K random performance in the main window usually holds up better here than a drive that only shines in sequential speed. If Program is painfully slow while ISO looks fine, your bottleneck is small-file handling, not raw bandwidth.

Game: a mixed folder

The Game scenario copies a folder modelled on a game installation: a mixture of large data files and smaller files. It sits between the other two in character and usually in speed as well, closer to ISO when the large files dominate and dragged down by the small files mixed in with them.

For many users Game is the most realistic single row. Modern applications, game libraries, and project folders are rarely one giant file or thousands of tiny ones; they are a blend. When someone asks how fast this drive copies in real life, the Game row, read together with Program, is a more honest answer than the headline sequential number from the main window.

How to read the results table

Each scenario reports a time and a speed in MB/s. Use the speed to compare scenarios and drives, and the time to sanity-check the experience: it is literally how long that copy took. Read the three rows as a profile, not as three separate scores to average. A healthy profile usually slopes downward from ISO through Game to Program. A profile that is flat and low everywhere points to a general limit. A profile with a strong ISO and a collapsed Program points to small-file weakness.

• ISO high, Game mid, Program low: the normal pattern; the spread shows the cost of small files.
• All three low: check free space, background load, connection type for external drives, and sustained-write behaviour.
• Program unusually close to ISO: a strong result; small-file handling is a strength of this setup.
• Results vary wildly between runs: something else was using the drive; re-run on a quiet system before concluding anything.

The Copy-Benchmark produces no total score and does not change your main AS SSD score. That is a feature. It keeps a practical, file-based measurement separate from the synthetic score so you can use each for what it is good at. Save the copy window alongside your main screenshot when you document a drive, especially if your real workload is file-heavy.

Copy-Benchmark vs the Compression-Benchmark

The other Tools menu test, the Compression-Benchmark, varies the data itself, testing how speed changes as data becomes more or less compressible. The Copy-Benchmark instead varies the shape of the workload: large files, small files, and a mix. Drives can look different under each lens. Reading both gives a fuller picture, so continue with the Compression-Benchmark explained when you are ready.

Used well, the Copy-Benchmark stops an argument that synthetic numbers often start. Instead of debating whether a drive is fast in the abstract, you can point to how long it took to copy a program-like folder on your own system. That is slower to run than the main test, but for upgrades, external drives, and troubleshooting slow file operations, it is often the more persuasive evidence.