Suggestions: Statistics toggle and reduced TRIM interval

Hello.
Your RAM disk is getting better and better—that's great to see, and I'd like to help make it even better. I am writing not only on my own behalf, but on behalf of a large group of people. I'd like to suggest two changes:

1. Add the option to disable all statistics, such as a checkbox in the GUI.
The goal is to minimize memory usage and disk operations as much as possible. Most people have absolutely no use for these statistics.

2. Reduce the time until TRIM is triggered after files are deleted from the RAM disk, to speed up memory release.
According to our measurements, it's 70 seconds. That's too long! Other RAM disks, such as Primo or Ultra RAMDisk, clear their cache in 10–15 seconds. 15 seconds is still a bit too long. 10 seconds would be ideal.
You could add an option in the GUI or at least in the registry to specify the cleanup interval, or simply make it run faster all the time, or, for example, reduce the interval when the disk is more than 50% full. This is necessary to avoid a BSOD when the disk occupies all of the RAM. This happens when the disk size has to be set to almost the full amount of RAM. So you can delete unnecessary files in time in situations like these.

We've addressed both in the latest version (26.7), just a bit differently than proposed.

Faster memory release: you were right that ~70 s was too long. Rather than a setting, we simply made it faster across the board: deleted-file memory on NTFS now clears in about 10–15 seconds. exFAT takes a little longer by design (no native TRIM, so we scan the allocation bitmap with confirmation passes).

BSOD risk: already covered. When free RAM drops to ~1 GB, the driver stops allocating and returns an "out of space" error instead of crashing, so you get a clean failure and time to delete files. The faster reclaim widens that margin further.

Statistics: rather than a full off switch, we reduced their cost directly. The counters use a tiny fixed amount of memory, and 26.7 writes them out less often. We've noted the request for a complete on/off option for a future update.

Hello.
I was pleasantly surprised that you not only agreed to do it, but also finished updating the program right away.

I tested TRIM, and now the cleanup speed is acceptable at 15–17 seconds!

However, during additional tests on version 26.7, I found a bug that occurs consistently every time. When copying files to the disk, if there is about 280 MB of free space left—sometimes even more—the copying process freezes. In other words, it's impossible to fill the disk completely. For example, if there's 1.2 GB of free space left on the disk, copying a 1 GB file to the disk will cause the process to freeze. In other programs for creating RAM disks, the disk fills up all the way to the end without any problems, leaving about 1 MB of free space. I recorded a screen video showing an example of this bug, if you need it.

Thanks for the video. It's very clear.

What you're seeing is a deliberate safeguard we put in place to prevent the system from running out of memory and crashing. This behaviour is documented in dynamic allocation:

Quote

Memory pressure: To maintain system stability, the driver will stop allocating memory when 1 GB or less of RAM remains available. If this occurs, I/O operations to the RAM disk will appear stalled until memory is freed. You can resolve this by closing other applications to free up memory, or by deleting files from the RAM disk. In the latter case, NTFS handles this scenario better than exFAT due to its more immediate TRIM processing.

If you don't like this behaviour, you can use static allocation instead. In that mode there is no safeguard, but it also means the system can run out of memory, which may lead to instability or a crash. For that reason, we don't recommend using it unless you understand the risks.

Got it. I missed that part in the description.

I tested this "protection" feature, and overall I think it's a good option. It works reliably within the 900-950 MB range, and the copy process resumes once disk space is freed up. It's exactly as you describe.

Thank you.