RAMdisk in virtual memory

Just wondering: isn't the creation of a RAMdisk in virtual memory, as opposed to physical memory, a contradiction in terms?

In the end, both virtual and physical memory end up being the physical RAM. Virtual memory is simply an abstraction that represents a contiguous address space, see this article in Wikipedia for a more detailed explanation.

In the Windows kernel world we can choose whether to use virtual memory or physical memory. The former is easier to use, the latter is faster due to the absence of abstraction.

In our RAM Disk app we use the physical memory with a number of optimisations. In the screenshot below you can see how it outperforms ImDisk in both virtual and physical configurations:

Sorry for being off-topic with this, but one of your earlier screenshots had the "SoftPerfect Rapid Ram Disk" header. Is this from an upcoming RAM Disk release? If yes, do you have an estimated time frame for release?

I am always excited to see this particular software product develop and move forward. I believe that your team is still the only RAM Disk software dev team which has their kernel-mode drivers cross-signed by Microsoft Windows (sha256) which is really important for all of the latest Windows 10 releases. So myself and many other paid users of your software really appreciate the work that your team does. Thank you.

The 'Rapid' isn't a new version I am afraid, but rather a custom build for our Japanese partners who market it in Japan under a different name.

We consider the RAM Disk product to be feature-complete, i.e. it does the job and does it well. Not much else can be added to it feature-wise. Having said that, we are open to suggestions on the product improvements. One of the things still planned to be implemented is compressed or dynamic images, which would offer more compact image files and possibly faster loading from them.

By the way, regarding the Microsoft's signature: it is very difficult to get. There is an easy way called "attestation" where a developer just uploads their driver to Microsoft and they cross-sign it. The problem is, the easy-way-signature doesn't work on Windows Sever platforms based on NT 10. To qualify for a driver signature, Microsoft requires a series of tests known as HLK testing to succeed. The full set of the tests takes nearly 24 hours and should one fail, the developer has to start over. While going through that process we have discovered and fixed a few issues with our drivers and now they should be rock-stable on any platform.

Thank you for the clarification regarding the different naming of the program for the Japanese market, Andrew. That eases my curiosity in that regard.

Also, I agree on the "feature complete" statement for RAM Disk. I personally could not think of any additional features required at this time and the stability, performance and efficiency are fantastic.

Thank you for the additional details regarding Microsoft's more stringent signing requirements for kernel-mode drivers as well, very interesting stuff. I am thankful that your team has "gone the extra mile" here to go above and beyond for the digital signing.

Again, sorry for chiming in on a thread that is not my own. I appreciate your time and I will drop my end of this conversation now since you covered my questions and curiosities thoroughly.

@WildByDesign: not need to apologise for asking questions, especially the relevant ones. That's what this forum is for

Also, we are very glad that you are happy with SoftPerfect RAM Disk. We pride ourselves on doing our best to create hight-quality, reliable software; and it is always nice to hear that our users appreciate it.