Using RAM FS for trading software

I am a day trader and have been using your RAM Disk for many years. I'm extremely happy with it.

Installed on one RAM disk are two trading programs, and the client application from my live data provider. This RAM disk makes a copy to my SSD when I shut down and reboot.

I have used Crystal disc to test the speed of my SSD and my RAM disk. The RAM disk is about 2 to 5 times faster. I can see the difference in my trading programs, and the downside risk is essentially zero: if I lose power, I can return to my starting point and link to live data again when the power returns.

In this situation would the speed of these operations be increased by switching to RAM FS?

I use:
Investor/RT trading program
DTN IQ feed client to link streaming live data for Investor/RT
Ninja Trader 7 trading program streaming live data from my broker and used to place live trades

RAM File System is significantly faster than RAM Disk in some types of operations, for example when copying large files. In situations where data comes from external sources via the internet, the limiting factor is usually the speed of the internet connection rather than the speed or RAM Disk / RAM FS.

To find out with certainty whether RAM FS can give you a noticeable increase in performance with your particular setup, you can test it in situ: it offers a free 30-day trial and can be installed alongside RAM Disk.

You may also find it helpful to have a look at RAM Disk vs RAM FS comparison.

One thing worth noting for the nearest future: as RAM Disk is a much older product, it went through years of testing and improvements. So currently is offers greater reliability, while RAM FS is very new and we are still ironing out some edge cases.

Thank you for your prompt and thorough reply.

I have linked operations between my two trading programs with extensive code that I wrote to export and import files from drive letter "S", my RAM Disk, so to test RAM File System I would need to create a virtual drive using letter "S". It seems unlikely that I could make another drive on Windows using RAM File System and designate it as drive "S", while leaving my RAM Disk in place that is also designated as "S". I just don't think Windows could or should let me do that.

Assuming I'm correct about this part, I would make a clone of my current SSD (using Macrolium Reflect) and then delete my RAM Disk letter "S" on the clone SSD. Then create a new virtual drive using RAM FS, designate this drive with letter "S" and then install my trading programs and run some tests.

Let me know if I'm correct about my assumptions for testing? Or is there is a better way?

Thanks again for the great tech support. I don't write in very often, but I'm always impressed when I do.

Appreciation,
Dan

Your assumption is correct: Windows will not let two volumes hold the same letter at the same time. But cloning your SSD is not necessary - the swap is simple and fully reversible, which makes it perfect for testing:

1. In RAM Disk, dismount or delete the S volume. Its image file on your SSD stays untouched, so nothing will be lost.
2. In RAM FS, create a volume and assign it the letter S.
3. Copy your files onto it and run your tests.

To go back, delete the RAM FS volume and re-create the RAM Disk volume on S as before. Nothing on your SSD is modified along the way. The two products install side by side without conflict; it is only the drive letter they cannot share.

One thing to be aware of: RAM FS can also save its contents to an image at shutdown and restore them at boot, so your current workflow will be preserved, but the image format is not compatible with RAM Disk (because they have different file systems). So before switching, copy the contents of your RAM Disk volume to a temporary folder on your SSD, then copy them back once the RAM FS volume is ready. As the paths stay the same (S:...), your programs should keep working without reinstalling. But if any of them misbehaves, reinstalling on the new volume is a clean fallback.

A good basic test option, if you wish: install RAM FS, create a small volume on a spare letter such as Q, and run Crystal Disk Mark against both Q and your current S. This will give you the basic comparison on your own hardware in a few minutes without disturbing anything. If the difference is not compelling for your workload, you can stop there. And if you see a significant improvement, then you will be making a well-grounded decision for investing more time into running the tests described above.

Finally, as this is a live trading setup: RAM FS is a new product and we are still working on its robustness, so please test outside trading hours and ideally not on the machine you rely on during a session.

Thank you for the compliments about our support. We are happy to hear that our efforts make a difference.

Excellent advice! I was thinking that I might be able to dismount the RAM disk, but I haven't done that before, so I wasn't sure whether it was a good idea or not.

And yes, I am aware of the risk involved with day trading on short time frames, and we'll make sure that everything is operating correctly before I use any newly installed programs live.

Once again, I appreciate the great tech support!

Thank you,
Dan