How do Raid-0 setups appear inside harddrivebenchmark ?

[cnd]

I'm wondering what happens with my disk performance results when I've engaged hardware raid-0 ? I assume the software is smart enough to not report the improved performance as if it was a single drive ? Do my HDD results just get discarded in that case?

I'm planning to install two Firecuda's into my new ASUS laptop in raid-0 - here is the only benchmark posted so far for this exact machine I'm getting: https://www.passmark.com/baselines/V...d=156482135855
The existing PM9A1 NVMe drive is already nearly half the performance of a Firecuda, so in raid-0 it should be about 300% faster...

Is there any way to search uploaded baselines to see if anyone has already posted a disk rating of (say) over 75,000+ ?

[David (PassMark)]

In most cases the RAID controllers block the collection of the drive model number so they are going to be hard to search for. In many cases they appear to the operating system as a "Intel RAID 0 Volume" or "AMD-RAID Array" or "Unknown Drive" and that is what is reported in PerformanceTest. But in this case you don't know what drive models make up the RAID set (and this is kind of the point of RAID as well, as it combines many drives into one).

RAID controllers also tend to increase latency due to the extra layer of software / electronics. So overall performance with already good SSDs doesn't improve nearly as much as you would expect. Some use cases might even be worse.

[cnd]

Thanks for the explanation! I wonder if there's an SPI or I3C/I2C interface that your software might be able to query, like the ones most modern memory modules support? I notice in the spec that there is mention of "I2C" in some pinouts (and others have interesting pins like "SIM DETECT", and the spec also cover SPI)... all those would be too slow to be useful for SSD data these days, so presumably they'd have to exist to report info about the module installed ?

[frazzlesnap]

I've been running 4 2TB FireCuda 530s in a RAID10 for about a year now, and I am very happy with the results. The Passmark scores are virtually identical to a single 530, but under some loads, it improves, sometimes dramatically. Latency, which I expected to be worse, seems to be the same. I have some screencaps around here somewhere that show ridiculous Crystal Disk scores in decidedly unrealistic use scenarios. I wouldn't recommend a RAID0. The performance benefits will be hard to justify if one of your drives craps out. I do not like the built-in AMD version of RAID support, however. Maybe the Intel version is better? I sure hope so!