Quickly spot-check any USB mass storage drive
for fraudulent deliberately missing storage.

After the ValiDrive's testing has completed and its user has closed its map display, the introductory contents of ValiDrive's main dialog is replaced with a report providing additional information about the drive tested and ValiDrive's interaction with the drive:


The various items of the drive's report are detailed below:



The ValiDrive performance details of two different thumb drives are shown above. The SanDisk 512GB drive is a classic tiny tumb drive, but it performs nicely. The total time required to run ValiDrive's test was 12.24 seconds (the sum of the total read and write times). By comparison, the 32GB "Micro Center" drives (no-name thumb drives purchased from Amazon in a 10-pack for ~$30 or $3 each) required a whopping 736.37 seconds or 12.27 minutes! – in other words about 60 times longer! The charts above reveal the reason for the massive discrepancy in performance:

Looking at the total time spent reading we see that the inexpensive 32GB drives significantly outperformed the much larger and more expensive SanDisk drive by more than 2 to 1. The inexpensive drive's average read time was 1.665 milliseconds whereas the average for the 512GB SanDisk drive was 4.467 milliseconds. However... writing is an entirely different story!

The $3 drive was nearly 100 times slower (93.95 times) to write than the larger 512GB SanDisk drive.

This massive discrepancy in apparent write performance is due to the fact that ValiDrive switches back and forth between reading and writing – transferring only 4k bytes of data each time – and less expensive drives require far more time to begin writing than more expensive and more sophisticated drives.

Writing to NAND flash memory requires the generation of high voltage, but reading requires low voltage. Before any NAND chip can write data into NAND flash memory it must internally generate a high voltage. This uses a “voltage pump” which takes time to charge up, and lower-end solid state storage does this much more slowly. Then, before reading, that high voltage must be dumped.

Since operating systems typically write entire files and even multiple files at once, a single “voltage charging” cycle can be amortized across the entire large write. So large writes are able to proceed at a reasonable pace. But since ValiDrive only writes 4k bytes before reading it back, causing the “writing charge” to be dumped before each read, ValiDrive accentuates this delay for drives which require a long write-cycle.

The potential importance of “Variance”

For any form of random access mass storage, the “Variance” in read and write timing can be an important indication of the general health of the storage system. We would expect healthy drives to generally spend the same amount of time reading or writing fixed amounts of data, independent of the data's location, whereas drives where something is wrong, very wrong, or beginning to go wrong, would be expected to exhibit more varying read and write performance. This is why ValiDrive introduces this potentially important concept of Variance.

As noted above, “standard deviation” is a measure of the dispersion of values within a sample set. Looking at the standard deviations of the write times of the two drives above, we see 5,605 for the fast-write cycle drive and a whopping 676,662 for the ultra-slow write-cycle drive. This would appear to indicate that the slower drive's write times have a massive dispersion. But that's not actually true because the standard deviation fails to take the average value into consideration.

The average write time for the faster write-cycle drive is 6,858 milliseconds and the average write time for the slower write-cycle drive is 637,543. So, for the standard deviation to have any diagnostic value it must be taken within the context of the average time. This is why ValiDrive calculates the Variance as the standard deviation divided by the average.

Now, when we compare the two drives' Variance we have 0.817 for the faster write-cycling drive and 1.061 for the slower write-cycling drive. So, even though their non-contextual raw standard deviations differ by a factor of 93!, their contextualized variance only differs by a factor of 0.77.

In other words, even though the inexpensive $3, 32GB thumb drive is exhibiting an extremely slow write-cycle under ValiDrive, that write-cycle timing is not varying very much, which suggests that all is well inside that drive... even though it takes a while to get going at the start of any writing activity. And it certainly does read quickly... more than twice as fast as the larger more expensive 512GB drive.


Jump to top of page
Gibson Research Corporation is owned and operated by Steve Gibson.  The contents
of this page are Copyright (c) 2023 Gibson Research Corporation. SpinRite, ShieldsUP,
NanoProbe, and any other indicated trademarks are registered trademarks of Gibson
Research Corporation, Laguna Hills, CA, USA. GRC's web and customer privacy policy.
Jump to top of page

Last Edit: Oct 09, 2023 at 06:25 (57.58 days ago)Viewed 68 times per day