Don't be sad! Be glad!
The only thing we can say for
certain is that your SSD is really fast, and that the write speeds you are getting in the (specific) Blackmagic App for an SSD of your size is NOT unusual.
Note that my knowledge here is limited - someone smarter than me could correct me if I make a mistake or further elaborate if they care to.
Different benchmarks of the same SSD can show
very different results (this example shows the difference between the benchmarks of the 256 GB MBP nTB using Blackmagic vs. QuickBench.) Simple things like the size of the SSD (larger sizes = higher write speeds, usually), to the size of the transfer files, to the amount of free space on the SSD, to the programs running in the background, to the (recovery) pauses in test intervals, to whether or not indexing is enabled or not (as the OS may try to index the test file), to the compressibility of the data can have a significant impact on outcome - then there's the technical variables such as the workload's IO intensity and the testing's queue depth, to which my own understanding is limited-at-best.
To my understanding, Blackmagic uses incompressible data. I do not believe it tests across multiple transfer sizes - even though this can hugely affect measured performance:
960 Pro
This carries significant real-world implications given how much one's personal workload could be oriented in one direction (while at the same time all Users [and their Apps/OS] use heterogeneous file sizes, to some extent.) To me, this limits most of the generalizability of the test, and makes it more of a fun novelty than anything else.
You could try running your own benchmarks based on your own usage. For example, I have my own benchmark folder, which contains four different types of file sizes totaling about 30 GB of many very small Excel/Word/Visio documents, many small AACs and JPEGs, some medium-sized datasets/visualization outputs, and several large video files and datasets. Essentially, these mimic some of the files I work with daily, and I can time how long it takes the very small/small/medium/large file sets individually, or as a whole, and compare the results of one SSD to another to determine how this affects my real-world performance. The transfer speeds of the many small files is dramatically slower than the large files (think like 1/100th the performance!)
