So, let’s start with the first part of the benchmarks and the feel-good mode for marketing. The synthetics are a good opportunity to really put the big numbers to the test. We’ll see later how well this works in reality with the real application benchmarks. That’s why I’m starting with CrystalDiskMark and four different file sizes. The SSDs were no longer new at the time of testing and I had already had fill levels of more than 50% before deleting the data several times. The good thing is that the maximum values can still be safely reached even after several tests and impressive figures can be achieved.
You can see very clearly that the dynamic pSLC does exactly what it is supposed to do, mind you with an empty (albeit not virgin) SSD. The nice thing about the 2 TB SSD is that there is plenty of space and it is therefore better never to fill it more than 2/3 full with data. A higher load does not affect reading, but the dynamic SLC will certainly reach its limits when writing. And if you do it over and over again, it will eventually no longer be possible to switch the memory modules between the two methods.
ATTO is very interesting this time, although I only work with two sizes here, which ultimately amounts to the same thing. Even if the figures are already somewhat smaller, although the cache still plays a role, synthetic benchmarks of this kind should not be taken too seriously, especially if the firmware has perhaps also been optimized for such benchmarks. But I didn’t want to withhold this result from you either.
Video streams
But what happens when you stream a larger video? The industry uses the AJA benchmark for this, which is effectively an interface between synthetic benchmarks and practical applications. Here, too, the Corsair MP600 Elite unfortunately fails brutally due to the trimmed cache, especially as it already deviates somewhat from the theoretical write and read rate with the cache. We can see that the comments made on the previous page about the dynamic pSLC cache and the behavior with larger file blocks are completely true. At just under 48 GB, the cache virtually collapses.
When reading, the compact stream also does not achieve the promised performance, but is almost below the cached write performance.
The measurement log shows quite clearly that this SSD is definitely not suitable for streaming large amounts of data:
BENCHMARK-TABLE
11 Antworten
Kommentar
Lade neue Kommentare
Urgestein
Veteran
Urgestein
Veteran
Urgestein
Urgestein
Urgestein
Urgestein
Urgestein
Urgestein
Urgestein
Alle Kommentare lesen unter igor´sLAB Community →