The Corsair MP400 1TB QLC NVMe SSD: A Quick Review
by Billy Tallis on December 11, 2020 2:00 PM ESTWhole-Drive Fill
This test starts with a freshly-erased drive and fills it with 128kB sequential writes at queue depth 32, recording the write speed for each 1GB segment. This test is not representative of any ordinary client/consumer usage pattern, but it does allow us to observe transitions in the drive's behavior as it fills up. This can allow us to estimate the size of any SLC write cache, and get a sense for how much performance remains on the rare occasions where real-world usage keeps writing data after filling the cache.
The Corsair MP400 opts for the largest possible SLC cache size, allowing for a quarter of the drive's advertised capacity to be written to the cache before performance plummets. This is the same strategy used by the Sabrent Rocket Q and probably all other QLC drives using Phison controllers. The Intel and Crucial QLC drives based on Silicon Motion's SM2263 controller have somewhat smaller variable-sized SLC caches, while the Samsung QLC SATA drives use the same small SLC cache sizes as their TLC counterparts.
Average Throughput for last 16 GB | Overall Average Throughput |
Overall drive fill performance for the Corsair MP400 is marginally faster than for the other 1TB QLC drives we've tested, but the DRAMless TLC drives and the larger Sabrent Rocket Q are considerably faster. However, all of the budget NVMe drives are clearly much slower for sustained writes than the mainstream and high-end TLC drives.
Working Set Size
The random read latency from the Corsair MP400 is quite similar to what we saw from the 8TB Sabrent Rocket Q, especially for medium to large working sets. The 1TB MP400 doesn't exhibit the same unsteady performance for small working sets as the 8TB Rocket Q, but instead only shows poor performance for the absolute smallest working set size tested.
The performance drop-off when performing random reads across the entire drive is expected and normal, because the Crucial P1 is the only QLC drive in this bunch to include the full 1GB per 1TB of DRAM. Most low-end DRAMless TLC drives will show an even earlier drop in performance, and many of the more mainstream TLC drives that have switched to a lower DRAM ratio will also show the same drop that the Rocket Q and MP400 show.
75 Comments
View All Comments
GeoffreyA - Tuesday, December 15, 2020 - link
My mistake! Made a blunder here. LBAs written (241) are the host/OS writes. Well, I can't seem to find SMART value 249 (NAND writes). Cunning work by Samsung, not exposing that value.Is where "Wear Leveling Count" the one to use with the P/E cycles? Mine is at 5. Calculating 5 / 2,000 (for TLC, roughly) doesn't seem to yield a sensible value, against the drive's TBW and the OS writes. I'm stumped.
GeoffreyA - Tuesday, December 15, 2020 - link
Another AT article helped here. Seems to be that 1 P/E cycle equals the capacity of the drive in NAND writes, roughly; and the raw Wear Levelling Count equates to used P/E cycles.Roughly: Raw WLC x drive capacity == NAND writes
For me, I get: 5 x 500 == 2,500 GB
I'm scratching my head though because that yields a write amplification that seems erroneously small, only 1.09x (2,500 nand / 2,300 host).
https://www.anandtech.com/show/8239/update-on-sams...
MDD1963 - Sunday, December 13, 2020 - link
'only 200 TBW'...based on my 10 TB per year as a 4 hours per day 'ordinary desktop user/surfer', the drive would be ending it's life in writes at merely 20 years for me... Best to shop elsewhere! :)nucc1 - Wednesday, December 16, 2020 - link
I just looked at the SMART data for my solid state devices, and I'm averaging 0.03 TB per day. over 660 (total power on) days on my oldest SSD. According to Aida64, the drive still has about 95% remaining lifetime. It's a Samsung Sata 850 EVO 500GB.I like to think I'm not an average user, because this machine is running boinc always, and I use these SSDs for running virtual machines whose lifetime is typically less than 90 days each and, I'm also hosting a media server on this machine for in-home streaming.
I used to worry about longevity too, thinking myself to be a heavy user, but the numbers have not borne this suspicion out.
My current OS boot drive, with bitlocker ON is a sabrent 1TB nvme device, currently averaging 0.022 TB written per day over it's 160 day lifetime, and 99% lifetime remaining. If endurance is the only thing that kills these drives, I have many years of worry-free operation left in all these budget, consumer drives packed into my system.
shabby - Friday, December 11, 2020 - link
Go home corsair, you're drunk...You guys should stop accepting these overpriced qlc junk drives for review.
boozed - Friday, December 11, 2020 - link
Why, so we'll never know how they perform?shabby - Friday, December 11, 2020 - link
We don't need to know how they perform, just their price, we know they all perform like shit.Maverick009 - Friday, December 11, 2020 - link
Not exactly true. QLC drives are still a step up from mechanical drives in certain case scenarios, and depending on the pairing of the controller, can even come close or on par with some of their counterparts.Rather have choice with competition rather then your narrow sight of vision as not every drive is built the same.
Gigaplex - Wednesday, December 16, 2020 - link
And then when a QLC drive comes out that's actually good, we'd never know...kpb321 - Friday, December 11, 2020 - link
Price is always where the QLC drives have failed for me. As the article notes it is usually pretty easy to find an 8 channel with dram drive for minimal price premium over one of the QLC drives. Not to mention the other low cost variants with 4 channels and/or dramless that are almost always available at similar price points and while they might have some compromises also compared the better drives they are usually smaller compromises and easier to accept. I just don't see the QLC drive providing the value if there are TLC alternatives available.