Random Read Performance

Our first test of random read performance uses very short bursts of operations issued one at a time with no queuing. The drives are given enough idle time between bursts to yield an overall duty cycle of 20%, so thermal throttling is impossible. Each burst consists of a total of 32MB of 4kB random reads, from a 16GB span of the disk. The total data read is 1GB.

Burst 4kB Random Read (Queue Depth 1)

The Crucial MX500 delivers a remarkable QD1 burst random read performance that is much faster than any SATA drive we've tested, and competitive with many NVMe SSDs, even those using MLC NAND. This is a 60% improvement over the MX300.

Our sustained random read performance is similar to the random read test from our 2015 test suite: queue depths from 1 to 32 are tested, and the average performance and power efficiency across QD1, QD2 and QD4 are reported as the primary scores. Each queue depth is tested for one minute or 32GB of data transferred, whichever is shorter. After each queue depth is tested, the drive is given up to one minute to cool off so that the higher queue depths are unlikely to be affected by accumulated heat build-up. The individual read operations are again 4kB, and cover a 64GB span of the drive.

Sustained 4kB Random Read

The sustained random read speed of the Crucial MX500 when some higher queue depths are involved is merely average, but still a big improvement over the MX300.

Sustained 4kB Random Read (Power Efficiency)

The power efficiency of the Crucial MX500 during random reads is a bit of a disappointment relative to the Intel 545s that uses Intel's similar 64L 3D TLC NAND and a slightly updated Silicon Motion controller. The 545s may have a slight advantage due to our sample being just a 512GB model, but the Crucial MX500's efficiency score is also slightly worse than the drives using Toshiba/SanDisk 64L 3D TLC. Compared to older drives, the efficiency of the MX500 looks great.

The low queue depth performance of the Crucial MX500 is great, and at higher queue depths the Samsung drives and the SanDisk Ultra 3D/WD Blue 3D  are the only ones that have a substantial lead over the MX500. At QD32, the MX500 catches up and is tied for first place with the Samsung drives. Through QD8, the Intel 545s offers very similar performance at lower power.

Random Write Performance

Our test of random write burst performance is structured similarly to the random read burst test, but each burst is only 4MB and the total test length is 128MB. The 4kB random write operations are distributed over a 16GB span of the drive, and the operations are issued one at a time with no queuing.

Burst 4kB Random Write (Queue Depth 1)

The Crucial MX500 sets another record with its burst QD1 random write performance, but the margin isn't quite as wide as for the random read performance. The other mainstream 3D TLC drives form a clear second tier of performance that is about 15% slower.

As with the sustained random read test, our sustained 4kB random write test runs for up to one minute or 32GB per queue depth, covering a 64GB span of the drive and giving the drive up to 1 minute of idle time between queue depths to allow for write caches to be flushed and for the drive to cool down.

Sustained 4kB Random Write

The sustained random write performance of the Crucial MX500 is a bit of a regression compared to the Crucial MX300, but still above average and better than the other 64L 3D TLC drives.

Sustained 4kB Random Write (Power Efficiency)

The Crucial MX500's power efficiency during random writes is second only to the MX300, and substantially better than any other mainstream SATA SSD.

As compared to its predecessor, the MX500's random write performance doesn't increase as quickly with higher queue depths, but the MX300 saturates at QD4 leaving the MX500 to catch up and surpass it beyond QD8.

The Samsung 850 PRO and EVO both saturate at about the same level of performance, but they reach that level with much lower queue depths. They also require far more power across the entire range of queue depths.

AnandTech Storage Bench - Light Sequential Performance
Comments Locked

90 Comments

View All Comments

  • Kristian Vättö - Tuesday, December 19, 2017 - link

    The BX300 is merely a vehicle to consume low yielding MLC wafers.
  • oRAirwolf - Tuesday, December 19, 2017 - link

    Can you expand on that statement? Low yielding in what ways?
  • hojnikb - Tuesday, December 19, 2017 - link

    Probably not as good performing (endurance, write performance) dies get used with bx300 instead of discarding them or using them elsewhere.
  • jabber - Tuesday, December 19, 2017 - link

    But where else would Crucial be using the good MLC ones if everything else is TLC?
  • cblakely - Tuesday, December 19, 2017 - link

    Enterprise products
  • Kristian Vättö - Wednesday, December 20, 2017 - link

    Most of it is going to mobile and industrial applications, of which both have very strict quality requirements. The ones that don't meet the criteria can either be sold as wafer/component or used in a retail client SSD, the latter obviously having a better profit margin.

    Memory is yielded at wafer level. Dies from a low yielding wafer statistically have a higher probability of failure in long-term, even if individual dies are OK in initial probing.
  • sonny73n - Wednesday, December 20, 2017 - link

    WTH are you talking about? Testing endurance and write performance of every die? It's not common sense at all. Therefore, you're a bullshit.
  • FunBunny2 - Wednesday, December 20, 2017 - link

    "WTH are you talking about? Testing endurance and write performance of every die? It's not common sense at all. Therefore, you're a bullshit. "

    that's not what he said. what he said was, post mortem analyses have shown that wafers with a low yield produce dies with short lifespans. those dies are then shipped out as retail SSD.
  • emvonline - Tuesday, December 19, 2017 - link

    Kristian is a wise man. BX300 is a pragmatic response ... not a strategy. MX500 is very well positioned
  • malventano - Tuesday, December 19, 2017 - link

    Low yield MLC wafers would likely not produce usable TLC dies with a higher endurance rating than Samsung's currently shipping V-NAND.

Log in

Don't have an account? Sign up now