The Western Digital WD Black SN750 SSD Review: Why Fix What Isn't Broken?
by Billy Tallis on January 18, 2019 8:01 AM ESTRandom 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.
The Toshiba/SanDisk 3D NAND has consistently shown worse burst random read performance than its competition, and the WD Black SN750 does nothing to change that pattern. The SN750 is insignificantly faster than last year's model, but both the Plextor M9Pe and Corsair MP510 get better read latency out of this NAND by pairing it with different controllers.
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.
On the longer random read test that also brings in some higher queue depths, the WD Black SN750 is still quite a bit slower than the best combinations of Silicon Motion controller and Micron NAND, but at least this time the drives that use the same Toshiba/SanDisk BiCS3 NAND with different controllers aren't so far ahead of the WD Black.
 |
|||||||||
| Power Efficiency in MB/s/W | Average Power in W | ||||||||
The WD Black SN750 is just a touch more efficient than its predecessor on the random read test, allowing it to retake a small lead among TLC-based drives here rather than being in a tie with the Toshiba XG6.
 |
|||||||||
When the WD Black SN750 is pitted against most of last year's competition, it does at least catch up if not pull ahead in random read performance at higher queue depths, while never requiring as much power. But the Silicon Motion SM2262EN drives are faster at all queue depths, though not in proportion to how much more power they require.
At low queue depths the random read performance of the WD Black SN750 is down in SATA performance territory where some drives can beat it on performance and power consumption. But at higher queue depths, it is obviously the most efficient NAND-based SSD we've tested for random reads.
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.
The WD Black SN750 shows a small regression in burst random write performance compared to last year's model. The Corsair MP510 and Silicon Motion SM2262EN engineering sample are nearly tied for first place, with the WD Blacks 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.
On the longer random write test that brings in higher queue depths, the WD Black SN750 is roughly tied with the SM2262EN sample for first place and is a few percent faster than competitors like the Samsung 970 EVO and Corsair MP510.
 |
|||||||||
| Power Efficiency in MB/s/W | Average Power in W | ||||||||
The SN750 is the most power efficient NVMe drive on our random write test, pulling slightly ahead of the Toshiba XG6 and the older WD Black, and maintaining a substantial lead over most other high-end NVMe drives.
 |
|||||||||
The WD Black SN750 hits full speed for random writes with a queue depth of 4 or higher, and maintains steady performance for the rest of the test while drawing just over 3 W.
The WD Black SN750 is very close to the being the most efficient flash-based SSD for random writes that we've ever tested, but there are a few data points in the archive that are slightly faster at similar power levels.
Facebook
Twitter
RSS
54 Comments
View All Comments
namechamps - Friday, January 18, 2019 - link
Strangely enough I don't think anandtech has ever reviewed the 970 Pro which is likely why it isn't in the comparison. They have done the 960 pro and the 970 evo but not the 970 pro.Ryan Smith - Friday, January 18, 2019 - link
"Whatever happend to the Samsung PRO in those comparisons?"Unfortunately Samsung never sampled the 970 PRO, so we don't have it on hand. And all indications are that they're just about done with it, having never released a 2TB version (like they did the 960 PRO).
I'm really not sure if we're going to see any new consumer MLC drives in 2019. The market has bifurcated into TLC and then more boutique solutions like Z-NAND and 3D XPoint.
althaz - Friday, January 18, 2019 - link
The 512Gb drive is a ~$150 part, just buy it and test it? To many this remains the most prestigious tech site around for at least some things. SSDs have been one of those things. Not having the latest of Samsung's pro drive in your results kinda makes this whole thing not worth it, IMO.If you need an excuse do a "MLC Redux" review where you look at what was probably the faster ever MLC drive and talk about how SSD tech has changed.
eldakka - Friday, January 18, 2019 - link
" just buy it and test it?"I'm sure anandtech would be delighted to test it if you sample one to them, or donate $150 to them to purchase one with.
GreenReaper - Saturday, January 19, 2019 - link
This is the level of a business expense though. Even if you did one a day for every day of the year it would be $53,400. OK, you could get a full-time journalist for that in some places, but that assumes they are kept and not sold on or used for anything else.Solandri - Monday, January 21, 2019 - link
More troubling is that unwillingness to buy a product for review smacks of demanding bribes. "Give us a free sample or we won't review your product." Which implies that companies which shower the reviewer with gifts will get more favorable reviews.Ideally, a review site should *never* accept free samples, and do all their product reviews with samples bought from the store. That's the only way to completely eliminate any undue influence the product manufacturer may have on the product review.
Dark_wizzie - Friday, January 18, 2019 - link
I don't think they actually reviewed the 970 Pro?StevoLincolnite - Saturday, January 19, 2019 - link
Yeah. Comparisons have been a bit crap on Anandtech lately.The RTX 2080 review is lacking a good lineup of GPU's to compare with... Despite promises of "Adding more later". (Never happened sadly.)
joesiv - Friday, January 18, 2019 - link
I would love to see at least some endurance testing in anandtechs reviews. I think it's something that's missing pretty much all SSD testing I see. Yes, speed is great, but endurance is something perspective owners will care more about in the long term.NAND technology is changing over time, is it changing for the better? More layers, but often at the expense of P/E Cycles.
SSD Firmwares as well have gotten both a lot more reliable, but also a lot more complicated, with different companies doing different things to maximize performance and also endurance.
Without testing these things, we don't know if there are duds in terms of firmwares/drives that don't strike a good balance between performance/reliability.
Bare minimimum I would love to see ananadtech do the following:
- Talk to manufacturers about the actual specifications of the NAND, what is the P/E Cycle rating for the NAND. Then list it in the specs for the drives so we can do our own research/homework if we want to. This spec is very often missed in the most marketing material.
- Do a quick capture of the SMART data at the very start of the testing
- Do a capture of the SMART data at the very end of your testing.
- Check out the Block Erase Count specifically. The Average Block Erase count would give a good approximation of how much life you've used just within your benchmark suite. It's much more grainular than Percent Life Used. This SMART attribute can vary between different manufacturer/firmware/controllers, so you might need to contact the manufacturer for this info.
In the end, you'll have something like: Rated P/E count / Average Block Erase, and since your test suite is likely pretty similar between SSD's, it would be a useful metric.
Bonus points, you could also look at NAND writes (not to be confused with Host Writes), during your testing, as it's related to life expectancy and can be corrolated with Total Byte Written.
I wouldn't be surprised to see that some drives write a lot more to NAND than you are expecting, and thus have higher Average Block Erases during your testing.
At my company when I did this test, while we were evaluating a drive for our products, Tier 1 memory manufacturers and their drives, failed bitterly at this endurance test where most did fine. I attribute this to the firmware. Working with the firmware developement team, we determined that the fix would not be feasible in that generation of product, so we had to skip that product.
Specs aren't everything, we need to test it, endurance is an important aspect, and I'd love to see it represented in your reviews.
romrunning - Friday, January 18, 2019 - link
Whom do you consider to be "Tier 1 memory manufacturers" - Samsung or Intel? I can't imagine their products failing endurance tests. Or are you talking about those you would consider to be lower than Samsung - like AData, Team Group, etc. - as a Tier 2 or Tier 3?