ADATA likes to produce a broad range of SSDs, sampling from all the controller and NAND manufacturers. To that end, they have wholeheartedly embraced the use of 3D MLC NAND even as most brands are using the 3D NAND transition to entirely remove MLC from their consumer product lines or relegate it to niche models instead of treating it as the mainstream default. ADATA is selling multiple SATA and NVMe models using Micron's first-generation 3D MLC NAND. On the SATA side of things, they have the Ultimate SU900 and XPG SX950 as the MLC models (positioned above the SU800 and SU700 using 3D TLC). For the NVMe market, they have the XPG SX8000 and XPG SX9000 pairing 3D MLC with Silicon Motion and Marvell controllers respectively. So far, all of their 3D NAND SSDs have relied on Micron's first-generation 32-layer 3D NAND, the only 3D NAND available in volume on the open market.

The ADATA XPG SX950 is their top of the line SATA SSD. Technologically, it is very similar to their Ultimate SU900: both use the same Micron 3D MLC NAND and Silicon Motion SM2258 controller. The SX950 is distinguished by reserving more spare area (yielding usable capacities like 480GB instead of 512GB) and a six-year warranty instead of five. There may be significant firmware tuning differences, but there are no obvious signs such as one model providing TCG Opal encryption support (both lack that feature). ADATA did their own NAND packaging for the SX950, so it may be using flash binned for better endurance. The SU900 has a 2TB model listed but not yet available, while the SX950 line only goes up to 960GB, with limited availability of the largest model.

ADATA XPG SX950 Specifications
Capacity	240 GB	480 GB	960 GB
Model Number	ASX950SS-240GM-C	ASX950SS-480GM-C	ASX950SS-960GM-C
Controller	Silicon Motion SM2258
NAND Flash	Micron 256Gbit 32-layer 3D MLC NAND
Sequential Read	560 MB/s
Sequential Write	520 MB/s	530 MB/s
Random Read IOPS	80k IOPS	90k IOPS	90k IOPS
Random Write IOPS	90k IOPS	90k IOPS	85k IOPS
TCG Opal Encryption	No
Power Consumption	Active: 0.82 W Slumber: 0.41 W
Warranty	6 years
Write Endurance	200 TB	400 TB	800 TB
MSRP	$139.99	$249.99	$449.99

The construction of the ADATA XPG SX950 is similar to ADATA's other recent SATA SSDs: a metal base and plastic lid are joined by a single screw through the center of the drive. The PCB occupies only half of the case, and has pads for eight NAND packages and two DRAM packages. On our 480GB sample, all eight NAND pads are occupied with dual-die packages, for a total raw capacity of 512GB.

Micron recently re-introduced MLC NAND to their consumer product line with the Crucial BX300, using their 32L 3D MLC and the SM2258 controller. The Crucial BX300 is positioned as more of an entry-level model, with a focus on lower capacities that cannot perform well when using their 3D TLC. The ADATA SX950 has twice the warranty period and up to five times the rated write endurance, but any performance differences will come down to firmware tweaks, and ADATA hasn't updated the firmware since the SX950 launched earlier this year. Current pricing for the SX950 shows that ADATA has not responded to the introduction of the BX300: for the two capacities where the models match up, the ADATA is more expensive by 25% and 45%. That's quite a premium for a longer warranty.

AnandTech 2017 SSD Testbed
CPU	Intel Xeon E3 1240 v5
Motherboard	ASRock Fatal1ty E3V5 Performance Gaming/OC
Chipset	Intel C232
Memory	4x 8GB G.SKILL Ripjaws DDR4-2400 CL15
Graphics	AMD Radeon HD 5450, 1920x1200@60Hz
Software	Windows 10 x64, version 1703
	Linux kernel version 4.12, fio version 2.21

• Thanks to Intel for the Xeon E3 1240 v5 CPU
• Thanks to ASRock for the E3V5 Performance Gaming/OC
• Thanks to G.SKILL for the Ripjaws DDR4-2400 RAM
• Thanks to Corsair for the RM750 power supply, Carbide 200R case, and Hydro H60 CPU cooler

AnandTech Storage Bench - The Destroyer

The Destroyer is an extremely long test replicating the access patterns of very IO-intensive desktop usage. A detailed breakdown can be found in this article. Like real-world usage, the drives do get the occasional break that allows for some background garbage collection and flushing caches, but those idle times are limited to 25ms so that it doesn't take all week to run the test. These AnandTech Storage Bench (ATSB) tests do not involve running the actual applications that generated the workloads, so the scores are relatively insensitive to changes in CPU performance and RAM from our new testbed, but the jump to a newer version of Windows and the newer storage drivers can have an impact.

We quantify performance on this test by reporting the drive's average data throughput, the average latency of the I/O operations, and the total energy used by the drive over the course of the test.

The ADATA XPG SX950 is the slowest MLC SSD in this group, going by its average data rate on The Destroyer, while the similarly-equipped Crucial BX300 is the second-fastest SATA drive in the half-TB capacity class.

ADATA has a latency problem on The Destroyer. The SX950's average latency is much worse than any other MLC SSD, and the 99th percentile latency as bad as the TLC-based SU800, which was already an extreme outlier.

Splitting the average latency up by read and write operations, it's clear that the SX950's troubles are mostly on the write side, though the average read latency is also more typical of a TLC SSD than one with 3D MLC.

The 99th percentile read latency of the ADATA SX950 is not much worse than other 3D MLC SSDs, but the 99th percentile write latency on The Destroyer is unusually high at over 81ms. It appears that the SX950 is being quite aggressive with its SLC caching, leading to a serious backlog when it is finally forced to perform garbage collection. The BX300 avoids this by using relatively small fixed-size SLC caches.

Given the mediocre data rate and poor QoS indicating a lot of background work, it's not too surprising to see that the SX950's energy usage on The Destroyer is substantially higher than the Crucial BX300 and most other 3D NAND SSDs. The SX950 does shave 25% off the energy usage of the TLC-based ADATA SU800, but Crucial still does much better with the same controller.

AnandTech Storage Bench - Heavy

Our Heavy storage benchmark is proportionally more write-heavy than The Destroyer, but much shorter overall. The total writes in the Heavy test aren't enough to fill the drive, so performance never drops down to steady state. This test is far more representative of a power user's day to day usage, and is heavily influenced by the drive's peak performance. The Heavy workload test details can be found here. This test is run twice, once on a freshly erased drive and once after filling the drive with sequential writes.

When the Heavy test is run on a fresh drive, the ADATA XPG SX950 delivers a good average data rate that is a bit higher than the Crucial BX300 and close to the level of Samsung's SATA SSDs. But when the drive is full, the SX950 suffers greatly in the manner of the SU800 and the Crucial MX300, while the BX300 is minimally affected.

Both the average and 99th percentile latencies highlight how poorly the SX950 performs when the Heavy test is run on a full drive, but the latency when the test is run on a fresh drive is normal.

The average read and write latencies of the SX950 are both significantly higher when the Heavy test is run on a full drive, but the write latency is again far more strongly affected. For both reads and writes, the full-drive performance is better than the TLC-based ADATA SU800, but nowhere close to the standard set by the Crucial BX300.

The MLC-based SSDs almost all show very little degradation in 99th percentile read and write latencies when the drive is full. The exceptions are the ADATA SX950 and the DRAMless OCZ VX500. The 99th percentile read latency of the SX950 is higher when the drive is full, but still better than the planar TLC drives. The 99th percentile write latency on the other hand grows by more than an order of magnitude to almost 95ms.

The ADATA SX950 scores very well on energy usage when the Heavy test is run on an empty drive: it matches the DRAMless OCZ VX500 and comes close to the Crucial MX300, which uses a Marvell controller fabbed on a newer and lower-power process. The Crucial BX300 uses 14% more energy largely due to taking longer overall to complete the test. The situation is reversed when running the test on a full drive: the ADATA SX950 takes much longer to complete the test and is doing a lot of costly background garbage collection, though it still at least beats the planar TLC SSDs and ADATA's own TLC-based SU800.

AnandTech Storage Bench - Light

Our Light storage test has relatively more sequential accesses and lower queue depths than The Destroyer or the Heavy test, and it's by far the shortest test overall. It's based largely on applications that aren't highly dependent on storage performance, so this is a test more of application launch times and file load times. This test can be seen as the sum of all the little delays in daily usage, but with the idle times trimmed to 25ms it takes less than half an hour to run. Details of the Light test can be found here. As with the ATSB Heavy test, this test is run with the drive both freshly erased and empty, and after filling the drive with sequential writes.

The ADATA XPG SX950 continues the trend of being reasonably fast when fresh, but much slower when full. Its average data rate on the Light test is almost as fast as the Samsung 850 Pro when the test is run on an empty drive, but when the drive is full it returns the lowest average data rate score in its class.

There usually isn't much to look at in the latency results for the Light test; most SATA drives perform very similarly. The ADATA SX950 is normal when the drive is fresh, but when the drive is filled before running the test, the results are very unlike any other SATA SSD we've tested. With a 99th percentile latency of over 30ms (dwarfing the Crucial MX300's 6.5ms), it's clear the SX950 does not manage its background processing properly.

The Light test is easy enough that the average read latency of the SX950 is normal whether or not the drive was filled before running the test. The average write latency is still several times higher in the full-drive case, and is twice that of the next-slowest drive in this bunch.

As with the average latency results, the 99th percentile read latency of the SX950 is not appreciably higher than normal when the Light test is run on a full drive, but on the write side latency is out of control. The Crucial BX300 has a substantially higher 99th percentile write latency when the test is run on an empty drive, but it doesn't go to pieces when the test is run on a full drive.

Energy usage by the ADATA SX950 is again quite good when the test is run on an empty drive. When the drive is full, the energy usage still falls within the acceptable range, but is unimpressive.

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.

The short-burst QD1 random read speed of the ADATA XPG SX950 is good, though Samsung still has a clear lead. The SX950 is 16% faster than the Crucial BX300 that uses the same NAND and controller.

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.

With a longer test and higher queue depths, the ADATA SX950's standing drops somewhat, but it still outperforms most TLC SSDs and the Crucial BX300.

The power efficiency of the SX950 during random reads is decent but not outstanding. The planar MLC-based PNY CS2211 beats the SX950 on both performance and power efficiency, but otherwise the SX950 is only beat by a few other 3D NAND SSDs.

ADATA XPG SX950 480GBIntel SSD 545s 512GBSamsung 850 EVO 500GBSamsung 850 PRO 512GBPNY CS1311 480GBPNY CS2211 480GBSanDisk Extreme Pro 480GBCrucial MX200 500GBCrucial MX300 525GBADATA Ultimate SU800 512GBOCZ Trion 150 480GBOCZ VX500 512GBCrucial BX300 480GB

For an MLC-based SSD, the ADATA SX950 does a poor job of scaling up random read performance with higher queue depths. Its QD1 performance is fine, but by QD8 it's lagging behind many of its competitors. The Crucial BX300 is marginally slower until QD32, where the SX950's performance starts to level off while the BX300 gets very close to the peak speeds reached by the best SATA SSDs.

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.

After seeing clear indicators from the ATSB tests that the ADATA SX950 is very aggressive with SLC write caching, it's good to see that the QD1 burst random write performance is near the top of the charts.

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 with higher queue depths included, the ADATA XPG SX950's random write score is unimpressive: barely faster than the TLC-based ADATA SU800 and well behind the Crucial and Samsung drives.

In spite of mediocre performance on the longer random write test, the SX950 delivers great power efficiency, in second place only slightly behind the Crucial MX300.

ADATA XPG SX950 480GBIntel SSD 545s 512GBSamsung 850 EVO 500GBSamsung 850 PRO 512GBPNY CS1311 480GBPNY CS2211 480GBSanDisk Extreme Pro 480GBCrucial MX200 500GBCrucial MX300 525GBADATA Ultimate SU800 512GBOCZ Trion 150 480GBOCZ VX500 512GBCrucial BX300 480GB

The SX950's random write speed mostly levels off in the second half of the test as queue depths grow beyond 4, and even at QD32 the SX950 is not really close to saturating the SATA bus. But at QD1 and QD2, its performance is as good as any SATA drive and its power consumption is unbeatable.

Sequential Read Performance

Our first test of sequential read performance uses short bursts of 128MB, issued as 128kB operations with no queuing. The test averages performance across eight bursts for a total of 1GB of data transferred from a drive containing 16GB of data. Between each burst the drive is given enough idle time to keep the overall duty cycle at 20%.

For short bursts of sequential reads at QD1, the ADATA XPG SX950 comes reasonably close to saturating the SATA bus and is not significantly outperformed by any other SATA drive.

Our test of sustained sequential reads uses queue depths from 1 to 32, with the performance and power scores computed as the average of QD1, QD2 and QD4. Each queue depth is tested for up to one minute or 32GB transferred, from a drive containing 64GB of data.

On the longer sequential read test, the SX950's performance drops dramatically, leaving it as one of the slowest SSDs in this bunch.

With such poor sustained read performance, the SX950's efficiency is quite low, though all the other drives with Intel/Micron 32L 3D NAND also rank at the bottom of this chart with the SX950 even when they deliver higher performance.

ADATA XPG SX950 480GBIntel SSD 545s 512GBSamsung 850 EVO 500GBSamsung 850 PRO 512GBPNY CS1311 480GBPNY CS2211 480GBSanDisk Extreme Pro 480GBCrucial MX200 500GBCrucial MX300 525GBADATA Ultimate SU800 512GBOCZ Trion 150 480GBOCZ VX500 512GBCrucial BX300 480GB

Sustained sequential read speeds on the ADATA SX950 do not increase with higher queue depths. The burst test showed that higher speeds are possible in favorable conditions, but giving the SSD more work to do when it is already bottlenecked internally doesn't help things.

Sequential Write Performance

Our test of sequential write burst performance is structured identically to the sequential read burst performance test save for the direction of the data transfer. Each burst writes 128MB as 128kB operations issued at QD1, for a total of 1GB of data written to a drive containing 16GB of data.

The burst sequential write speed of the SX950 is about average, and virtually identical to its TLC-based SU800 sibling; this test is primarily hitting the SLC caches so it's unsurprising to see the two perform the same.

Our test of sustained sequential writes is structured identically to our sustained sequential read test, save for the direction of the data transfers. Queue depths range from 1 to 32 and each queue depth is tested for up to one minute or 32GB, followed by up to one minute of idle time for the drive to cool off and perform garbage collection. The test is confined to a 64GB span of the drive.

The sustained sequential write speeds of the SX950 fall short of the top SATA drives and are instead slightly lower than the Crucial BX300, but since this test doesn't fill the drive the SX950's write performance doesn't end up falling through the floor.

The SX950's power efficiency during the sustained sequential write test is pretty good, though the Crucial MX300 and OCZ VX500 still manage to beat it. The SX950 has about a 20% advantage over the BX300, which is the next most efficient drive using 3D MLC.

ADATA XPG SX950 480GBIntel SSD 545s 512GBSamsung 850 EVO 500GBSamsung 850 PRO 512GBPNY CS1311 480GBPNY CS2211 480GBSanDisk Extreme Pro 480GBCrucial MX200 500GBCrucial MX300 525GBADATA Ultimate SU800 512GBOCZ Trion 150 480GBOCZ VX500 512GBCrucial BX300 480GB

The SX950 doesn't reach full sequential write performance until QD4, and its saturation speed is slightly slower than drives like the Samsung 850 PRO (which is very nearly at full speed by QD2). Performance and power consumption remain steady through the second half of the test as the drive still has plenty of SLC cache remaining.

Mixed Random Performance

Our test of mixed random reads and writes covers mixes varying from pure reads to pure writes at 10% increments. Each mix is tested for up to 1 minute or 32GB of data transferred. The test is conducted with a queue depth of 4, and is limited to a 64GB span of the drive. In between each mix, the drive is given idle time of up to one minute so that the overall duty cycle is 50%.

The overall performance of the ADATA XPG SX950 on the mixed random I/O test is just a bit average, and about 20% slower than Samsung's SATA drives.

The power efficiency of the SX950 looks a bit better, but the Samsung 850 EVO managed to tie for first place on performance wile holding a substantial lead on power efficiency. The SX950 does hold a slight efficiency advantage over the Samsung 850 PRO on this test.

ADATA XPG SX950 480GBIntel SSD 545s 512GBSamsung 850 EVO 500GBSamsung 850 PRO 512GBPNY CS1311 480GBPNY CS2211 480GBSanDisk Extreme Pro 480GBCrucial MX200 500GBCrucial MX300 525GBADATA Ultimate SU800 512GBOCZ Trion 150 480GBOCZ VX500 512GBCrucial BX300 480GB

The SX950's power consumption is very flat across the varying workloads of this test, while the performance improves steadily as the share of writes grows. The top-performing Samsung drives start out about 40MB/s faster than the SX950 and most drives show a more significant performance spike when they reach the all-writes phase of the test.

Mixed Sequential Performance

Our test of mixed sequential reads and writes differs from the mixed random I/O test by performing 128kB sequential accesses rather than 4kB accesses at random locations, and the sequential test is conducted at queue depth 1. The range of mixes tested is the same, and the timing and limits on data transfers are also the same as above.

The mixed sequential I/O performance of the ADATA XPG SX950 is sub-par, though there are actually three planar MLC SSDs that are even slower. The Crucial BX300 is about 14% faster overall.

The SX950's efficiency is tied with the Crucial BX300, and both are near the top of the chart. The DRAMless OCZ VX500 is far more efficient that the rest of these drives, and without it the SX950's efficiency score would look very good.

ADATA XPG SX950 480GBIntel SSD 545s 512GBSamsung 850 EVO 500GBSamsung 850 PRO 512GBPNY CS1311 480GBPNY CS2211 480GBSanDisk Extreme Pro 480GBCrucial MX200 500GBCrucial MX300 525GBADATA Ultimate SU800 512GBOCZ Trion 150 480GBOCZ VX500 512GBCrucial BX300 480GB

During the first half of this test as the proportion of writes grows, the SX950's performance increases and the power consumption drops. During the second half, the improvement falters but neither performance nor power consumption regress significantly. The Crucial BX300 is faster across the board, especially during the read-heavy parts of the test, but it requires significantly more power to deliver that higher read performance.

Power Management

Real-world client storage workloads leave SSDs idle most of the time, so the active power measurements presented earlier in this review only account for a small part of what determines a drive's suitability for battery-powered use. Especially under light use, the power efficiency of a SSD is determined mostly be how well it can save power when idle.

SATA SSDs are tested with SATA link power management disabled to measure their active idle power draw, and with it enabled for the deeper idle power consumption score and the idle wake-up latency test. Our testbed, like any ordinary desktop system, cannot trigger the deepest DevSleep idle state.

Idle power management for NVMe SSDs is far more complicated than for SATA SSDs. NVMe SSDs can support several different idle power states, and through the Autonomous Power State Transition (APST) feature the operating system can set a drive's policy for when to drop down to a lower power state. There is typically a tradeoff in that lower-power states take longer to enter and wake up from, so the choice about what power states to use may differ for desktop and notebooks.

We report two idle power measurements. Active idle is representative of a typical desktop, where none of the advanced PCIe link or NVMe power saving features are enabled and the drive is immediately ready to process new commands. The idle power consumption metric is measured with PCIe Active State Power Management L1.2 state enabled and NVMe APST enabled.

Our ADATA XPG SX950 didn't seem to ever make use of the slumber power state; it draws about 0.8W with or without power management enabled. This is a relatively high active idle power draw for a SATA SSD, and the lack of deeper idle is unacceptable for mobile use.

The bright side of not using any deep idle states is that the ADATA XPG SX950 doesn't take any time to wake up; the 16µs latency we measured is pretty much all on the host side. Phison's SATA drives also do very well on this metric, and the rest all need more than half a millisecond to wake up.

Conclusion

At no point in our testing did the ADATA XPG SX950 convince us that it deserves to be regarded as a high-end SATA SSD, nor did it provide any evidence that the high-end SATA segment is still relevant. Under the right conditions, the SX950 can perform as fast as any other SATA SSD, but those are all the same tests where a low-end SATA SSD also performs fine.

Without the performance headroom that PCIe SSDs enjoy, a premium SATA SSD needs to distinguish itself by offering great performance in all conditions, under light or heavy workloads. The SX950 does the opposite. The aggressive SLC caching it uses to deliver high peak performance is a double-edged sword. When subjected to a large volume of writes, the SX950 accrues a large debt of cache flushing and garbage collection that have been deferred. Once the SLC cache fills up, the SX950's performance tanks. Both reads and writes suffer, though write performance much moreso. Worse, it takes the SX950 too long to finish cleaning up even when given the opportunity. The five minutes of idle time our test protocol reserves after filling the drive in preparation for some of the ATSB tests is clearly not long enough, and even during the ATSB Light test the SX950 can't finish catching up on its garbage collection.

The more recent Crucial BX300 uses the same Micron 32L 3D MLC and the same Silicon Motion SM2258 controller, but exhibits a completely different performance profile. Micron learned their lesson about taking SLC caching unnecessarily far on MLC drives with the Crucial MX200, and the BX300 has fairly small fixed-size SLC caches. This leads it to have lower performance than the SX950 under favorable conditions, but the BX300 holds up well under pressure.

The ADATA SX950 offers twice the warranty period of a typical budget SSD and a fairly high write endurance rating, but those are the only ways in which it can be regarded as a premium product. It doesn't even provide TCG Opal encryption support, a distinguishing feature that only a handful of SSD vendors implement for retail SSDs. Micron's first-generation 3D NAND is simply too slow to compete against Samsung's 3D NAND, and the SM2258 is a low-cost/low-power SSD controller that is ill-suited for competing against Marvell and Samsung controllers on performance. The result is a drive that not only falls far short of its lofty performance goals, but a drive that has unbalanced performance and makes poor use of the resources it has at hand.

SATA SSD Price Comparison
	240-275GB	480-525GB	960-1050GB
ADATA XPG SX950	$135.22 (56¢/GB)	$269.99 (50¢/GB)
ADATA SU800	$89.99 (35¢/GB)	$158.65 (31¢/GB)	$274.99 (27¢/GB)
Crucial BX300	$89.99 (38¢/GB)	$149.99 (31¢/GB)
Crucial MX300	$92.99 (34¢/GB)	$149.99 (29¢/GB)	$279.99 (27¢/GB)
Intel SSD 545s	$99.99 (39¢/GB)	$179.99 (35¢/GB)
Samsung 850 PRO	$128.98 (50¢/GB)	$223.32 (44¢/GB)	$447.87 (44¢/GB)
Samsung 850 EVO	$99.95 (40¢/GB)	$159.99 (32¢/GB)	$327.99 (33¢/GB)
SanDisk Ultra 3D	$99.99 (40¢/GB)	$164.99 (33¢/GB)	$284.99 (29¢/GB)
WD Blue 3D NAND	$98.39 (38¢/GB)	$164.65 (33¢/GB)	$299.99 (30¢/GB)

On a budget SSD, the problems with the ADATA XPG SX950 would be mildly disappointing but reasonable. For light desktop use, the SX950's weaknesses wouldn't come into play. But given the premium pricing, the SX950's failings are unacceptable. ADATA can't beat Samsung's 850 PRO on price, let alone the Crucial BX300.

As the Crucial BX300 shows, most of this could be fixed with radically different firmware. But ADATA can't get anywhere by trying to compete directly against the BX300 and its unbelievably low pricing. Instead, ADATA should take the lessons learned from the SX950 and prepare to offer a more sensible drive when they can get their hands on Micron's second generation 3D NAND, which the Intel 545s suggests will be far faster and more able to match the Samsung 850 PRO, especially if used with Silicon Motion's newest SM2259 SATA controller. However, that would still leave ADATA competing in a very narrow segment of the SSD market, as almost all premium products are now using NVMe and focusing more on performance than endurance and warranty period.