Anandtech reviews products independently. When you click links to buy products we may earn money that supports our work.


Original Link: https://www.anandtech.com/show/10131/the-adata-premier-sp550-ssd-review



Silicon Motion's SM2246EN SSD controller was a highly successful solution for SSDs using MLC NAND flash. It offered a great combination of power efficiency, solid performance, and low cost. Last year we got an early look at the successor SM2256 controller for drives using TLC NAND and found that despite being made for TLC it didn't do a very good job of mitigating the performance and power consumption problems inherent to TLC NAND. At the time there were only a few mainstream TLC drives and the SM2256 was one of the first controllers designed specifically with TLC in mind. That preview was based on a reference design drive using Samsung 19nm TLC and prerelease firmware, so we were hopeful for improvement in final retail products that would be using different NAND and firmware.

In November we looked at the performance of the Crucial BX200 based on the SM2256 controller and Micron 16nm TLC NAND, and we found that the situation had gotten much worse: performance was bad and power consumption was high. In the meantime, Phison's S10 controller (and Toshiba's custom branded variant of it) became the most popular solution for TLC SSDs and the market was flooded with nearly-identical drives from new and old players in the market.

That brings us to today's review of ADATA's Premier SP550 SSD. ADATA manufactures drives with almost every controller available on the open market, and the Premier SP550 is their SM2256 offering. ADATA has in turn positioned the drive as the entry-level model of their mainstream consumer Premier product line. Meanwhile with a different choice of flash (SK Hynix 16nm TLC NAND) and updated firmware, the SP550 has the chance to partially redeem the SM2256 controller.

ADATA Premier SP550 Specifications
Capacity 120GB 240GB 480GB 960GB
Controller Silicon Motion SM2256
NAND Flash SK Hynix 16nm TLC
Sequential Read 560MB/s 560MB/s 560MB/s 520MB/s
Sequential Write 410MB/s 510MB/s 510MB/s 490MB/s
Random Read IOPS 60k 75k 75k 80k
Random Write IOPS 70k 75k 75k 35k
TCG Opal Encryption No
Power Management Slumber and DevSleep
Warranty 3 years
Current Retail Price $38.99 $57.99 $112.99 $219.99

With pricing starting at $39 for 120GB and even the 480GB model staying under $0.25/GB, the pricing is a good start, as the SP550 has consistently been one of the cheapest drives on the market. No TLC drive has yet reached a price low enough to completely excuse its shortcomings relative to MLC drives and earn an unconditional recommendation, but at least the SP550 has kept pace with the price drops in recent months. Its primary competitors are Phison/Toshiba-based SSDs with Toshiba's A19nm TLC, as their successors based on Toshiba's 15nm TLC are still new enough to the market to be selling for higher prices closer to the MSRP. This means we'll be looking at the OCZ Trion 100 as the most likely alternative, though similar drives from Silicon Power and PNY are available in the same price range and newer drives like the OCZ Trion 150 may match its price it in the near future.

AnandTech 2015 SSD Test System
CPU Intel Core i7-4770K running at 3.5GHz (Turbo & EIST enabled, C-states disabled)
Motherboard ASUS Z97 Deluxe (BIOS 2501)
Chipset Intel Z97
Memory Corsair Vengeance DDR3-1866 2x8GB (9-10-9-27 2T)
Graphics Intel HD Graphics 4600
Desktop Resolution 1920 x 1200
OS Windows 8.1 x64


Performance Consistency

Starting things off, our performance consistency test explores the extent to which a drive can reliably sustain performance during a long-duration random write test. Specifications for consumer drives typically list peak performance numbers only attainable in ideal conditions. The performance in a worst-case scenario can be drastically different as over the course of a long test drives can run out of spare area, have to start performing garbage collection, and sometimes even reach power or thermal limits.

In addition to an overall decline in performance, a long test can show patterns in how performance varies on shorter timescales. Some drives will exhibit very little variance in performance from second to second, while others will show massive drops in performance during each garbage collection cycle but otherwise maintain good performance, and others show constantly wide variance. If a drive periodically slows to hard drive levels of performance, it may feel slow to use even if its overall average performance is very high.

To maximally stress the drive's controller and force it to perform garbage collection and wear leveling, this test conducts 4kB random writes with a queue depth of 32. The drive is filled before the start of the test, and the test duration is one hour. Any spare area will be exhausted early in the test and by the end of the hour even the largest drives with the most overprovisioning will have reached a steady state. We use the last 400 seconds of the test to score the drive both on steady-state average writes per second and on its performance divided by the standard deviation.

Steady-State 4KB Random Write Performance

For its capacity, the SP550 maintains a higher average random write speed than most low-end drives and even manages to beat some MLC drives, including a narrow advantage over the SM2246EN-based Crucial BX100. This is significant as the only major change between the SM2246EN and the SM2256 was the addition of LDPC error correction, so the firmware on the SP550 is achieving better performance with slower flash on a controller that is basically the same where this benchmark is concerned.

Steady-State 4KB Random Write Consistency

The SM2256 is a budget-oriented controller based around a single-core processor. This makes it extremely difficult to completely eliminate pauses or periods of greatly reduced performance during garbage collection. The SP550 is hardly alone in this, and its consistency scores are typical for its market segment.

IOPS over time
Default
25% Over-Provisioning

Upon reaching steady state, the SP550 shows a clear baseline of performance plus some random variation that is capped at the average IOPS attained before the spare area was exhausted. At the higher capacities a much higher share of the writes are completed at full speed and the distribution becomes mostly bimodal.

Steady-State IOPS over time
Default
25% Over-Provisioning

Looking more closely at the steady state, we can see that the garbage collection cycles take longer on the higher capacity drives, though not as long as on the Toshiba Q300. Neither the baseline performance nor the peak performance is strongly dependent on capacity.



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 and unlike our Iometer tests, 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.

We quantify performance on this test by reporting the drive's average data throughput, a few data points about its latency, and the total energy used by the drive over the course of the test.

AnandTech Storage Bench - The Destroyer (Data Rate)

The average data rates sustained by the SP550 on The Destroyer are pretty good by budget drive standards. It's generally performing on par with other TLC drives of twice the capacity.

AnandTech Storage Bench - The Destroyer (Latency)

The latency situation for the SP550 is not good. Each model has a higher average service time than any other drive of equal capacity, except for the Crucial BX200, with the two smaller drives reaching values close to HDD rotational latency. While the Phison-based drives have lower average data rates, they are better at keeping the latency under control.

AnandTech Storage Bench - The Destroyer (Latency)

AnandTech Storage Bench - The Destroyer (Latency)

The frequency of moderate latency outliers is a bit higher for the SP550 than the competition, but it does a decent job of preventing the most extreme outliers that would not pass unnoticed by a user.

AnandTech Storage Bench - The Destroyer (Power)

The BX200 posted an incredibly bad energy usage score but the SP550 completely turns that around and is the most energy-efficient drive with planar TLC NAND. Most MLC drives still score better due to completing the test in a shorter time span, but the SP550 is bridging that gap.



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.

AnandTech Storage Bench - Heavy (Data Rate)

The SP550's average data rates on the Heavy test are what we should have seen from the BX200 and are close to what the prototype delivered. The SP550 is trading blows with the OCZ Trion 100, one of its primary competitors. The difference in performance between starting the test with the drive full or empty is much smaller for the SM2256 drives than almost everything else, so for day to day use there's little reason to be concerned about reserving some free space.

AnandTech Storage Bench - Heavy (Latency)

The average service time of the SP550 is again poor, beating only the BX200 and losing to everything else by a substantial margin.

AnandTech Storage Bench - Heavy (Latency)

The high average service time of the SP550 is due in large part to the high number of latency outliers. Taking capacity into account, it's still a huge improvement over the Crucial BX200.

AnandTech Storage Bench - Heavy (Power)

The energy usage of the SP550 under our Heavy test is unremarkable. Aside from the BX200 there isn't a broad spread of scores and the SP550 is typical of a TLC-based SSD.



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.

AnandTech Storage Bench - Light (Data Rate)

The average data rate of the SP550 on the Light test can't quite keep up with the competition when starting with an empty drive, but the performance penalty of a full drive is again relatively small, allowing it to clearly surpass the Trion 100 in that scenario. It's interesting to note that the performance of the 480GB SP550 is almost exactly the same as the 120GB Crucial BX100 which uses MLC NAND and the earlier SM2246EN controller: the performance penalty of TLC is clear even on the lightest ATSB test.

AnandTech Storage Bench - Light (Latency)

The average service time scores of most drives are pretty similar, but this test clearly isn't light enough for the BX200 and smaller capacities of SP550 to completely maintain peak performance.

AnandTech Storage Bench - Light (Latency)

As with the average service time, the frequency of latency outliers for the smaller capacities of SP550 show that it manages to get bogged down in a few parts of this test where most other drives breeze through.

AnandTech Storage Bench - Light (Power)

The energy usage of the SP550 doesn't stand out, and for this test that's a good thing. Its efficiency is on par with most other TLC drives.



Random Read Performance

The random read test requests 4kB blocks and tests queue depths ranging from 1 to 32. The queue depth is doubled every three minutes, for a total test duration of 18 minutes. The test spans the entire drive, which is filled before the test starts. The primary score we report is an average of performances at queue depths 1, 2 and 4, as client usage typically consists mostly of low queue depth operations.

Iometer - 4KB Random Read

The SP550's random read performance is slightly higher than most of the TLC-based competition, but significantly slower than the SM2256 prototype with Samsung 19nm TLC managed. There's still a clear gap between the MLC drives and most of the TLC drives. The Samsung 850 EVO and SanDisk Ultra II are still the only TLC drives that can compete against MLC.

Iometer - 4KB Random Read (Power)

The SP550's power consumption is about average. TLC flash usually only has a slight disadvantage for read power and Silicon Motion's very efficient controller offsets that for the SP550.

Looking at performance over various queue depths, all three capacities of the SP550 perform very similarly at low depths, with performance that starts quite low and doesn't increase very steeply. At higher queue depths the 120GB SP550 isn't able to scale up performance as much as the larger capacities, and even the 480GB model can't crack 300MB/s while high-end SATA drives top out around 400MB/s.

Random Write Performance

The random write test writes 4kB blocks and tests queue depths ranging from 1 to 32. The queue depth is doubled every three minutes, for a total test duration of 18 minutes. The test is limited to a 16GB portion of the drive, and the drive is empty save for the 16GB test file. The primary score we report is an average of performances at queue depths 1, 2 and 4, as client usage typically consists mostly of low queue depth operations.

Iometer - 4KB Random Write

Random write performance is the SP550's biggest weakness. It beats the Crucial BX200 at equivalent capacities, but falls well behind everything else. While the steady-state random write performance was relatively good, this performance on this shorter test does not impress.

Iometer - 4KB Random Write (Power)

The power consumption of the SP550 on the random write test is very good, but given the bad performance there's no efficiency advantage.

The SP550's random write speed doesn't scale much with queue depth. Fortunately, neither does the power consumption—a welcome change from the Crucial BX200. The OCZ Trion 100 and Toshiba Q300 also don't scale, but everything else does to at least some extent, including the SM2256 prototype.



Sequential Read Performance

The sequential read test requests 128kB blocks and tests queue depths ranging from 1 to 32. The queue depth is doubled every three minutes, for a total test duration of 18 minutes. The test spans the entire drive, and the drive is filled before the test begins. The primary score we report is an average of performances at queue depths 1, 2 and 4, as client usage typically consists mostly of low queue depth operations.

Iometer - 128KB Sequential Read

The sequential read speed of the SP550 places it in the second tier, but there's nothing to complain about.

Iometer - 128KB Sequential Read (Power)

Power consumption during sequential reads is not quite low enough for the SP550 to compete against the best MLC-based drives, but its efficiency is otherwise pretty good.

The SP550's sequential read performance saturates by the time queue depth reaches 4 and is close at QD2. This drive isn't quite the fastest to reach its limit, but it's still easy to get full performance from it. The only real flaw here is the relatively low QD1 performance.

Sequential Write Performance

The sequential write test writes 128kB blocks and tests queue depths ranging from 1 to 32. The queue depth is doubled every three minutes, for a total test duration of 18 minutes. The test spans the entire drive, and the drive is filled before the test begins. The primary score we report is an average of performances at queue depths 1, 2 and 4, as client usage typically consists mostly of low queue depth operations.

Iometer - 128KB Sequential Write

As with the random write performance, the sequential write performance is very low. At this point even a good mechanical hard drive would be much faster, though several other TLC drives and some smaller MLC drives also fall below that threshold.

Iometer - 128KB Sequential Write (Power)

Power consumption is almost halved relative to the Crucial BX200, but that's still not enough for the SP550's efficiency on this test to be considered good.

Performance and power consumption are slightly better at QD1 as the SP550 gets to burn through its SLC cache at the beginning of the test, but after that there's no scaling with queue depth.



Mixed Random Read/Write Performance

The mixed random I/O benchmark starts with a pure read test and gradually increases the proprotion of writes, finishing with pure writes. The queue depth is 3 for the entire test and each subtest lasts for 3 minutes, for a total test duration of 18 minutes. As with the pure random write test, this test is restricted to a 16GB span of the drive, which is empty save for the 16GB test file.

Iometer - Mixed 4KB Random Read/Write

Despite its poor random write speed, the SP550 keeps pace overall with its low-end competition on the mixed random I/O test.

Iometer - Mixed 4KB Random Read/Write (Power)

The SP550's power consumption is good but not great, so its efficiency ends up better than competitors like the OCZ Trion 100 and the SanDisk Ultra II, but the high-end drives are delivering performance that more than justifies their higher power consumption.

The drives that score best on the mixed random I/O test benefit from a big spike in performance in the write-heavy final stages, but the SP550 doesn't have that peak. Its performance variation across this test is relatively small, and the power consumption only grows modestly as the workload becomes more write-heavy.

Mixed Sequential Read/Write Performance

The mixed sequential access test covers the entire span of the drive and uses a queue depth of one. It starts with a pure read test and gradually increases the proprotion of writes, finishing with pure writes. Each subtest lasts for 3 minutes, for a total test duration of 18 minutes. The drive is filled before the test starts.

Iometer - Mixed 128KB Sequential Read/Write

Unlike the mixed random test, the SP550's poor write speeds really hurt it relative to the competition on the mixed sequential test and keep it near the bottom of the chart.

Iometer - Mixed 128KB Sequential Read/Write (Power)

The SP550's power consumption isn't out of control like the Crucial BX200's, but its efficiency is still poor.

Top performers on the mixed sequential test show a U-shaped performance curve and the winner is often determined by which drives suffer least during the middle stages of this test. The SP550 starts out great in the pure-read phase of the test, but it's all downhill from there.



ATTO

ATTO's Disk Benchmark is a quick and easy freeware tool to measure drive performance across various transfer sizes.

ATTO Performance

Most drives should maintain very steady performance once they hit their limit.  The uneven performance on the second half of the test is a sign that the SP550 has some performance problems that deserve investigation with the rest of our benchmarks. That said, the graphs aren't as inconsistent as the Crucial BX200's, where it was obvious that it had severe issues.

AS-SSD

AS-SSD is another quick and free benchmark tool. It uses incompressible data for all of its tests, making it an easy way to keep an eye on which drives are relying on transparent data compression. The short duration of the test makes it a decent indicator of peak drive performance.

Incompressible Sequential Read PerformanceIncompressible Sequential Write Performance

The AS-SSD test is short enough that the SP550 doesn't run out of SLC cache headroom and into trouble, and its peak performance is fine for a low-end drive.

Idle Power Consumption

Since the ATSB tests based on real-world usage cut idle times short to 25ms, their power consumption scores paint an inaccurate picture of the relative suitability of drives for mobile use. During real-world client use, a solid state drive will spend far more time idle than actively processing commands. Our testbed doesn't support the deepest DevSlp power saving mode that SATA drives can implement, but we can measure the power usage in the intermediate slumber state where both the host and device ends of the SATA link enter a low-power state and the drive is free to engage its internal power savings measures.

We also report the drive's idle power consumption while the SATA link is active and not in any power saving state. Drives are required to be able to wake from the slumber state in under 10 milliseconds, but that still leaves plenty of room for them to add latency to a burst of I/O. Because of this, many desktops default to either not using SATA Aggressive Link Power Management (ALPM) at all or to only enable it partially without making use of the device-initiated power management (DIPM) capability. Additionally, SATA Hot-Swap is incompatible with the use of DIPM, so our SSD testbed usually has DIPM turned off during performance testing.

Active Idle Power Consumption (No ALPM)

When aggressive link power measurement is disabled, the SP550 doesn't do a particularly good job of saving power, and is much worse than its Phison-based competition.

Idle Power Consumption (HIPM+DIPM)

However when the SP550's power saving modes can kick in, Silicon Motion's efficiency shines and the idle power drops to near the limit of our meter's resolution. The Phison-based TLC drives fall slightly behind and the drives that don't implement device-initiated power management stand out.



Final Words

Wrapping things up, I want to return to the broader question of what the ADATA SP550 does for the SM2256's reputation. In short, it helps a lot. The SM2256 clearly isn't the blockbuster that the SM2246EN controller was and isn't fully living up to the original expectations, but it does now have a place in the market.

To that end the ADATA Premier SP550 takes shape as a properly working implementation of an SM2256-based SSD, without the ugly surprises of the Crucial BX200. Write speeds - both sequential and random - are an acute weakness, though this admittedly is nothing new for a low-end TLC drives and at this point is becoming a known trade-off to reach lower prices. Unfortunately this also means that it still lags behind the prototype that used Samsung NAND and that it rarely pulls ahead of the pack of retail competitors. Meanwhile latency is consistently worse than drives based on other controller architectures that have multi-core processors.

The unexpected bright spot is that the SP550 exhibits relatively little performance degradation from a full drive. On the ATSB Light and Heavy tests the SP550's average data rate is lower than that of the OCZ Trion 100 when the test is run on an empty drive, but the SP550 comes out ahead when the drives are full. This effect also shows up when comparing the random write test against the steady-state performance: the SP550 is in last place on the shorter test but ranks much higher on the longer test. This is an important advantage for the 120GB model, which is the easiest to accidentally fill up. Since the 120GB SP550 is also the cheapest 120GB SSD at the moment, it's a reasonable pick for consumers trying to save every dollar possible.

At higher capacities it is much easier to justify spending a few dollars more for improved performance, and it's much easier to keep some free space on the drive. The 960GB model in particular won't make sense unless its price comes down substantially; that capacity was released more recently but probably can't compete against newer Phison drives, let alone the MLC-based Mushkin Reactor 1TB that is only $10 more. In the 480GB range the OCZ Trion 100 is priced very close and performs close albeit with different strengths and weaknesses. But again MLC drives like PNY's CS2211 aren't that much more expensive. Otherwise in the 240GB class there's not much that can compete with the SP550's price, and none of them can clearly beat it on performance.

Value SSD Price Comparison
Drive 960GB 480GB 240GB 120GB
ADATA SP550 $219.99 $112.99 $57.99 $38.99
PNY CS1311 $229.99 $119.99 $59.99 $39.99
OCZ Trion 100 $199.99 $114.99 $64.95 $54.99
OCZ Trion 150 $255.99 $133.49 $72.26 $52.96
Crucial BX200 $259.27 $128.50 $64.99  
SanDisk Ultra II $224.65 $129.99 $74.99 $54.99

Ultimately the bottom of the SSD market is crowded and it's hard to make a product stand out, especially as the limiting factor in both cost and performance is frequently the NAND. No value drive is without its faults and there's no clear top performer in this segment. But the ADATA SP550 does manage to stand out unambiguously with the best pricing at 120GB and 240GB while not earning notoriety for its performance. We can't reasonably demand more than that from a value drive.

As for the SM2256 controller, it's clear that Silicon Motion is learning to deal with the challenges of managing TLC flash, and many of the remaining weaknesses exhibited by the SP550 can be blamed on the limitations of the flash rather than the controller. For the impending transition to 3D NAND, the SM2246EN controller will be sticking around for MLC drives, but the SM2256 will be updated to SM2258 for 3D TLC. This suggests that Silicon Motion may be addressing more shortcomings in order to be ready to play a bigger role in the budget TLC SSD market with the next generation of drives.

Log in

Don't have an account? Sign up now