Original Link: https://www.anandtech.com/show/13421/the-mushkin-source-sata-ssd-review
The Mushkin Source 500GB SATA SSD Review: A Value Proposition For An Everyday PC
by Billy Tallis on November 21, 2018 10:00 AM ESTQLC SATA SSDs haven't quite hit the consumer market just yet, so the most affordable entry-level SSDs are still DRAMless designs with TLC NAND. The Mushkin Source is a current-generation example that provides only modest improvements over older DRAMless SSDs, but is adequate for most home systems.
Source: The Secret Sauce is $0.13/GB
The Mushkin Source is an entry-level SSD designed to offer the absolute lowest cost possible from current-generation components. The 500GB drive we're testing today is currently for sale at $65, putting the price per GB at a very low $0.13.
Specifications for our drive look pretty standard for a drive in this price range: peak sequential read speeds of 560 MB/s, writes at 520 MB/s, and 75k/81k IOPs respectively. Mushkin rates the drive at 2 W active and 80 mW idle, and with a three-year warranty that is typical for entry-level SSDs. The write endurance ratings are pretty low - for this 500GB drive that is 200TB total writes, which equates to a paltry 0.22 drive writes per day over those three years, equivalent to about 0.13 DWPD over five years (our usual metric).
Mushkin Source SSD Specifications | ||||||
AnandTech | 120 GB | 250 GB | 500 GB (Tested) |
1 TB | ||
Form Factor | 2.5" SATA or M.2 2280 SATA | |||||
Controller | Silicon Motion SM2258XT | |||||
NAND Flash | Micron 64L 3D TLC | |||||
Sequential Read | 510 MB/s | 560 MB/s | 560 MB/s | 560 MB/s | ||
Sequential Write | 440 MB/s | 515 MB/s | 520 MB/s | 520 MB/s | ||
Random Read | 29k IOPS | 54k IOPS | 75k IOPS | 78k IOPS | ||
Random Write | 79k IOPS | 81k IOPS | 81k IOPS | 81k IOPS | ||
Power | Active | 2 W | ||||
Idle | 80 mW | |||||
Warranty | 3 years | |||||
Write Endurance | 50 TB 0.22 DWPD |
100 TB 0.22 DWPD |
200 TB 0.22 DWPD |
400 TB 0.22 DWPD |
||
Current Retail | $27.99 (23¢ / GB) |
$41.99 (17¢ / GB) |
$64.99 (13¢ / GB) |
$135.99 (14¢ / GB) |
That endurance value is almost as low as what we're seeing from SSDs that use QLC NAND, and is likely due to a combination of the DRAMless controller making it harder to keep write amplification low, and Mushkin potentially using somewhat lower-grade flash in the Source. Nevertheless, the aim of this drive is primarily cost, and is aimed at a home user unlikely to do 100GB of writes per day every day.
How to Get Cheap before QLC Hits The Market
SATA SSDs with four-bit-per-cell QLC NAND flash memory are coming, but are not yet available in the consumer market and yield issues are preventing QLC from yet offering a meaningful discount relative to TLC NAND. That has led Mushkin to use a more familiar formula for the Source: 3D TLC NAND combined with a DRAMless SSD controller. Drives in this market segment always make sacrifices to reach the lowest price points, but they don't all make the same tradeoffs.
NAND flash memory prices have been in steep decline in recent months, and Mushkin has for the most part been able to keep pace with price drops for the Source. There have been occasions where more mainstream SATA drives with DRAM caches have gone on sale for lower prices than DRAMless drives, but the Mushkin Source has usually been quite close to being the cheapest SSD on the market from a reputable brand. Almost any SSD can garner some measure of success simply by being the cheapest. Being the cheapest *and* offering better than bottom-of-the-barrel specifications is harder to pull off.
The Mushkin Source uses Silicon Motion's SM2258XT controller, the DRAMless variant of the familiar SM2258. There is a newer SM2259 SATA controller but it doesn't offer any significant benefits for consumer SSDs, so the 2258XT can still be regarded as the current-generation solution from Silicon Motion. The Source uses Micron 64-layer 3D TLC NAND, which has proven to be much faster than their previous 32-layer 3D NAND and has been used in great products in all segments of the SSD market.
With only four NAND channels coming off the small SM2258XT controller and no external DRAM packages, the PCB inside the Mushkin Source is quite compact. Unusually, the SATA connector is not a discrete component attached to the edge of the circuit board but instead is formed from a combination of exposed contacts on the PCB itself and a plastic part to fill out the rest of the connector's shape. This may not be as durable as a discrete connector and certainly doesn't look as neat, but it works fine and probably saves a few cents. There's a token thermal pad between two of the NAND flash packages and the case, but is seems unlikely to have much effect on the drive's operation.
Mainstream SATA SSD product lines seldom offer a 120/128GB option these days due to the lower performance such small drives suffer from with only four NAND flash chips to work with. For entry-level products like the Mushkin Source, 120GB options are still relatively common even though they carry a significantly higher price per GB than 240GB and larger drives. As can be expected, the specifications for the smallest Mushkin Source reflect this lower performance, and the 250GB model improves on it but is still not up to full speed for things like random read throughput.
The primary competition for the Mushkin Source is other dirt-cheap SATA SSDs. Most of these are also DRAMless SSDs, usually based on Silicon Motion or Phison reference designs and sold by numerous brands. We have the Toshiba TR200 standing in for the Phison-based drives - availability of the TR200 has been inconsistent and pricing has often been far too high, but its performance is representative of cheaper products that use the same Phison S11 DRAMless controller with Toshiba 64L 3D TLC. The other DRAMless SSD in this review is the HP S700, which uses the same SM2258XT controller but older 32-layer Micron 3D TLC NAND.
It's also fair to compare the Source against mainstream SATA SSDs with DRAM caches, because the best deals in that market segment are usually within a few dollars of the going rate for DRAMless SSDs, and the more popular and thus higher-volume mainstream products occasionally end up undercutting DRAMless drives on price.
AnandTech 2018 Consumer 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 1709 |
Linux kernel version 4.14, fio version 3.6 | |
Spectre/Meltdown microcode and OS patches current as of May 2018 |
- 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
- Thanks to Quarch for the XLC Programmable Power Module and accessories
- Thanks to StarTech for providing a RK2236BKF 22U rack cabinet.
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 overall performance of the Mushkin Source on The Destroyer is a step backwards from other DRAMless SSDs we've tested recently, including the HP S700 that uses an older generation of Micron 3D TLC NAND.
The average and 99th percentile latencies from the Mushkin Source during The Destroyer are quite high, though not unprecedented given how other DRAMless SSDs have performed. The average latency from the Mushkin Source is slightly better than the Toshiba TR200, but the 99th percentile latency is in last place among this collection of drives.
The three DRAMless drives stand out with substantially higher average read latencies than the mainstream SATA SSDs, and the Mushkin Source is the worst of the three. Average write latencies are very high for the Mushkin Source and the Toshiba TR200, but the HP S700 is only slightly worse off than the slower mainstream drives with DRAM caches.
The DRAMless drives have much higher 99th percentile write latencies than the mainstream SATA drives, but the Mushkin Source's 233ms is not as bad as the 327ms from the TR200. For 99th percentile read latency, the Mushkin Source is in last place and the other two DRAMless drives have scores that are competitive with many of the mainstream drives.
The Mushkin Source requires more energy than most SATA drives to complete The Destroyer, but some of the older mainstream drives that also have generally poor performance use even more energy. The other two DRAMless SSDs have average or slightly above average efficiency on this test despite their low performance.
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 an empty drive, the Mushkin Source offers performance that is not too far behind many mainstream SATA drives. When the test is run on a full drive, the story is very different, and the Source's average data rate is lower than any of the other drives in this bunch, including the other two DRAMless SSDs.
The average and 99th percentile latencies from the Mushkin Source during the Heavy test are reasonable when the test is run on an empty drive, but when the drive is full both scores are worse than either of the other DRAMless SSDs in this bunch.
The average read and write latencies of the Mushkin Source are both severely inflated when the drive is full, beyond what the other DRAMless SSDs suffer. But unlike the Toshiba TR200, the Mushkin Source handles both reads and writes well when the drive isn't full.
When the Heavy test is run on a full drive, the Mushkin Source has far worse read QoS than any of the other drives; even the HP S700's 99th percentile read latency is no more than twice as high when the test drive is full, but the Mushkin Source's 99th percentile latency gets an order of magnitude worse. For 99th percentile write latency, the Mushkin Source isn't much worse off than other DRAMless SSDs, and at least it does not have the empty-drive performance problems of the TR200.
The disparity in performance between full and empty drive test runs carries through to the energy usage measurements. When the Heavy test is run on an empty drive and the Mushkin Source is able to make good use of its SLC cache, the power efficiency is also very good. When the drive is full and bogged down with too much background garbage collection, the energy usage is significantly worse than the rest of the drives in this review.
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 average data rate from the Mushkin Source running the Light test is a bit slower than most mainstream SSDs, but the performance deficit is only severe when the test is run on a full drive.
The average and 99th percentile latency scores for the Mushkin Source are typical for SATA SSDs when the Light test is run on an empty drive, but when the drive is full both latency measures are out of control (though not quite the worst we've seen).
When the Mushkin Source is full, the average write latency degrades much more than the average read latency, which merely doubles rather than growing by a factor of four.
The 99th percentile read and write latencies from running the Light test on a completely full Mushkin Source are both higher than even the other DRAMless SSDs. The 99th percentile write latency grows much more than read score, but both are stark outliers.
In spite of the poor performance when full, the Mushkin Source manages to have reasonable power efficiency when running the Light test on a full drive. When testing an empty drive, the Mushkin Source is significantly more efficient than the mainstream SATA SSDs that must power a DRAM cache.
Working Set Size
DRAMless SSDs may still have a bit of SRAM on the SSD controller to cache a portion of the flash translation layer's mapping tables. Just like a GB of external DRAM can hold the mapping tables for a TB of flash, a few MB of onboard RAM in the controller can hold the mappings for a few GB of flash. We can probe for such a cache by performing random reads with a varying working set size.
The Mushkin Source shows some decline in random read performance as the working set size grows beyond just a few GB, but there's no sharp drop in performance like we see with DRAMless NVMe drives that are using the Host Memory Buffer feature. Most of the other drives show roughly the same random read performance regardless of working set size.
Whole-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 Mushkin Source lasts a very long time before its SLC cache is filled. While the Toshiba TR200's write speed ends up in the gutter very quickly, it will be almost impossible for a real-world consumer workload to fill the cache of the Source unless the drive starts out nearly full. From an empty drive, the apparent initial SLC cache size is well over 150GB—essentially the entire drive operating as SLC until free space runs out.
Average Throughput for last 16 GB | Overall Average Throughput |
In the unlikely event that the SLC write cache on the Mushkin Source should overflow, its sequential write speed drops to the slowest in its class at just over 70 MB/s—slower than most mechanical hard drives. But this is only relevant when writing tens of GB at once.
BAPCo SYSmark 2018
BAPCo's SYSmark 2018 is an application-based benchmark that uses real-world applications to replay usage patterns of business users, with subscores for productivity, creativity and responsiveness. Scores represnt overall system performance and are calibrated against a reference system that is defined to score 1000 in each of the scenarios. A score of, say, 2000, would imply that the system under test is twice as fast as the reference system.
SYSmark scores are based on total application response time as seen by the user, including not only storage latency but time spent by the processor. This means there's a limit to how much a storage improvement could possibly increase scores, because the SSD is only in use for a small fraction of the total test duration. This is a significant difference from our ATSB tests where only the storage portion of the workload is replicated and disk idle times are cut short to a maximum of 25ms.
AnandTech SYSmark SSD Testbed | |
CPU | Intel Core i5-7400 |
Motherboard | ASUS PRIME Z270-A |
Chipset | Intel Z270 |
Memory | 2x 8GB Corsair Vengeance DDR4-2400 CL17 |
Case | In Win C583 |
Power Supply | Cooler Master G550M |
OS | Windows 10 64-bit, version 1803 |
Our SSD testing with SYSmark uses a different test system than the rest of our SSD tests. This machine is set up to measure total system power consumption rather than just the drive's power.
For the Creativity and Productivity subtests, the Mushkin Source and other DRAMless SSDs offer performance indistinguishable from a mainstream SATA SSD with a DRAM cache; these scenarios are more bottlenecked by CPU power and system RAM. The responsiveness test does reveal some of the reduced performance of DRAMless SSDs, but even so they greatly outperform a mechanical hard drive.
Energy Use
The SYSmark energy use scores measure total system power consumption, excluding the display. Our SYSmark test system idles at around 26 W and peaks at over 60 W measured at the wall during the benchmark run. SATA SSDs seldom exceed 5 W and idle at a fraction of a watt, and the SSDs spend most of the test idle. This means the energy usage scores will inevitably be very close. A typical notebook system will tend to be better optimized for power efficiency than this desktop system, so the SSD would account for a much larger portion of the total and the score difference between SSDs would be more noticeable.
The DRAMless SSDs use a bit less energy during a SYSmark run than the mainstream Crucial MX500, but the HP S700 is the only one that really stands out as saving a non-trivial amount of power, and even in that case the difference is small enough to only matter for low-power notebooks, not our desktop testbed.
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 burst random read performance from the Mushkin Source is low, as expected from a DRAMless SSD. The Crucial MX500 manages almost twice the performance with the same NAND flash. Compared to other DRAMless SSDs, the Toshiba TR200 has the clear lead over the Mushkin Source, but the Source is an improvement over the HP S700's older 32L TLC.
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 brings in some slightly higher queue depths, the Mushkin Source pulls ahead of the other DRAMless SSDs but is still unable to match the mainstream SATA SSDs with full-size DRAM caches.
Power Efficiency in MB/s/W | Average Power in W |
The Mushkin Source uses substantially more power during the random read test than the other two DRAMless SSDs, so it has one of the lowest efficiency scores while the Toshiba TR200 manages a slightly above-average score.
The random read performance of the Mushkin Source does scale with increasing queue depth (especially past QD2), but the mainstream SATA SSDs greatly outpace it and several end up with almost twice the performance at high queue depths.
Comparing the Mushkin Source against all the other SATA SSDs that have completed this test, it is clear that the Source doesn't come close to reaching the limits of the SATA interface, and its power consumption at high queue depths is rather high for the performance it enables.
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 burst random write performance of the Mushkin Source is a bit on the slow side, but the DRAMless SSDs don't stand out as extreme outliers thanks to their reasonably quick SLC write caches.
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, the difficulty of managing an overflowing SLC cache on a DRAMless drive is readily apparent, and the Mushkin Source suffers more than the other two DRAMless SSDs. The mainstream SATA SSDs are 3-4 times faster than the Source.
Power Efficiency in MB/s/W | Average Power in W |
The Mushkin Source and other DRAMless SSDs do use significantly less power than the mainstream SATA SSDs on the random write test, but the savings are nowhere near large enough to make up for the huge performance deficit. The Source is both slower and more power-hungry than the other DRAMless SSDs, so it comes in last place for efficiency.
The random write performance of the Mushkin Source does not improve appreciably with increasing performance, because each phase of this test writes more than enough data to exhaust the SLC cache. The Toshiba TR200 DRAMless SSD does manage a significant increase from QD1 to QD8, but even so it offers less than half the performance of a good mainstream drive.
Plotted against all the other SATA SSDs that have completed this test, it is clear that the Mushkin Source is one of the slowest, and other drives have offered much higher performance within a lower power budget.
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%.
The Mushkin Source turns in respectable performance on the burst sequential read test, and the DRAMless design is clearly not a handicap at all for this workload.
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. This test is run twice: once with the drive prepared by sequentially writing the test data, and again after the random write test has mixed things up, causing fragmentation inside the SSD that isn't visible to the OS. These two scores represent the two extremes of how the drive would perform under real-world usage, where wear leveling and modifications to some existing data will create some internal fragmentation that degrades performance, but usually not to the extent shown here.
On the longer sequential read test, the Mushkin Source maintains high performance both for the initial test run reading back data that was written sequentially, and when reading from a dataset that was fragmented by the random write tests.
Power Efficiency in MB/s/W | Average Power in W |
With sequential read performance that is competitive against mainstream drives and a slight power consumption advantage from not having to power DRAM, the Mushkin Source turns in very good power efficiency scores, though not quite as high as the HP S700 managed.
Like any SATA SSD, the Mushkin Source can't quite saturate the SATA interface at QD1. At any higher queue depth, the Source delivers steady full-speed sequential reads.
While the random access performance and power consumption of the Mushkin Source may not have looked good in the broader picture, for sequential reads the Mushkin Source is clearly as fast as any other SATA drive and draws less power than almost any other.
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 performance of the Mushkin Source isn't quite top-tier, but it is comparable to most mainstream SATA SSDs.
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.
On the longer sequential write test, the Mushkin Source's standing falls somewhat, but it still offers decent performance, unlike the Toshiba TR200.
Power Efficiency in MB/s/W | Average Power in W |
Every other SATA SSD in this batch uses at least as much power during this test as the Mushkin Source, which easily tops the efficiency ranking.
The Mushkin Source doesn't reach its full sequential write speed until QD4, which is later than some drives but not uncommon even among mainstream SATA SSDs. The Source maintains that performance through the remainder of the test, with no sign of the SLC cache overflowing.
The Mushkin Source may not quite reach the highest sequential write speeds possible for a SATA SSD, but it gets pretty close and is obviously the most efficient SATA drive we've put through this test. That's a strange win, but we'll take it.
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 Mushkin Source does not handle the mixed random I/O test well at all, with performance that is barely faster than the HP S700 and much slower than the Toshiba TR200, let alone the mainstream SATA SSDs that have a DRAM cache.
Power Efficiency in MB/s/W | Average Power in W |
The Mushkin Source uses a bit more power than the HP S700 on this test, so it ends up falling to last place on the power efficiency ranking. The Toshiba TR200 uses less power than the other two DRAMless drives and substantially outperforms them, so its efficiency is close to competitive against the mainstream drives.
The Mushkin Source displays a typical bathtub curve for performance on the mixed random I/O test. With pure reads or pure writes, it is substantially faster than with any mix, and the lowest performance is with around a 50/50 mix of reads and writes. Most mainstream SATA SSDs these days show significant performance growth throughout the test as the workload becomes more write-heavy, finishing with a significant spike in performance at the end when the SLC cache can be used to the greatest extent possible.
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 Mushkin Source comes close to the performance of mainstream SATA SSDs on the mixed sequential I/O test, but it is still handicapped slightly by the DRAMless controller despite having turned in competitive scores in the tests of pure sequential reads and writes.
Power Efficiency in MB/s/W | Average Power in W |
Unsurprisingly, the Mushkin Source continues its trend of very good power efficiency during sequential I/O, ranking just below the HP S700 and above all the mainstream SATA SSDs.
The Mushkin Source's behavior on the mixed sequential I/O test is quite similar to the HP S700: a steep decline in performance during the first half of the test as the workload shifts from pure reads to a 50/50 mix, followed by a moderate performance recovery. The mainstream SATA SSDs that show higher overall performance win out during the middle phases of the test, where they lose less performance as writes are introduced.
Idle Power Measurement
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 if supported.
The active idle power consumption of the Mushkin Source seems a bit high given the lack of external DRAM to power, but in the slumber state it is quite competitive against other SATA drives.
The Mushkin Source has a notably slow wake-up from sleep, taking more than 2.6ms when a typical time for SATA SSDs is around 1ms. The Toshiba TR200 also has a similarly long wake-up latency, and the HP S700 likely only avoids that by not having a working slumber state at all.
Conclusion: Cheap, But Don't Fill It
Mushkin has been basing its SSDs off Silicon Motion controllers for years, and they have produced a few hits when they have managed to deliver a good-performing SMI platform at great prices. That was clearly the case with the last Mushkin drive we reviewed, the Reactor. The Mushkin Source isn't one of those clear winners. Instead, it's a cheap drive that makes some compromises and comes with some caveats.
The Mushkin Source provides about what we expect from an entry-level SATA SSD. The choice of a DRAMless SSD controller means that random I/O performance lags well behind mainstream SATA SSDs, but sequential I/O performance is for the most part competitive. The lack of DRAM lowers power consumption a bit, so for sequential workloads where the Source performs well, the power efficiency tends to be better than most mainstream SATA SSDs, but the power savings aren't enough to match the much lower performance on workloads with random I/O.
Like any other consumer SSD using TLC NAND, the SLC write caching used by the Mushkin Source is an important part of the overall performance profile. Without it, write performance would be unacceptable, but the cache on the 500GB Source we tested is adequately sized to handle most of our tests and most real-world consumer workloads. For the few tests that wrote enough data to overflow the SLC cache, the effects can be seen as substantially lowered write performance and worse read performance as the controller is too busy shuffling data around. The worst-case performance of the Mushkin Source is just as bad as other DRAMless SSDs we've tested, which is a bit of a disappointment given how mainstream SATA and NVMe SSDs have shown Micron's 64L 3D TLC to be significantly faster than their first generation 3D NAND. The Mushkin Source doesn't offer the kind of generational improvement we wanted to see over older DRAMless drives like the HP S700.
For lighter consumer workloads that don't routinely throw around multi-GB datasets, the DRAMless design of the Mushkin Source is a minor hindrance, not a crippling flaw. It's slower than mainstream SATA SSDs, but still far faster than a hard drive for both random and sequential I/O. When subjected to a more intense workload, the Source can offer adequate performance for a while before the SLC cache runs out and latency goes through the roof. Given its entry-level positioning, the Source is quite capable of handling the kind of workloads it can be reasonably expected to encounter.
120GB | 240-256GB | 480-525GB | 1TB | 2TB | |
Mushkin Source |
$27.99 (23¢/GB) | $41.99 (17¢/GB) | $64.99 (13¢/GB) | $135.99 (14¢/GB) | |
Toshiba TR200 |
$39.99 (17¢/GB) | $79.99 (17¢/GB) | $279.99 (29¢/GB) | ||
Crucial BX500 |
$26.77 (22¢/GB) | $42.95 (18¢/GB) | $76.95 (16¢/GB) | ||
Crucial MX500 |
$52.99 (21¢/GB) | $74.99 (15¢/GB) | $139.99 (14¢/GB) | $326.37 (16¢/GB) | |
Samsung 860 EVO | $57.99 (23¢/GB) | $72.99 (15¢/GB) | $127.98 (13¢/GB) | $294.88 (15¢/GB) | |
WD Blue 3D NAND | $53.65 (21¢/GB) | $79.99 (16¢/GB) | $134.99 (13¢/GB) | $311.00 (16¢/GB) |
Black Friday and Cyber Monday sales are just days away (and in some cases, already underway), but there's a chance that this may not actually be the best week out of this year to buy a new SSD. NAND flash memory prices have been falling fast in recent months and that trend is set to continue well into 2019. A record-low price per GB this week could be the everyday going rate in a few weeks or months. In addition to this volatility, it appears that the sales are causing major online retailers to run out of stock of some of the SSDs they are offering the best deals on. So any assessment of the current market conditions and the pricing on the Mushkin Source will have a short shelf life.
But as things stand now, the Mushkin Source doesn't stack up very well against the best prices for mainstream SATA SSDs. The clearest niche for drives like the Source is at the 120GB capacity point that the mainstream drives have abandoned. At larger capacities, the Source is either more expensive than something like a Samsung 860 EVO, or jumping up to an 860 EVO or MX500 is only a few well-justified dollars more. Unless price per GB is the ultimate goal at the 500GB, then the Source is difficult to recommend. The annoying thing is, the performance when full means that users get the best when they do not fill it, so if the 'useable capacity' is reduced in this way, then price per GB goes up. That's when the MX500 comes into play at 500GB, or the 860 EVO at 1TB.
The Future Value is in QLC
Looking forward, DRAMless SATA drives like the Mushkin Source will be competing against the coming wave of QLC SATA SSDs. Many of these should be able to include DRAM caches and still offer lower prices per GB than DRAMless TLC SSDs. At high capacities (1TB+), QLC SATA drives will probably destroy the market for DRAMless TLC SATA SSDs. At smaller capacities, QLC drives will suffer even more severely from the performance limitations of only having a few NAND flash dies to use in parallel, and may not be able to beat a DRAMless TLC drive.