Original Link: https://www.anandtech.com/show/14601/the-kingston-kc2000-ssd-review
The Kingston KC2000 SSD Review: Bringing BiCS4 To Retail
by Billy Tallis on July 22, 2019 8:00 AM ESTThe Kingston KC2000 is a new high-end consumer M.2 NVMe SSD. Kingston had been neglecting this market segment for a while, but the two year old KC1000 has finally been replaced by a proper successor using the latest flash memory and controllers.
The SSD market is in the midst of two significant-but-protracted transitions. First off, PCIe 4.0 has arrived in one small corner of the market, but so far it offers little aside from increased power usage. Meanwhile the second transition is the move from 64-layer to 96-layer 3D NAND; this is a bit more widespread, but is still moving slowly. Toshiba started shipping their 96L NAND almost a year ago, but in the first half of 2019 we saw more new retail SSDs launch with 64L NAND than with 9xL NAND. Kingston's not the first to use 96-layer 3D TLC in a retail drive, but they're still generally ahead of the curve and have even beat Toshiba and Western Digital to market using their BiCS4 96L NAND.
The Kingston KC2000 illustrates a broader trend of diversification that we've been seeing signs of with this new generation of flash memory. SSD vendors that don't make their own controllers or NAND are exploring a wider range of options than in recent years. With 32-layer and 64-layer NAND, there was a strong association between Silicon Motion SSD controllers and Intel/Micron NAND, while at the same time Toshiba's 15nm and 64L NAND were mostly used with Phison SSD controllers. (Marvell's presence in the retail consumer SSD market has dropped significantly, and newer controller suppliers like Realtek haven't had much impact.) With the 96-layer generation, Silicon Motion and Phison are both fully supporting multiple NAND vendors, and SSD vendors are trying out the new options.
Kingston has been a fan of the Phison+Toshiba combinations for quite a while, but the KC2000 pairs Toshiba's 96L NAND with the Silicon Motion SM2262EN controller instead of the Phison E12 or E16 we would have predicted. That gives us the chance to make a very direct comparison between Micron 64L TLC and Toshiba 96L TLC, since we have tested several drives using the SM2262EN controller with the Micron NAND. We already have some idea how Toshiba's 96L NAND compares against their own previous generation thanks to our testing of the XG5 and XG6, but those are OEM-only drives with performance constrained by a somewhat outdated controller. The Kingston KC2000 is one of the first drives to bring Toshiba's 96L NAND to the retail market, and it will be followed by many others over the next year.
Kingston KC2000 SSD Specifications | ||||||
Capacity | 250 GB | 500 GB | 1 TB | 2 TB | ||
Form Factor | M.2 2280 PCIe 3.0 x4 | |||||
Controller | Silicon Motion SM2262EN | |||||
NAND Flash | Toshiba 96L BiCS4 3D TLC | |||||
Sequential Read | 3.0 GB/s | 3.0 GB/s | 3.2 GB/s | 3.2 GB/s | ||
Sequential Write | 1.1 GB/s | 2.0 GB/s | 2.2 GB/s | 2.2 GB/s | ||
Random Read IOPS | 350k | 350k | 350k | 250k | ||
Random Write IOPS | 200k | 250k | 275k | 250k | ||
Power Consumption | 3 mW Idle, 2.1 W Max Read, 7 W Max Write | |||||
Encryption | AES 256, TCG Opal 2.0, eDrive | |||||
Warranty | 5 years | |||||
Write Endurance | 150 TB 0.3 DWPD |
200 TB 0.3 DWPD |
600 TB 0.3 DWPD |
1200 TB 0.3 DWPD |
The Silicon Motion SM2262EN controller is a very familiar chip at this point, especially this late in the product cycle. It's a minor update over last year's SM2262 that mostly brings firmware optimizations for improved peak performance. When paired with Micron 64L NAND it has repeatedly proven capable of delivering very high performance for a drive with a PCIe 3 x4 interface., and it's unlikely to be the bottleneck when used with Toshiba's 96L NAND. Kingston's specs for the KC2000 agree: most of the performance ratings are a bit lower than the competition using Micron NAND. The power consumption specs suggest that we may see some of the impressive efficiency that Toshiba/WD BiCS NAND has been providing in first-party drives like the Toshiba XG series and the recent WD Black models.
The Kingston KC2000 lineup spans from 250GB to 2TB. That smallest capacity point has started to disappear from some flagship high-end product lines because it isn't able to deliver the same level of performance that larger models can. For the KC2000, Kingston is keeping the smallest and most affordable capacity around, but with significantly slower write speed ratings.
The 5-year warranty and 0.3 DWPD write endurance ratings for the KC2000 are standard for this market segment. The support for TCG Opal and eDrive encryption are still relatively uncommon among retail NVMe drives, and probably result from Kingston's strong focus on the business market.
Kingston uses a double-sided layout for the KC2000. The DRAM and NAND were binned and packaged by Kingston, and they generally prefer to use more packages with fewer dies each. This means we have eight total NAND packages of 128GB (1024Gb) each and two DRAM packages of 512MB each on our 1TB sample; a tight fit overall, especially compared to the loose spacing seen on previous SM2262EN drives.
The Competition
The most interesting comparison for the Kingston KC2000 is another drive using the same Silicon Motion SM2262EN controller. For this review, that role is filled by the ADATA SX8200 Pro. The Toshiba XG6 is the only other drive we have so far that uses Toshiba 96L TLC, but it's an OEM-only product. The Silicon Power P34A80 is the representative from the wide field of products that combine Toshiba's 64L TLC with the Phison E12 controller. Also from the Phison camp is Kingston's entry-level NVMe drive, the A1000 that uses the Phison E8 controller and Toshiba 64L TLC. We also include results for flagship drives from Intel, Samsung and Western Digital.
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 HD 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 average data rate from the Kingston KC2000 on The Destroyer is slower than any of the competing drives we've tested, though the ADATA SX8200 Pro that uses the same SM2262EN controller is only slightly faster. The KC2000 is still almost twice as fast overall as the SATA and entry-level NVMe drives.
The KC2000 falls behind other high-end drives in terms of latency on The Destroyer, but the average latency is still within reason. The 99th percentile latency is several times higher than it should be.
The average read latency of the KC2000 on The Destroyer is decent and outperforms several other high-end drives. The average write latency is significantly worse than the competition, but still well ahead of the entry-level NVMe drives.
Breaking down the 99th percentile scores, the KC2000 again handles reads well, but has much higher write latency than is typical for today's high-end NVMe drives. The ADATA SX8200 Pro that uses the same controller with different NAND scores even worse for both QoS metrics.
The Kingston KC2000's energy consumption during The Destroyer makes it the most power-hungry drive out of the several here that use Toshiba/WD BiCS NAND. However, compared against the broader field, it still provides about average efficiency, comparable to the ADATA SX8200 Pro that uses the same controller but Micron NAND.
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.
The overall performance from the Kingston KC2000 on the Heavy test is disappointing. Its performance when full isn't quite as bad as the other drives that use recent Silicon Motion controllers with aggressive SLC caching, but the empty-drive performance is less than half what the ADATA SX8200 Pro provides. The KC2000 doesn't seem to be making tradeoffs to handle one case better than the other; it's just slow either way.
The overall average latency scores for the KC2000 on the Heavy test are worse than expected for a high-end NVMe drive and are pushing into entry-level NVMe territory. The 99th percentile latency scores are a bigger problem, since the KC2000 falls behind even mainstream SATA drives.
Breaking down the average latency scores, the KC2000 is competitive with read latency, though its read latency is a bit high when the test is run on a full drive. The average write latency scores are generally the worst among high-end NVMe drives, but are not a serious problem.
The poor 99th percentile latency scores for the KC2000 are due entirely to its behavior for writes, where it delivers worse QoS than a decent SATA drive (though not as bad as a full QLC drive). The QoS for read operations is competitive with many of the best TLC drives on the market.
The energy usage of the Kingston KC2000 over the course of the Heavy test is a little bit on the high side of normal, but isn't an outlier like the Samsung drives. Like the other two Silicon Motion drives, the KC2000's power requirements are notably higher when the test is run on a full drive, but the impact isn't quite as large on the KC2000 as it is for the ADATA drive that uses Micron NAND instead of Toshiba's 96L TLC.
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 overall performance of the Kingston KC2000 on the Light test is another disappointment, since it is basically the same speed as last year's entry-level NVMe drive from Kingston that used the less powerful Phison E8 controller and an older generation of Toshiba NAND. The KC2000 handles a full drive better than other recent Silicon Motion drives, but even in that worst-case scenario it's still substantially slower than most high-end NVMe drives.
The average latencies from the KC2000 during the Light test are a bit high compared to most high-end drives, but it's quick enough to not be a problem for lighter workloads. The 99th percentile latency is fine when the Light test is run on an empty drive, but when the drive is full it starts to stutter more than a decent SATA drive.
Splitting the average latencies by reads and writes, we see that both write latency scores for the KC2000 are a bit on the slow side for something aspiring to be a high-end drive, while the read latency is very competitive for the empty-drive test run and only falls a bit behind when the drive is full.
Breaking down the 99th percentile latency scores reveals where the KC2000 really gets into trouble: when dealing with a full drive and the unavoidable pressure of background work, the KC2000's read QoS suffers with 99th percentile read latencies jumping to several milliseconds—close to hard drive seek times. This is a known issue for the Silicon Motion SM2262EN controller, which doesn't seem to be very good at interrupting background work to quickly handle more important reads. Fortunately, the 99th percentile write latency is nowhere near as bad as we've seen from drives like the ADATA SX8200 Pro.
The Kingston KC2000 doesn't win any prizes for energy efficiency during the Light test. When the test is run on an empty drive the energy usage is decent but like the other Silicon Motion drives it gets more power hungry when the drive is full and there's more background work to be done. Even in that case, it is more efficient than Samsung's drives, which burn a lot of power to offer performance that simply doesn't matter on a light workload like this test.
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 Kingston KC2000 continues the trend of drives using Silicon Motion's NVMe controllers delivering top-notch burst random read performance, but the ADATA SX8200 Pro that uses Micron 64L TLC is still faster than the KC2000's 96L Toshiba NAND.
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, the Samsung 970 PRO catches up to the KC2000 but otherwise the rankings and spread of scores are similar to the burst random read results.
Power Efficiency in MB/s/W | Average Power in W |
The KC2000 turns in an excellent power efficiency score for random reads, but is still outclassed by the ADATA SX8200 Pro. The KC2000 manages a slightly higher efficiency score than the WD Black SN750, which is quite a bit slower than the KC2000 on this test but also uses less power than any of the other NVMe drives in this bunch.
Several of the competing drives are able to eventually outperform the Kingston KC2000 for random reads, given a sufficiently high queue depth. The WD Black and Toshiba XG6 manage to do so while drawing much less power, so the KC2000's excellent efficiency score on this test only holds up at low queue depths.
Comparing the KC2000's random read test results against our entire database of results shows that its peak peak performance is good but not quite up to the limits of what TLC drives can deliver, and the power consumption at higher queue depths is on the high side.
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 Kingston KC2000 is pretty good, but is overshadowed by the Phison E12-based Silicon Power drive that has a wide lead over the rest of the high-end drives thanks to a very fast SLC write cache.
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 KC2000 falls slightly behind most top NVMe drives but still performs very well.
Power Efficiency in MB/s/W | Average Power in W |
The power efficiency of the KC2000 during random writes is top tier, but drives with the same NAND and different controller or vice versa are slightly ahead of the KC2000.
At queue depths of 4 or higher, the high-end drives offer roughly similar performance but power consumption varies significantly. Of the drives that match the power consumption of the KC2000, the ADATA SX8200 Pro is consistently a little bit faster, while the Toshiba XG6 underperforms at low QDs and surpasses the KC2000 at high queue depths.
Comparing the KC2000's random write results against the entire database shows that the power efficiency of the KC2000 is quite good, with only a handful of scores from other drives offering similar performance at slightly lower power levels.
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 burst sequential read performance of the Kingston KC2000 is excellent, though slightly slower than the same SSD controller manages when paired with Micron 64L TLC rather than the Toshiba 96L TLC that Kingston is using.
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 that brings in some higher queue depths, the KC2000's performance doesn't stand out from other top TLC NVMe drives. The Samsung 970 EVO Plus is roughly 5-15% faster, but you have to step up to an MLC or Optane drive to see a huge boost to the worst-case fragmented data performance.
Power Efficiency in MB/s/W | Average Power in W |
The power efficiency of The Kingston KC2000 during the sequential read test is not quite top tier, but it is still acceptable for a high-end NVMe SSD. In absolute terms, the KC2000 is one of the most power-hungry M.2 drives in this bunch, and it doesn't quite have enough performance to match.
The Kingston KC2000 is fairly well-behaved across the range of queue depths tested, reaching full speed at QD4 or higher. However, it is always a bit slower and more power-hungry than the ADATA SX8200 Pro.
Comparing the KC2000 against the entire library of test results shows that it mostly offers the performance expected from a PCIe 3 x4 drive, but doesn't quite saturate the link as well as the competition, and its power efficiency is nothing special.
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 Kingston KC2000 is much slower than the fastest high-end NVMe SSDs offer, but still acceptable for this market segment. The SSD controller clearly isn't the bottleneck since the ADATA SX8200 Pro is faster by over 200MB/s, and judging by the WD Black it appears the 96L BiCS4 TLC isn't really any faster than the 64L BiCS3 TLC.
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 KC2000 again isn't the fastest, but it holds up better than the Phison E12 drive and doesn't lose much ground relative to the other SM2262EN drive.
Power Efficiency in MB/s/W | Average Power in W |
The power efficiency of the Kingston KC2000 during the sequential write test is good, but a bit lower than the fastest drives that only require slightly more power.
The Kingston KC2000 hits its full sequential write speed at QD2 and has no trouble maintaining that speed for the rest of the test thanks to the large SLC cache. However, it's still a much slower SLC write speed than Intel/Micron and Samsung NAND offers.
Plotting the KC2000's sequential write results against the broader landscape shows that the KC2000 doesn't make it very far into true high-end performance territory, and its power consumption is a bit high for the maximum speeds it does attain.
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 Kingston KC2000 performs well on the mixed random IO test. It's a bit slower overall than the ADATA SX8200 Pro that uses the same controller, but still fast enough for this market segment.
Power Efficiency in MB/s/W | Average Power in W |
The power efficiency of the KC2000 during the mixed random IO test is second-tier, falling behind the Toshiba and WD drives that also use BiCS TLC and the ADATA SX8200 Pro that uses Micron NAND with the same SM2262EN controller as the KC2000.
The ADATA SX8200 Pro with Micron NAND earns a higher overall score than the KC2000 largely due to better performance on the more read-heavy half of the test, while the KC2000 mostly catches up during the write-intensive portions. The Samsung 970 EVO Plus by contrast earns its advantage primarily from better performance during the write-heavy half 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.
On the mixed sequential read/write test, the Kingston KC2000 is slower than most high-end NVMe SSDs, though it does significantly outperform the Toshiba XG6 that relies on the same 96L TLC NAND.
Power Efficiency in MB/s/W | Average Power in W |
The subpar performance of the KC2000 on the mixed sequential test carries over to its efficiency score. The KC2000's power draw is about average, but it doesn't deliver quite as much performance within that power envelope.
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 SLC cache behavior of the Kingston KC2000 is quite similar to the other SM2262EN-based drive, the ADATA SX8200 Pro. Both have large caches of a little over 150GB. The performance levels differ slightly: the KC2000's 96L BiCS can handle only about 2.2GB/s of writes to its SLC cache while the ADATA's Micron 64L TLC in SLC mode is just shy of 3GB/s. Once the caches are filled and the drives start writing direct to TLC, the 96L BiCS4 TLC is a bit faster than the Micron 64L TLC. Performance drops again when the TLC starts to fill up and space has to be reclaimed from the SLC cache.
Average Throughput for last 16 GB | Overall Average Throughput |
The average write speed from the KC2000 for the last 16GB and the overall average for the entire drive fill are both subpar for a high-end drive, but faster than the ADATA SX8200 Pro–which was tied for the fastest SLC cache at the beginning of the test.
Power Management Features
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.
For many NVMe SSDs, the closely related matter of thermal management can also be important. M.2 SSDs can concentrate a lot of power in a very small space. They may also be used in locations with high ambient temperatures and poor cooling, such as tucked under a GPU on a desktop motherboard, or in a poorly-ventilated notebook.
Kingston KC2000 1TB NVMe Power and Thermal Management Features |
|||
Controller | Silicon Motion SM2262EN | ||
Firmware | S2681101 | ||
NVMe Version |
Feature | Status | |
1.0 | Number of operational (active) power states | 3 | |
1.1 | Number of non-operational (idle) power states | 2 | |
Autonomous Power State Transition (APST) | Supported | ||
1.2 | Warning Temperature | 75 °C | |
Critical Temperature | 80 °C | ||
1.3 | Host Controlled Thermal Management | Supported | |
Non-Operational Power State Permissive Mode | Not Supported |
The power management features supported by the Kingston KC2000 match other drives using the SM2262EN controller, and the configuration of power states is identical. The firmware still promises excellent idle power savings and very quick transition latencies.
Kingston KC2000 1TB NVMe Power States |
|||||
Controller | Silicon Motion SM2262EN | ||||
Firmware | S2681101 | ||||
Power State |
Maximum Power |
Active/Idle | Entry Latency |
Exit Latency |
|
PS 0 | 9.0 W | Active | - | - | |
PS 1 | 4.6 W | Active | - | - | |
PS 2 | 3.8 W | Active | - | - | |
PS 3 | 45 mW | Idle | 2 ms | 2 ms | |
PS 4 | 4 mW | Idle | 6 ms | 8 ms |
Note that the above tables reflect only the information provided by the drive to the OS. The power and latency numbers are often very conservative estimates, but they are what the OS uses to determine which idle states to use and how long to wait before dropping to a deeper idle state.
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, and depending on which NVMe driver is in use. Additionally, there are multiple degrees of PCIe link power savings possible through Active State Power Management (APSM).
We report three 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. Our Desktop Idle number represents what can usually be expected from a desktop system that is configured to enable SATA link power management, PCIe ASPM and NVMe APST, but where the lowest PCIe L1.2 link power states are not available. The Laptop Idle number represents the maximum power savings possible with all the NVMe and PCIe power management features in use—usually the default for a battery-powered system but rarely achievable on a desktop even after changing BIOS and OS settings. Since we don't have a way to enable SATA DevSleep on any of our testbeds, SATA drives are omitted from the Laptop Idle charts.
Note: We recently upgraded our power measurement equipment and switched to measuring idle power on our Coffee Lake desktop, our first SSD testbed to have fully-functional PCIe power management. The below measurements are all new, and are not a perfect match for the older measurements in our previous reviews and the Bench database.
The active idle power consumption of the KC2000 is a bit on the high side even compared to other SM2262EN drives, but when the idle power states are engaged it is excellent at saving power, even in the desktop configuration that does not enable all of the PCIe power management features supported by the drive.
The downside to the excellent idle power savings possible with the KC2000 is that it takes a surprisingly long time to wake up, at about 136ms in both desktop and laptop configurations. This is more than twice as long as the already sluggish wake-up times exhibited by previous SM2262EN drives, and many times higher than what the drive's own firmware indicates.
Conclusion
Kingston is a well-known and high-volume SSD brand, but they have not had much of an impact on the high-end segment of the market. The Kingston KC2000 is much better try than several other enthusiast-oriented SSDs we've seen from them. On paper, the KC2000 is one of the more advanced SSDs currently on the market thanks to its inclusion of Toshiba's 96-layer 3D TLC NAND while the bulk of its competition is still using 64-layer NAND flash memory. The SSD controller used by the KC2000 is Silicon Motion's SM2262EN, which has a proven track record and still holds the record for some of our tests.
The combination of the Silicon Motion SM2262EN controller and Toshiba's 96-layer 3D TLC NAND flash memory works well to produce a decent high-end consumer SSD, but it doesn't break new ground. We've seen Toshiba get better power efficiency out of the same NAND using their own controller, and ADATA get better performance (and often better efficiency as well) from the SM2262EN controller by pairing it with Micron 64L TLC NAND. This novel controller+NAND combination doesn't have any compelling advantages, but it is competitive against the existing lineup and should suffice until Silicon Motion's next generation of controllers is ready. When Micron's 96L TLC starts showing up in competing drives, it may have a clear performance advantage over Toshiba's NAND.
The KC2000's biggest weaknesses show up on our ATSB tests of real-world IO patterns. Other drives based on the SM2262EN controller seem highly tuned for peak performance on lighter workloads, and suffer greatly on the harder tests. The Kingston KC2000 is missing that great peak performance, but it still suffers about as much on the harder tests. The KC2000 manages to generally stay ahead of entry-level NVMe drives on those harder tests, but it is outperformed by even Intel's QLC-based 660p in more realistic conditions where the drives aren't completely full and a faster SLC cache can be a big help.
NVMe SSD Price Comparison (July 22, 2019) |
|||||||
240-280GB | 480-512GB | 960GB-1TB | 2TB | ||||
Kingston KC2000 | $63.82 (26¢/GB) | $113.04 (23¢/GB) | $218.06 (22¢/GB) | $413.32 (21¢/GB) | |||
ADATA XPG SX8200 Pro |
$49.99 (20¢/GB) | $74.99 (15¢/GB) | $149.99 (15¢/GB) | ||||
HP EX950 | $86.99 (17¢/GB) | $139.99 (14¢/GB) | $274.99 (14¢/GB) | ||||
Silicon Power P34A80 | $38.99 (15¢/GB) | $63.99 (12¢/GB) | $127.99 (12¢/GB) | $299.99 (15¢/GB) | |||
MyDigitalSSD BPX Pro | $44.99 (19¢/GB) | $79.99 (17¢/GB) | $109.99 (11¢/GB) | $229.99 (12¢/GB) | |||
Intel 660p | $59.99 (12¢/GB) | $94.99 (9¢/GB) |
$194.95 (10¢/GB) | ||||
Intel Optane 900P/905P | $254.99 (91¢/GB) | $469.99 (98¢/GB) | $1141.99 (119¢/GB) | $2199.99 (147¢/GB) | |||
Samsung 970 EVO Plus |
$69.99 (28¢/GB) | $108.98 (22¢/GB) | $217.99 (22¢/GB) | $492.99 (25¢/GB) | |||
Samsung 970 PRO | $159.99 (31¢/GB) | $332.99 (33¢/GB) | |||||
Western Digital WD Black SN750 |
$69.99 (28¢/GB) | $99.99 (20¢/GB) | $189.99 (19¢/GB) | $499.99 (25¢/GB) |
Kingston seems to be pricing the KC2000 against flagship TLC drives from the top tier brands like Samsung and Western Digital, but Kingston isn't a top-tier brand. They are one of the largest second-tier brands, but they buy their components on the open market and didn't add any special sauce to the KC2000. They need to be competing against the likes of ADATA and can't really expect to charge much of a premium over even smaller brands.
There are simply too many alternatives to the KC2000 that are far cheaper. ADATA and HP have better drives using the same controller, and they're at least 25% cheaper per GB. There are Phison E12 drives approaching half the price of the KC2000, and the Kingston loses quite a few benchmarks against them.
96-layer 3D TLC NAND may be the future, but for now it's not doing Kingston any favors. When 64L NAND production starts to wind down and the NAND manufacturers try to get their margins back up, the KC2000 may end up looking somewhat competitive. But in the near future, the going rate for this grade of SSD will be staying much cheaper than what Kingston is asking for.