Original Link: https://www.anandtech.com/show/10258/the-samsung-750-evo-120gb-250gb-ssd-review-a-return-to-planar-nand
The Samsung 750 EVO (120GB & 250GB) SSD Review: A Return To Planar NAND
by Billy Tallis on April 22, 2016 8:00 AM EST
Since the introduction of their first consumer TLC SSD with the SSD 840, Samsung's consumer/retail SATA SSD lineup has consisted of two product families: the MLC-based Pro drives, and the TLC-based 840 and EVO drives. The strength of Samsung's SSD controllers and the advantages of 3D NAND have allowed the 850 EVO to maintain a solidly mid-range position in the SSD market well above any other drive using TLC NAND.
Meanwhile, with the strength of the 850 EVO as a near-midrange product - and with pricing to match - like so many other vendors over the last year, Samsung has also been developing a true low cost TLC drive for the mass market. The end result is the Samsung 750 EVO, which we're revieiwng today. The 750 EVO establishes a new budget-oriented product line that competes in the cutthroat low end of the market where price per GB is the most important aspect of the product.
There are several design choices that help minimize the cost of the 750 EVO, aside from the obvious choice of TLC over MLC. The MGX controller it borrows from the lower capacity 850 EVOs is a dual-core version of Samsung's usual triple-core architecture. Similarly, the 750 EVO is only available in 120GB and 250GB sizes, so there is no high-priced high capacity model. Consequently, with only two small capacities, the 750 EVO line is served by a single tiny PCB layout, made even smaller by the fact that Samsung has put the 256MB of DRAM in the same package as the SSD controller.
Samsung MGX controller with onboard 256MB DRAM
But the most significant aspect of the 750 EVO is that it doesn't use the 3D NAND that has been a key competitive advantage for Samsung's 850 product lines. Samsung has continued development of planar NAND even after transitioning their retail SSDs to 3D NAND, and the 750 EVO in turn uses 16nm planar TLC. It doesn't offer the same performance or endurance of Samsung's 3D NAND, but it does significantly lower the cost of the drive.
128GB of 16nm TLC NAND in a single package
The 16nm TLC NAND is the successor to Samsung's 19nm TLC that had a troubled tenure in the 840 EVO. With the 840 EVO, data at rest on the drive degraded over time and eventually required the use of more thorough and thus slower error correction when read back. As a result the 750 EVO inherits the benefits of all the firmware work Samsung did to mitigate the read speed degradation. This, for what it's worth, gives Samsung some degree of a leg-up against other 16/15nm TLC drives that face the same challenges.
| Samsung TLC SATA SSD Comparison | |||||
| Drive | 750 EVO 120GB | 750 EVO 250GB | 850 EVO 120GB | 850 EVO 250GB | |
| Controller | MGX | MGX | |||
| NAND | Samsung 128Gb 16nm TLC | Samsung 32-layer 128Gbit TLC V-NAND | |||
| DRAM | 256MB | 256MB | 256MB | 512MB | |
| Sequential Read | 540MB/s | 540MB/s | 540MB/s | 540MB/s | |
| Sequential Write | 520MB/s | 520MB/s | 520MB/s | 520MB/s | |
| 4KB Random Read | 94K IOPS | 97K IOPS | 94K IOPS | 97K IOPS | |
| 4KB Random Write | 88K IOPS | 88K IOPS | 88K IOPS | 88K IOPS | |
| 4KB Random Read QD1 | 10K IOPS | 10K IOPS | 10K IOPS | 10K IOPS | |
| 4KB Random Write QD1 | 35K IOPS | 35K IOPS | 40K IOPS | 40K IOPS | |
| DevSleep Power | 6mW | 2mW | |||
| Slumber Power | 50mW | 50mW | |||
| Active Power (Read/Write) | 2.1W / 2.4W (Average) | 2.4W / 2.8W (Average) | Max 3.7W / 4.4W | ||
| Encryption | AES-256, TCG Opal 2.0, IEEE-1667 (eDrive) | AES-256, TCG Opal 2.0, IEEE-1667 (eDrive) | |||
| Endurance | 35TB | 70TB | 75TB | ||
| Warranty | Three years | Five years | |||
By and large the performance specifications for the 750 EVO match the lower capacity 850 EVO, though a slightly lower random write speed at a queue depth of one hints that the 750 EVO may require a bit more background wear leveling work. The feature set is identical to the 850 EVO, making the 750 EVO one of the few low-end drives to support TCG Opal encryption.
Samsung's marketing strategy for the 750 EVO is a little different from their previous retail SSD products. The 750 EVO is targeted specifically at system integrators and system builders, rather than at users looking to upgrade an existing machine. Consequently, they aren't pushing to make the 750 EVO available from as wide a range of retail outlets. Enough of the major online retailers have it in stock that it is not difficult to obtain.
This review will focus on comparing the 750 EVO against Samsung's other SATA SSDs and against other current-generation low-end TLC SSDs of comparable capacity. Our Bench tool can assist in making other comparisons.
| AnandTech 2015 SSD Test System | |
| CPU | Intel Core i7-4770K running at 3.5GHz (Turbo & EIST enabled, C-states disabled) |
| Motherboard | ASUS Z97 Pro (BIOS 2701) |
| 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 |
- Thanks to Intel for the Core i7-4770K CPU
- Thanks to ASUS for the Z97 Deluxe motherboard
- Thanks to Corsair for the Vengeance 16GB DDR3-1866 DRAM kit, RM750 power supply, Carbide 200R case, and Hydro H60 CPU cooler
Performance Consistency
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.
Starting with a look at steady-state performance, the 750 EVO is clearly inferior to the 850 EVO, particularly at 250GB. But it outperforms most of the planar TLC competition and the occasional low-end MLC drive.
The high consistency score is a hallmark of Samsung's top-notch controller architecture. The 750 EVO is in an entirely different league from the planar TLC drives.
 |
|||||||||
| Default | |||||||||
| 25% Over-Provisioning | |||||||||
The 750 EVO's initial burst of high performance is relatively short-lived, but it transitions into a very well-regulated steady state. The gradual performance recovery before a second smaller drop in performance is less pronounced than on the other Samsung drives, but is still present.
With extra overprovisioning, the 750 EVO's steady state shows much looser performance regulation but is still delivering a better worst-case than its competition's best-case.
 |
|||||||||
| Default | |||||||||
| 25% Over-Provisioning | |||||||||
A closer look at the 750 EVO's steady state is pretty boring, with no clear patterns of periodic background maintenance or sporadic outliers.
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.
The Destroyer has earned its name here. The 750 EVO is clearly the slowest modern SSD on this test, showing that it is not suitable for sustained intense workloads with a high volume of writes. Almost any other SSD currently on the market will perform better under pressure, including competing TLC drives.
The 750 EVO also sets new records for slow responses, with average service times on par with standard hard drive seek times. Though since The Destroyer has an average queue depth of about 5.5, a mechanical hard drive would still be several times worse by this metric. Conversely, the best MLC SSDs are almost ten times quicker than the 750 EVO.
With over 10% of operations taking more than 10ms, we can't quite refer to them as outliers anymore. At the 100ms threshold, the 750 EVO has twice as many outliers as anything else.
The substantially higher energy usage of the 750 EVOs is a straightforward consequence of them taking much longer than everything else to complete the test: The 120GB 750 EVO took just over 17 hours to complete this test while the 120GB PNY CS1311 took only 13.5 hours and the 128GB 850 Pro needed only 10 hours.
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.
The Heavy test doesn't write enough data to fill a 120GB drive, and when there's still spare area the 750 EVO performs better than any other planar TLC drive. When the test is run on a full drive the 750 EVO suffers more than most and falls to the bottom of the rankings.
The penalty that the 750 EVO pays when filled is even more apparent when looking at average service times, pushing it into last place.
Even when starting on an empty drive, the 750 EVO's latency outlier situation isn't great, beating the ADATA SP550 but little else.
The energy usage of the 750 EVO only stands out for the full drive case. When starting with an empty drive, the 750 EVO uses less energy overall than the other planar TLC drives.
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.
The average data rate scores of the 750 EVOs allow it to blend in to the crowd as a solid performer and the fastest planar TLC drive on the Light test.
The average service time of the 750 EVO again places it as one of the fastest planar TLC drives and a reasonable performer compared to the SATA market overall. The 120GB 750 EVO is also relatively close to the 250GB version, while other 120GB models are considerably slower than their larger counterparts.
The number of latency outliers experienced by the 750 EVO in the Light test is normal for TLC drives but considerably higher than most MLC drives.
Power usage over the course of the Light test separates the drives pretty cleanly into different categories. The 750 EVO clearly requires more power than the 850 EVO, and all the other TLC drives require more power than the 750 EVO. The MLC drives all require less energy than all the TLC drives, except that the 850 Pro sacrifices efficiency to deliver its high performance.
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.
All the Samsung drives are crowding the top of the chart for low queue depth random read speeds, and the 750 EVO is way ahead of any other planar TLC drive here.
With power usage in the middle of the pack, the 750 EVO and 850 EVO are some of the most power-efficient drives on this test, but drives like the Plextor M6V are still at the top of the efficiency ranking.
 |
|||||||||
In addition to offering great performance at low queue depths, the 750 EVO scales up to reach higher speeds at QD32 than any non-Samsung drive, without power consumption getting out of hand.
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.
Taking capacity into account, the 750 EVO provides much better random write speeds than any other planar TLC drive, and the 250GB model is competitive against many MLC drives.
The 750 EVO is high-performing for a TLC drive but at the cost of requiring more power than most planar TLC drives and much more than the 850 EVO.
 |
|||||||||
The 120GB 750 EVO shows almost no scaling with queue depth, while the 250GB needs a queue depth of at least two to reach full performance. By contrast, the 120GB 850 EVO shows a little bit of performance scaling from QD1 to QD2 and the 250GB 850 EVO doesn't hit full performance until QD4.
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.
The 250GB 750 EVO comes in right behind the 850 EVO and Pro for sequential read speed, and the 120GB 750 EVO surprises by coming in fourth, significantly ahead of both the 120GB 850 EVO and 128GB 850 Pro.
Power consumption for the 750 EVO is a little high but nothing concerning. Only a few MLC drives distinguish themselves with particularly high efficiency during sequential reads.
 |
|||||||||
The high performance score of the 750 EVOs is due primarily to their unusually good QD1 speeds, which are quite close to the limit reached at higher queue depths.
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.
Sequential write speeds of the 750 EVO are significantly lower than the 850 EVO and not competitive with MLC drives, but are in the lead among planar TLC drives.
Power consumption for the 750 EVO is significantly higher than the 850 EVO, but it is still more efficient than the planar TLC competitors.
 |
|||||||||
Most drives show no scaling with queue depth in this test, and the 750 EVO follows suit.
Mixed Random Read/Write Performance
The mixed random I/O benchmark starts with a pure read test and gradually increases the proportion 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.
The 750 EVO outperforms any non-Samsung drive of similar capacity on the mixed random I/O test, and the 120GB 750 EVO even beats a few MLC drives with twice the capacity.
The 750 EVO has above-average power efficiency during the mixed random I/O test.
 |
|||||||||
Most drives lose performance during the middle phases of the mixed random I/O test. High performance scores on this test rely on performance bouncing back during the final phase of the test when the workload shifts to pure writes. The 750 EVO's spike at the end is noticeable though not huge, but it also benefits from comparatively good performance in the first two sub-tests.
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 proportion 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.
Unlike for the mixed random I/O test, most MLC drives are able to maintain a clear performance lead over the planar TLC drives—including the 750 EVO.
The 750 EVO's power consumption is low enough that it is clearly the most efficient planar TLC drive, but its efficiency doesn't beat a typical MLC drive.
 |
|||||||||
The 750 EVO has great read speed, but it bottoms out at a relatively low level during the middle of the test and doesn't bounce back at the end as much as most MLC drives and the 850 EVO.
ATTO
ATTO's Disk Benchmark is a quick and easy freeware tool to measure drive performance across various transfer sizes.
 |
|||||||||
The ATTO benchmark shows the 750 EVO has good performance on small transfers and a maximum write speed that is very close to the read speed. The 120GB 750 EVO even provides better write speeds than the 850 Pro 128GB thanks to the former's SLC write caching.
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.
The 750 EVO provides top-notch burst speeds for reads and writes. The write speeds in particular distinguish the 750 EVO from most other TLC drives and low-end MLC drives that suffer from a lack of parallelism at small capacities.
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.
Idle power consumption of the 750 EVO is comparable to other Samsung drives: great when ALPM is enabled, and average when it is disabled.
Final Words
The Samsung 750 EVO is a drive for a limited audience. It is intended for use as the primary boot drive of a system that will not be subjected to particularly intense storage workloads, and the limited capacity options preclude using it to hold a large game or multimedia library. Seen through this lens, the 750 EVO offers great performance for a budget drive. The peak performance of the 750 EVO is close to the Samsung 850 EVO and even the 850 Pro in many cases. On tests simulating lighter real-world usage the 750 EVO is generally the fastest budget TLC drive and also sometimes competes well against low-end MLC drives.
That said, if the 750 EVO is subjected to a more strenuous workload, things start to fall apart. The performance of this drive suffers greatly if it is operated in a near-full state and when sustained writes overflow its SLC cache. The same is also true of any other budget TLC drive, but most of the competition handles the pressure better than the 750 EVO. The best way to make use of the 750 EVO is probably to pair it with a large hard drive to hold bulk data and large applications. This is especially true of the 120GB model, as that much space can quickly fill up if used to store even a few movies or games.
As a cost-cutting exercise, the 750 EVO produces interesting results. Samsung's in-house SSD controller design was already the cheapest option for Samsung to use, and they didn't produce a crippled cut-down version for the 750 EVO. Instead, the 750 EVO gets the same higher-performance controller from the lower capacity 850 EVO, and broad feature set of the full 850 series. The NAND flash is where almost all of the cost savings occur and that does have an impact on performance, but under reasonable usage scenarios the Samsung controller is able to compensate for that better than most others. The warranty and endurance ratings on the 750 EVO are lower than for the 850 EVO but are normal for the budget segment of the market.
| SSD Price Comparison | ||
| Drive | 240GB/250GB | 120GB |
| ADATA SP550 | $57.99 | $34.99 |
| PNY CS1311 | $59.99 | $39.99 |
| PNY CS2211 | $79.99 | |
| OCZ Trion 150 | $59.99 | $45.99 |
| SanDisk Ultra II | $74.99 | $54.79 |
| Samsung 750 EVO | $79.95 | $59.99 |
| Samsung 850 EVO | $87.89 | $68.95 |
The current pricing for the Samsung 750 EVO accurately reflects where it ranks in terms of performance and features. For consumers who would otherwise consider getting a small 850 EVO, the 750 EVO saves some money while making only modest sacrifices in performance.
At the same time however due to this higher performance, Samsung is charging a higher price for it, and consequently compared to budget drives from other companies the Samsung 750 EVO doesn't look very attractive from a total price or price-per-gigabyte basis. There are MLC drives like the PNY CS2211 at the same price as the 250GB 750 EVO. There are 240GB TLC drives at or below the price of the 120GB 750 EVO, and the $25 gap between the 120GB ADATA SP550 and the 120GB Samsung 750 EVO is huge. In the end I suspect that most users who don't have a hard requirement for drive encryption would be better served by either a slightly lower performing drive with much better price per GB, or a higher-performing option than the 750 EVO.
Facebook
Twitter
RSS
