Kingston A400 Ssd 960gb Sata 3 25 Solid State Drive Sa400s37/960g Review

This week, I decided to option upwardly a Kingston A400 960GB SATA SSD at the "bargain" price of AU$120 posted to convert into a portable USB drive. At this price, it was cheaper than even some USB sticks which had less than a quarter of its capacity, making it a bargain. This particular model has been the bailiwick of a number of negative threads, suggesting it is a very unreliable model with poor performance and varying internals. I've previously endemic a number of SSDs which were considered past some to be "very unreliable" just not encountered any issues (e.g. OCZ Vertex 3, ADATA SU650) so I decided to take these reviews with a grain of salt.

The Bulldoze

The drive comes on a retail hanging parcel with nothing else but the drive itself, its distinctive branded design shown off through a clear plastic bubble. As a "value" serial SSD, the lack of any other inclusions is perhaps not unexpected. It claims to be 10x faster than a 7200RPM hard drive.

The rear clarifies that the performance claim is based on PCMark 8 Bandwidth Exam, and contains a number of production warranty related letters. A hole in the cardboard allows for seeing the bulldoze's series number label from the exterior. The unit is assembled in Taiwan.

The plastic chimera is near hands released past opening the packet from the side and sliding information technology out, revealing some boosted warranty information in different languages on the inside of the packet.

The drive has a sparse metallic shell, secured past four security Torx screws, i of which is behind a warranty seal. Because of the bulldoze's reputation, I did not opt to break the warranty seal and accept the unit apart, just apparently they do come with Phison or Silicon Motion controllers.

The rear shows the SATA connector being light amplification by stimulated emission of radiation etched with the chapters of the drive. The 960GB capacity suggests that the drive reserves a bit of its capacity as a cache to boost performance, every bit it is a DRAM-less bulldoze, as most "low-terminate" units are. In general, this is not skilful news, but when it comes to the price-sensitive segment of the market where the demands placed on the device are rather limited, this tin be an acceptable trade-off.

The other sides of the drive are elaborately "folded" together and the drive is a standard 7mm "thin" form factor. For use with 9.5mm enclosures, you lot will have to supply your own spacer.

Performance Tests

The drive was tested on my old desktop, an AMD Phenom II x6 1090T overclocked to three.8GHz with 16GB RAM and a Gigabyte 890FXA-UD7 motherboard running Windows 7.

CrystalDiskInfo

The drive started with a ability-on count of 3, two unsafe shutdowns and a few non-null values. Many of the SMART attributes are unknown and vendor-specific, with 0xE7 appearing to be the media vesture-out indicator while 0xAA seems to indicate pre-failure country and may be an indication of accumulated bad blocks at a guess.

After an all-encompassing amount of testing, attribute E7 was reduced by i, while 0xF1, 0xF2 appear to reflect the amount of reads and writes committed to the drive. Attributes 0xAD, 0xE9, 0xF4, 0xF5 and 0xF6 have besides changed as a upshot, some of which may be related to wheel count and error correction rates.

HDTune Pro

Sequential Reads

Initial tests showed the bulldoze to accept the expected SATA3-saturating read speeds averaging 516.5MB/s.

However, information technology was discovered that depending on the data pattern written to the bulldoze, the read speeds would vary but would be far less than when fresh, achieving averages of about 300-350MB/s instead.

Sequential Writes

On a make clean drive, writes averaged 460.4MB/southward which is a lilliputian less than what SATA3 is capable of, but is non a bad result.

Merely one time filled, rewrites are slower, recording 352.7MB/s and 448.4MB/s in 2 separate follow-up attempts. The drive temperature remained a absurd 31 degrees Celsius, and then this behaviour is virtually certainly not a result of overheating merely more than than the controller and flash behaves differently depending on the state of the bulldoze.

Random Fill

Considering of my discovery that the drives' behaviour is very data dependent and land dependent, I decided to try and fill the bulldoze with random data to see what its actual full-surface write speed is. Unfortunately, this process took over thirteen hours, averaging just 17.4MB/due south. This is a very un-SSD-like effect that could easily be bested by many hard drives and even USB two.0 flash drives. This makes the drive a very poor choice if you intend to make full it upwardly and this kind of behaviour is likely the cause of poor performance reported past users. Even reading actual data apace is a challenge for the drive, making this a rather big thwarting.

Other Tests

Testing IOPS functioning turns up some rather ordinary simply decent numbers.

Other tests were performed, no clear caching advantage is seen.

Full I/O performance is seen at nigh 128kB accesses.

CrystalDiskMark

Testing with CrystalDiskMark seems to testify read operation a piffling slower than what SATA3 is capable of, although the random 4k access performance does not seem too bad. I suppose this may be because these benchmarks use only a small portion of the drive (1GB) which makes it like shooting fish in a barrel for the bulldoze to deliver better numbers equally it might reside in a pSLC cache (assuming it has one).

Every bit SSD Criterion

Every bit SSD seems to show quite a similar event, with a score that is very average for a SATA SSD.

The copy benchmark resulted in some surprising numbers that exceed the adequacy of the SATA3 interface and may take been impacted by write caching on the system. I consider these results potentially anomalous.

The pinch benchmark shows a slight advantage on reads for compressible data – in that location may be some logic inside the controller that detects very thin patterns to speed up operation.

Anvil's Storage Utilities

The results from Anvil's Storage Utilities confirms the drive to be a very average SATA SSD, although over again, these tests only use a limited amount of the chapters, thus hiding the nasty behaviour of the SSD once it fills upward.

ATTO Disk Benchmark

Full performance is reached almost 64kB on the ATTO benchmark, with 32kB fairly shut.

H2testW

H2testW is another "acid" test that reveals the SSD'southward truthful graphic symbol – it slows downwards significantly once it's over 25% full, as a full fill takes over seven hours! This results in an actual average throughput of 35.2MB/s, which is barely enough to saturate a USB 2.0 link and could easily be bested by a decent microSDXC bill of fare. That is very disappointing performance, as even the read speed was a pedestrian 223MB/s on the test platform.

As I idea information technology was potentially an bibelot, I repeated the test using an Orico USB UASP three.0 enclosure on my new workstation (Ryzen 7 1700 @ iii.8Ghz, 64GB 2800MHz RAM) and received near identical results for writing and marginally faster reading. Information technology seems the read speed is definitely information dependent as uncovered with HDTune Pro.

Determination

Based on my testing of the Kingston A400, you should not try buy this at home. I should accept listened to the commenters, because in testing, while the bulldoze tin can plow out the expected numbers in a fresh state on the standard suite of benchmarks, this is an illusion based upon the express amount of storage used by the benchmarks and the controller'south behaviour based on compressible information patterns. Attempting to fill the drive completely reveals its true nature – functioning that drops rapidly one time the drive is over 25% full and averages a random-fill write speed of just 17.4MB/s on Hard disk drive Tune Pro and 35.2MB/s in H2testW. Read speeds likewise take a significant hit when real data is written to the bulldoze, falling to 300-350MB/s. While I could non assess reliability, the performance of this drive is virtually uncharacteristic of nearly SSDs.

While technology ordinarily gets cheaper and faster, in this instance, this doesn't seem to be the instance. This drive has shown that it is indeed possible to buy an SSD that is even slower than every other SATA SSD I have endemic to engagement. Spending a little more than for something else is strongly recommended.

dudleywhoun1964.blogspot.com

Source: https://goughlui.com/2020/08/02/quick-review-kingston-a400-960gb-sata-ssd-sa400s37-960gb/

Share this post