Table of Contents
Kingston is a well known company for their multitude of storage solutions, and entered the market way back in 1987 with a surface mount memory chip. So, Kingston is a veteran when it comes to the storage market. They have recently released their new HyperX 3K line of solid state drives, which aim to give users a high-performance storage solution while keeping the cost low. Kingston’s 240 GB HyperX 3K solid state drive has been put to the test over the past few weeks, so let’s see how it performed.
Specifications
A couple things to note in the specifications are the 3K program/erase cycles, which is where the drive gets its “3K” branding, and the MTBF (Mean Time Between Failures) rating of 1,000,000 hours. Both of those are below the normal 5K P/E cylces and MTBF of 2,000,000 hours seen in other SF-2281 drives, and these are reasons Kingston was able to cut costs on the HyperX 3K series.
Kingston HyperX 3K SSD Specifications | |
Form Factor | 2.5″ |
Controller | LSI SandForce SF-2281 |
Components | 25 nm Intel MLC Synchronous NAND (3K P/E Cycles) |
Interface | SATA Rev 3.0 (6 Gb/s) |
Capacities | 90 GB, 120 GB, 240 GB, 480 GB |
Sequential Reads ATTO Disk Benchmark 2.41 | 90/120/240 GB: 555 MB/s 480 GB: 540 MB/s |
Sequential Writes ATTO Disk Benchmark 2.41 | 90/120/240 GB: 510 MB/s 480 GB: 450 MB/s |
Sustained Random 4K R/W IOMeter08 | 90 GB: 20,000/50,000 IOPS 120 GB: 20,000/60,000 IOPS 240 GB: 40,000/57,000 IOPS 480 GB: 60,000/45,000 IOPS |
Maximum Random 4K R/W IOMeter08 | 90 GB: 85,000/74,000 IOPS 120 GB: 85,000/73,000 IOPS 240 GB: 86,000/60,000 IOPS 480 GB: 75,000/48,000 IOPS |
Power Consumption | 0.455 W (Idle), 1.58 W (Read), 2.11 W (Write) |
Dimensions | 69.85 x 100 x 9.5 mm |
Weight | 97 g |
Operating Temperature | 0 °C ~ 70 °C |
Storage Temperatures | -40 °C ~ 85 °C |
Shock Resistance | 1500G |
Vibration Operating | 2.17G |
Vibration Non-Operating | 20G |
MTBF | 1,000,000 hours |
Total Bytes Written (TBW) | 90 GB: 57.6 TB 120 GB: 76.8 TB 240 GB: 153.6 TB 480 GB: 307.2 TB |
Features | Supports SMART, TRIM, and Garbage Collection |
Packaging & Accessories
The box isn’t very flashy. The front has a picture of the HyperX 3K SSD, its name, drive size, read/write transfer rates according to ATTO 2.41, and the Kingston face logo creepin’ in at the bottom right. The back side of the box lists the contents and shows an image of them. It also has “Boots faster and loads applications faster. Increased durability and reliability.” printed in 22 different languages, and I’m not convinced that’s the most useful info to be printed in so many languages on the box. On the sides, it’s just product info and marketing.
Opening the box, we can see the SSD and accessories are packed into two pieces of foam that isolate everything from the edges of the box. So just in case something were to happen to the box, the contents would be undamaged. The CD including the installation manual and Acronis True Image software is placed between the two foam blocks.
The HyperX 3K SSD and its 2.5″ to 3.5″ mounting adapter are in the top piece of foam. There are also enough screws provided to mount the SSD to the adapter and to mount the adapter in a 3.5″ bay. Kingston even added detail to the foam by cutting out their “X” logo into the foam, between the SSD and mounting adapter.
Next, we have a couple accessories that usually aren’t included with new SSDs, a screwdriver and a 2.5″ USB 2.0 external enclosure. Not sure why Kingston chose to go the USB 2.0 route on the enclosure, especially with a SSD and USB 3.0 being backwards compatible. USB 3.0 would have been an amazing addition, it must have just been a cost thing. Nonetheless, an included external enclosure is awesome in itself!
Kingston HyperX 3K SSD
The HyperX 3K is a very nice looking drive. Its top casing is brushed aluminum with black trim, and the brushed aluminum makes an “X” shape for HyperX.
Nothing special about the bottom of the SSD, just the label and threaded holes for mounting.
Here’s the sticker I really wanted to break so I could show the internals of the SSD. However, the drive’s casing is attached with an odd type of screw. I thought they were just typical Torx screws, but they are actually like Torx plus an extrusion in the middle. Unfortunately, I don’t have the tools to remove this type of screw.
Acronis True Image Software
The included Acronis software is a bootable disk that will allow you to clone your current HDD/SSD over to the HyperX 3K SDD for an easy switch without having to reinstall the OS. When booting from the disk, the first screen you’ll see is the following. Just click on Acronis True Image to continue to clone your drive to the HyperX 3K.
On the next screen, you’ll click on Clone disk, which will take you to a screen to choose between Automatic or Manual cloning. Automatic mode will work fine, and then you’ll just choose the source disk (current OS disk) and the destination disk (HyperX 3K). Keep in mind that the destination disk needs to be large enough to hold the used space on your OS drive. Now, let the cloning do it’s thing and you’ll have a perfectly bootable HyperX 3K with everything as you had it before.
Test Setup & Methodology
Test Setup | |
CPU | Intel i7 2700K |
Motherboard | EVGA P67 FTW |
RAM | Corsair Dominator GT DDR3-1600 6-6-6-20 |
GPU | EVGA GTX 680 |
SSD | Kingston HyperX 3K 240 GB |
Operating System | Windows 7 Professional 64-bit |
Drivers | Windows 7 Default Driver Intel RST 10.8.0.1003 Driver |
Tests
- Boot Timer – Records time to boot
- ATTO v2.46 – Queue Depth set to 10 (max)
- CrystalDiskMark 3 – Default (Random), 0Fill, and 1Fill
- AS SSD Benchmark – Default Settings
- IOMeter 2008.6.18 – 4 KB Random (4 KB Aligned) and 2 MB Sequential (Sector Aligned); Queue Depth set to 32
- Windows Driver -vs- Intel RST Driver – Used IOMeter 2008.6.18 to test differences in drivers
- USB 2.0 -vs- USB 3.0 Enclosure – Used ATTO v2.46 to test speeds with the enclosures
Results
Boot Timer
Boot Timer measures the amount of time it takes to get into the OS from the end of POST. The key is it starts counting at the end of POST, and POST can vary depending on setup. Measuring that way removes that variance to make the boot timer results comparable across multiple hardware setups. Also, Boot Timer doesn’t stop counting until after everything in the OS is loaded and start-up programs/services are started. I didn’t have much of anything installed on the HyperX 3K except for Windows 7 Pro 64-bit, and I had a sub 10 second boot time of a VERY low 9.656 seconds.
ATTO 2.46
Looking at ATTO results, the HyperX 3K and Vertex 3 Max IOPS are neck-and-neck in reads and writes of 16 KB or greater. The HyperX 3K starts falling behind in the smaller transfers of less than 16 KB. The specifications state 555/510 MB/s for Read/Write in v2.41 of ATTO, and in my test with v2.46 I got 546/521 MB/s Read/Write, which I got by dividing the actual ATTO results in kilobytes by 1024 to get megabytes. So, the HyperX 3K didn’t quite meet the rating in read tests, but exceeding the rating in write tests.
IOMeter 2008
On to the IOMeter tests, and first we have the transfer rates. The small 4K reads and writes go to the Max IOPS drive, while the larger 2MB reads and writes unexpectedly go to the vanilla Vertex 3. However, the HyperX 3K isn’t too far behind OCZ’s drives.
On to the IOPS. According to the drive specifications, Kingston uses IOMeter08 to test their Random 4K IOPS and I’ll be using that program as well. For 4K Random Read/Write, I got 81,172/52,931 IOPS. So, in these results, the 4K Random Read IOPS aren’t quite up to the max rating of 86,000, but it’s well over the sustained rating of 40,000. Surprisingly, the 4K Random Write IOPS results don’t reach the max rating of 60,000 or the sustained rating of 57,000.
AS SSD Benchmark
The AS SSD transfer results are in line with the other others with the HyperX 3K keeping up with the Vertex 3 Max IOPS in most tests. The Max IOPS clearly comes out in front through half of the AS SSD tests, the HyperX 3K wins in two of the tests, and all the SSDs basically tie in the 4K read test.
With AS SSD we can also get the access times for the drives, and it turns out the HyperX 3K wins in Write access time by beating out the Vertex 3 Max IOPS, and comes in a close second to the Vertex 3 in Read access time. Realistically, we are talking fractions of a millisecond here, so no difference between the drives would be noticeable.
CrystalDiskMark 3
Using CrrystalDiskMark’s deafault (random) test, we continue to see the fight between the Vertex 3 Max IOPS and the HyperX 3K. The Max IOPS beats the HyperX 3K by a noticeable amount in the 512KB read and 4KB queue depth 32 read tests. In the other tests, the HyperX 3K either wins or comes very close to the Max IOPS.
In CrystalDiskMark’s 0 Fill test, the HyperX 3K is either equal to or beats the Vertx 3 Max IOPS in every test except for the 4K read at 32 queue depth again. I expected the Max IOPS to continue it’s dominance in the small file reads and writes, but that’s not the case with CrystalDiskMark.
Very interesting results in the 1 Fill test. In the Seq Write and 512K Write tests, the HyperX 3K SSD completely dominates all the other SSDs tested. In the other parts of the tests it’s a fight for the top top spot.
Windows Default Driver -vs- Intel Rapid Storage Technology Driver
When comparing the two drivers, the default driver used by Windows 7 is much better than the Intel RST Driver. The only test where the Intel RST driver comes out on top is the 4KB random write, beating the Windows driver in both transfers and IOPS.
Enclosure
Typical USB 2.0 speeds here (~35/27 MB/s), and it severely bottlenecks the HyperX 3K SSD as expected. The USB 3.0 enclosure still bottlenecks the HyperX 3K a lot, but it’s 5-6 times faster than USB 2.0, coming in at ~170 MB/s in both reads and writes. I’m unsure of the additional cost to make the enclosure USB 3.0, but it would have gotten us a lot more speed out of the HyperX 3K SSD in a portable package. Like I said earlier, having an enclosure of any type include with the drive is plus.
Conclusion
The HyperX 3K competed well in the tests used for this review. Transfers were high, up to 546 MB/s in reads and 521 MB/s writes in ATTO Disk Benchmark. The maximum IOPS I experienced with the HyperX 3K in IOMeter 2008.6.18 were 81,172 in reads and ~60,000 in writes, not quite up to the max ratings, but a lot better than the sustained ratings. The boot time for the HyperX 3K drive was an insanely fast 9.656 seconds, so there will be a lot less waiting to get into the OS.
So, comparing prices of stand-alone 240 GB drives, the Kingston HyperX 3K is $280, Vertex 3 is also $280, the Cucial M4 (256 GB) is $250, and the Max IOPS is $430. Performance-wise, the HyperX 3K comes in second because the Max IOPS wins outright in sub-16K transfer tests and also has better IOPS numbers. However, the HyperX 3K is $150 cheaper than the Max IOPS for very similar performance. The 240 GB HyperX 3K tested is an upgrade kit costing $290, and most other available SSDs don’t have upgrade bundles, which is a plus for Kingston. The included software and accessories are well worth the $10 difference between the stand-alone and upgrade versions.
The cut in price was made possible by using 3K Program/Erase cycle 25 nm NAND instead of 5K and by using NAND with a MTBF of 1,000,000 hours instead of 2,000,000. Both of the changes are irrelevant to the majority of SSD users. P/E cycles only come into play in heavy writing situations, but if your drives will experience heavy writing, then you shouldn’t get a consumer grade SSD. The mean time between failures (MTBF) of 1,000,000 hours turns out to be over 110 years, so if you experience this then you were just extremely unlucky.
I do believe Kingston has met their goal of providing high-performance SSDs at a low price.
Click the Approved stamp for an explanation of what it means.
– Matt T. Green (MattNo5ss)
3 replies
Loading new replies...
Registered
Premium Member
Registered
Join the full discussion at the Overclockers Forums →