Expert’s Rating
Pros
- Small form factor
- Inexpensive
- Good everyday performance
Cons
- Slows drastically during long file writes
Our Verdict
For the average user the low price of this good everyday performing 10Gbps SSD is the prime selling point. But it’s slow long write render it unsuitable for video editing and the like.
Price When Reviewed
1TB: $65 I 2TB: $110
Best Prices Today: Kingston XS1000 USB SSD
$65.83
Kingston’s new 10Gbps XS1000 portable SSD brings the 20Gbps SX2000‘s small form factor to the masses at a more affordable price point, and with surprisingly comparable performance.
Indeed, in our real-world 48GB transfers, it actually proved quite a bit faster. Go figure. Alas, neither drive handles super-large data sets very well, as you’ll see below in the performance section.
Further reading: See our roundup of the best external drives to learn about competing products.
XS1000 price and features
Kingston sells the XS1000 in 1TB/$65 and 2TB/$110 (tested) capacities, which is considerably cheaper than the XS2000, though the latter has dropped in price significantly since our review. And the XS1000 lacks the 512GB and 4TB capacities that the XS2000 is available in.
The XS1000 is approximately 2.73 inches long (69mm), 1.27-inches wide (32mm), and 0.53-inches (13mm) thick and features a Type-C connector.
When it came to the 48GB file transfers, the XS1000 was actually faster than the XS2000 by a fair margin, and faster than the 10Gbps competition.
The drive is USB 10Gbps, which I’ll refer to from now on as simply 10Gbps. Kingston includes a Type-C to Type-A cable, but there’s no silicon sock as is included with the XS2000.
That said, the XS1000’s semi-gloss black finish is less prone to blemishes than the XS2000’s silver. And SSDs are so shock resistant in general, that the lack of the extra buffer isn’t a big deal — though I did like the tactile sensation of the rubber boot.
The controller is a Silicon Motion SM2320, but Kingston declined to reveal the exact type of NAND inside. The company did admit that it’s 3D TLC (Triple-Level Cell, 3-bit). We’d guess lots of layers (176, 232, etc.) seeing as it’s available up to 2TB.
The XS1000’s internals are NVMe, as SATA would not be able to deliver anywhere near the performance that the XS1000 reaches.
How doe the XS1000 performan?
We included the 20Gbps Kingston XS2000 in the charts so you can see exactly what, if any, advantages the pricier drive delivers. The other two drives, the SK Hynix Beetle X31 and Crucial X9 Pro are both 10Gbps SSDs like the XS1000.
Obviously, the 20Gbps XS2000 was going to dominate the synthetic benchmarks, but the XS1000 is right there with the 10Gbps competition In CrystalDiskMark 8 and AS SSD (not shown).
Random performance was largely the same, though the 20Gbps throughput of XS2000 offers no advantage over the XS1000 in this department.
When it came to the 48GB file transfers, the XS1000 was actually faster than the XS2000 by a fair margin, and faster than the 10Gbps competition.
Where things fall apart for the XS1000 is the 450GB write time. It’s nothing short of abysmal, even compared to the XS2000’s weak performance. Kingston’s XS-series SSDs are not what you want for this type of operation, though it’s an admittedly rare type of workload.
The capture below shows that the XS1000 runs out of steam — aka secondary cache (NAND written as SLC rather than TLC/QLC) — at around the 150GB mark and speed drops to between 100MBps and 200MBps from that point on.
As long as you stay away from very large data sets, the XS1000 is a good performer. Indeed, while the 450GB write was a bit tragic, that’s really not an action the average user will perform more than once in a blue moon.
Should you buy the Kingston XS1000?
We certainly can’t argue with the price of the XS1000, or its performance in the vast majority of situations. For everyday computing and data transport, it’s aces and a much better value than the XS2000.
But it’s not for prosumers banging on it constantly with large video files. Check out Crucial’s X10 Pro for help with that scenario.
