|Samsung 32GB USB 3.0 Flash Drive Duo (MUF-32CB/AM)
|SanDisk Ultra Flair USB 3.0 16GB Flash Drive High Performance up to 130MB/S (SDCZ73-016G-G46)
|SanDisk Cruzer Force 8GB USB 2.0 Flash Drive With Metal Casing- SDCZ71-008G-B35
Samsung 32Gb Bar (Metal) Usb 3.0 Flash Drive (Muf-32Ba/Am)
Most helpful customer reviews
563 of 578 people found the following review helpful.
All the other info from the reviews seems true to me about the speed and whatnot, (It's not the fastest), but one thing that I noticed is that the top reviews didn't test the waterproof claims on the product. Well after washing it through almost a complete cycle, cancelling the wash, and fishing it out of some sopping pants, I can say that it is waterproof. I dried it off with a towel and blew out the port, plugged it in, and it still works just as it did before. Data was still there, new data copied on, and everything was fine, and for that I give it five stars.
377 of 402 people found the following review helpful.
The flash drive is super small at the same time very fast!
By Sam Ad
The flash drive is super small but also at the same time very fast compared to other metal USB 3.0 flash drives. The best part of this; the key chain holder is metal and strong and you can take it with you and your keys anywhere.
I bought mine in 2015 and I no longer have a use for it anymore but I tested the speed again in 09/2017 and still getting decent speeds. Some people have been disappointed for not getting the same kind of speed but one thing to be aware is, the speed tests depend on the type of motherboard and USB controllers. Other factors include the number of other USB devices you have connected to your machine. After all, the USB devices have USB bandwidth they share. Another factor is whether you plug it to a USB 2.0 port which ultimately limits your speed to a USB 2.0 speeds.
22 of 22 people found the following review helpful.
Review For: Samsung BAR 32GB and 128GB (Slow Writes?)
There seems to be some confusion about the performance of these drives, stemming from the benchmarks people are posting in reviews.
For the 32GB version of the drive, the real-world performance, using the default file system (FAT32), the default Windows policy (quick removal) and transferring LARGE files, is around 20MB/s write and 150MB/s read. For the 128GB model, the write speed increases to ~30MB/s, with read speed remaining unchanged. Transferring lots of small files at once will be slower than transferring large ones.
You're not going to get symmetrical 100+MB/s read/write speeds at this price point.
When it comes to writing data, these drives will have a small cache that's far quicker than the rest of the available storage. Any advertised write speeds may be referring to that cache. If you write a small enough amount of data to the drive, it will be written to that cache at a high speed, then in the background that data will be trickled out of the cache and into the slower part of the drive. However, when writing larger quantities of data, the cache will be saturated, at which point the write speed will drop to that of the regular TLC memory.
It's exactly the same for their low-mid range SSDs; the 840 EVO is advertised as having ~500MB/s sequential write speeds, and CrystalDiskMark will appear to back this up, but that's because it's not testing any deeper than the SLC "TurboWrite" cache. If you run a SUSTAINED sequential write operation on the 840 EVO, it'll be forced to write directly to the regular TLC NAND. You will see the speed plummet to <150MB/s after about 3GB of data has been written, and it will remain at that speed until you stop writing. To get sustained sequential write performance at 400+MB/s, you would need had to have bought the 840 Pro, which uses faster MLC NAND for the entire storage area.
make sure you look carefully at the benchmarks people are posting. One of the reviews here is showing the 32GB model/packaging, but the CrystalDiskMark result is for a 128GB device. Also pay close attention to the benchmark software version and settings. For example, new versions of CrystalDiskMark are running a different set of tests to old versions, so the results aren't necessarily going to look the same even if the device is performing identically. The default "test size" setting for CrystalDiskMark is 1GiB, but people are posting results taken with that set to 100MiB. This will make all the difference between getting a ~20MB/s write result and a ~100MB/s write result, presumably because 100MiB isn't big enough to saturate the abovementioned cache. I've provided CrystalDiskMark results from two different versions of the program and with different settings to illustrate all this.
Anyway, I have four of these drives. I haven't had them long enough to comment on longevity, but so far so good.
They get quite warm, but I haven't seen any overheating/thermal throttling. I pulled the entire 128GB of data off one drive, and it was reading at the same ~150MB/s the entire time.
The shape of the drive is slightly inconvenient because it can cover a neighboring port if the ports are stacked closely together (eg. motherboard IO cluster).