Sandisk Ultra 128gb Usb 30 Flash Drive Review

Recently, I had some bad luck with my Sandisk Ultra Flair 128GB USB three.0 Wink Drive which failed after a few years of read-by and large operation. While initial recovery seemingly went well, the RMA process was protracted as the replacement particular was somehow delivered back to the sender. Afterward calling and escalating the consequence, a lack of stock led to a asking from Sandisk to substitute their Ultra in identify of the Ultra Flair to which I accepted. At last, on 13th October, the replacement drive arrived – the whole process costing 52 days.

Given that this was a different unit of measurement and information technology is customary for all data storage devices to be subjected to an initial commissioning test, I decided to write this up as a quick review.

The Unit of measurement

The packaging is very much like other SanDisk products – reddish stripe and logo on the bottom, but the text seems to have been pared dorsum a piffling, in that location is no longer whatever artwork in the background, and the new "store MORE. FAST>" tag-line has been added. Peradventure the latter is to misfile prospective buyers, every bit the production seems to only claim a speed upward to 130MB/south for reads merely nix nigh writes!

This package is Made in Malaysia and has a model code of SDCZ48-128G-A46. Information technology is covered by a 5-year limited warranty and includes a RescuePRO Deluxe software download offering with a key inside the cardboard packaging.

The drive is sealed inside a plastic "chimera" with a plastic tear-off film.

Unlike the Ultra Flair, this drive has a retractable, proper USB 3.0 connector. The shape is iconic of the Ultra series, although the sliding detented insert is black rather than the orange of older USB ii.0 models and in that location seems to be no activity lite.

The rear has a section where it seems perfect for labelling the owner or use of the drive. No serial number etchings were observed, which is a little unexpected. A pigsty on the rear of the drive allows for zipper to a lanyard.

The side contour of the bulldoze shows its slightly "curved" design.

The drive appears with a VID of 0781 and a PID of 5581. Its default formatted capacity was 123,030,798,336 bytes total in FAT32 with eleven,862,016 bytes used.

Included is the Sandisk Secure Access software. Every bit I am using a Windows 10 system, the "Arrangement Volume Information" binder appears to accept been automatically written by the Bone.

The partitioning type is MBR with a partition offset of 32 sectors and no spare sectors at the stop. The total drive capacity is 123,060,879,360 bytes, which is 62,586,880 bytes (59.6875MiB) larger than the Ultra Flair it replaces.

Performance Testing

Functioning testing was done using the onboard AMD Chipset-based USB 3.0 ports on my workstation, an AMD Ryzen 7 1700 (overclocked to 3.75GHz), 64GB DDR4 RAM (overclocked to 2800MHz) on an Asus PRIME X370-PRO running the Windows x (21H1).

HDTune Pro

Fresh out of the box, sequential read was a scrap depression, averaging 54.3MB/due south. After a full write, however, the speeds increased to the expected levels, reaching an average of 152.6MB/s.

Write speeds, from fresh, were rather erratic, averaging just 24.3MB/s on the first write laissez passer, although reaching 43.1MB/southward in brusque spikes.

Random access tests showed 512 byte read IOPS at 1059, 4kB read at 1225. Write IOPS, however, were noticeably lower at 82 and 118 respectively, with pregnant variability in some very long latencies of >6s beingness measured.

Extra tests were run, although these figures are perhaps more relevant to hard drives.

CrystalDiskMark

CrystalDiskMark seemed to reaffirm the drive's read performance beingness decent, reaching 152MB/southward sequential. The write performance lags far backside at 31.88MB/s which is something perhaps the best USB two.0 drives can achieve. Small block 4kB accesses reached just shy of 7MB/s read and effectually 0.58MB/s write which seems pretty typical just the original Ultra Flair did perform a trivial better overall reaching 164.8MB/s, 37.72MB/s, 8.3MB/s and 0.65MB/s respectively.

ATTO Disk Criterion

The first ATTO run showed some inconsistency around 24MB accesses, cogitating of the inconsistent write IOPS performance noted before. A re-run showed more consistent results. The drive shows rather typical behaviour in ATTO, reaching total speed by 256kB accesses. Unlike the Ultra Flair, the performance seems to be more stable and consequent, which may be because the Ultra Flair may suffer from thermal degradation of performance.

H2Testw

A full surface test with H2testw showed speeds like with the other tests, with no integrity errors reported. Compared to the Ultra Flair, information technology did fare better in both write and verify metrics, maybe due to thermal constraints with the Ultra Flair which gets quite hot under workload.

Conclusion

The Sandisk Ultra USB 3.0 as a replacement for the Ultra Flair performed slightly worse than the Ultra Flair in most tests, although exhibited more functioning stability. Its larger pattern with slightly larger usable capacity and a proper retractable USB 3.0 connector is preferable as it does non exhibit the host port to pregnant heating and seems to reduce potential thermal-throttling induced performance variation. Write operation was significantly behind the advertised read speeds, every bit it common amongst "low-end" cost-sensitive flash memory products, reaching somewhere in the 24-40MB/south range depending on the benchmark and drive status. Write IOPS seem to exist quite inconsistent with some long pauses in throughput (>6s) observed, which suggests this bulldoze may non be skilful for real-fourth dimension recording needs.

Bonus: Imperfect Data Recovery?

Now that I had a replacement drive for the SanDisk Ultra Flair, I could write the image back to the drive and get back to concern … or so I idea.

Instead, I was greeted to near half of the files (by occupied space) non matching hashes from backup files.

Instead, they suffered random fleck-rot abuse despite ddrescue successfully recovering all sectors into the epitome. This highly suggests that the Ultra Flair was reporting successful reads despite the reads existence unsuccessful and means that software-based recovery tools may not be able to know if the drive is existence honest nearly its data recovery (even if it attempts the same recovery twice, or more).

Examining the binary differences between files, it seems that every false read was returned as a block of 2048-bytes of nothing (0x00) bytes. Inadvertently, this has ended up creating a binder of "glitch-art", but thankfully, I do have a backup of most of the files I can restore from.

That being said, I could accept discovered this earlier, earlier I RMA'd the drive, if I had mounted the drive paradigm on a loop-back device and examined it … but I didn't, every bit the recovery and postal service-recovery testing had killed the original drive to the point of not being detected and as a rarely-used backup, I could exercise without information technology for a while.

In the finish, this only reduces my trust in SanDisk products in general. I've had way too much bad luck with their Ultra microSD cards, and now, their Ultra USB drives have besides caused problem. I've also found two of three Strontium Nitro 433x 16GB microSDHC cards have besides destroyed themselves every bit well after about two to three years (with no data loss), although despite the lifetime warranty, the toll of returning the items brand it pointless to endeavor and claim on it. Perhaps the pursuit of small size and depression cost should terminate if nosotros expect consumer-grade microSD cards to accept any level of reliability in any but the nearly occasional use applications.

brownhicharach.blogspot.com

Source: https://goughlui.com/2021/10/24/review-sandisk-ultra-usb-3-0-flash-drive-128gb-sdcz48-128g-a46/

Share this post