- Obtained from: Amazon
- Advertised capacity: 64GB
- Logical capacity: 62,277,025,792 bytes
- Physical capacity: 62,277,025,792 bytes
- Fake/skimpy flash: Skimpy (2.69% skimp)
- Protected area: 62,411,243,520 bytes (inaccessible)
- Speed class markings: Class 10*, U3, V30, A2
- CID data:
- Manufacturer ID:
0x27
** - OEM ID:
0x5048
** - Product name:
0x5344363447
(ASCII:SD64G
) - Product revision:
0x60
- Manufacturer ID:
* The Class 10 marking appears on the product package, but does not appear on the card itself.
** This manufacturer ID/OEM ID combination is pretty well known to belong to Phison.
*** This item was purchased as a 3-pack. The price shown represents the cost of a single item in the 3-pack.
Discussion
This one came to me as a roundabout suggestion from Reddit: a Redditor had asked me if I had tested PNY’s 1.5TB card. I hadn’t, because (genuine) cards that large are typically out of my price range. However, I dug into it a little bit more and found that the only 1.5TB card that PNY shows on their website is the PRO Prime Elite; I then saw that a 3-pack of the smallest capacity — 64GB — would fit in my price range. (And on further investigation, it appears that PNY only sells the 64GB version in a 3-pack.) So, here you go random Redditor!
If you only pay attention to the CID data, it would appear that these are simply relabeled PNY Elite-X’s or HP MicroSDXC mx330’s; however, performance tells a different story: these cards managed to weasel their way into my top 10 with their impressive scores across all metrics. Sequential write scores, in particular, were the highest out of any card I’ve tested so far — even beating out my previous favorite, the Kingston Canvas Go! Plus. Its worst metric was random read speeds — but even these were nothing to balk at, coming in above average. As of this writing, sequential read speeds came in about 1.5 standard deviations above average; sequential write speeds came in a whole 3 standard deviations above average; random read speeds came in about 0.5 standard deviations above average; and random write speeds came in about 2.5 standard deviations above average. Overall, pretty impressive, and good enough to put it in the #2 spot on my top 10 list. (Well done, PNY!)
On the endurance testing front:
- Sample #1 has not yet reached the 2,000 read/write cycle mark. It is currently expected to get there sometime in May 2025.
- Sample #2’s first error was a write failure, affecting 128 contiguous sectors, during round 535. It has survived 1,352 read/write cycles in total so far.
- Sample #3 has not yet reached the 2,000 read/write cycle mark. It is currently expected to get there sometime in May 2025.
April 3, 2025