Two V60 cards that both hold V90

Why test two cards that look identical on paper

These two arrived as a deliberate pair. Same form factor, same UHS-II interface, same 128 GB capacity, same printed V60, and the same 250 MB/s read claim on the face. If two cards agree on every number a shopper can read, the only thing left to separate them is a bench. They ran on the same reader, on the same afternoon, under the same protocol. They join CKBench, the running index of every card on the bench, alongside the SanDisk High Endurance 128GB and the Kioxia Exceria G2 128GB.

What the cards claim, and what they did

Read figures come from the ProGrade PGM0.5, a dual-slot UHS-II reader. Every read number on this site names the reader that produced it, because the reader changes the answer.

ProGrade SDXC UHS-II V60 128GB

Lexar Professional 1667x 128GB

Both cards are badged one class below what they deliver

The video speed class on a card is a floor, not an average. A V60 card promises that its worst ten seconds under sustained writing will never drop below 60 MB/s. That is the number that matters when a camera is recording, because a single dip below the floor is a dropped frame or a stopped clip.

The ProGrade's worst ten seconds landed at 102.8 MB/s. The Lexar's landed at 109.5. Both are comfortably past the 90 MB/s V90 floor, so both cards would qualify for a badge one full step above the one they carry. The ProGrade clears V90 by 14 percent and the Lexar by 22 percent.

That is not a mistake by either manufacturer. Certifying to a higher class means guaranteeing it across every unit, every temperature, and the whole warranty period, and a card that scrapes V90 on a good day is safer sold as a V60. What it does mean for a buyer is that the badge is a promise rather than a description, and the real number can sit well above it.

The card that claims less is the one that delivers it

Neither card prints a write speed on its face. Both numbers come from the manufacturer's own product page, and that is where the two part company.

Lexar claims 120 MB/s. Across three passes the card averaged 119.3, with the best pass at 120.9. That is a card doing exactly what it said it would do, and the small shortfall on the average is inside the noise of the measurement.

ProGrade claims 130 MB/s. The card averaged 108.4 across the same three passes, and held 104.0 once the write was sustained. That is 17 percent under the claim, and it is not a bus limitation: 108 MB/s is nowhere near the ceiling the same card's 296.9 MB/s read is pressed against. The card simply does not write as fast as the spec sheet says.

It is worth being precise about what that does and does not mean. The ProGrade is still a fast card that holds a class above its badge. But of the two, the one that made the more modest claim is the one that kept it.

What happens if you put a UHS-II card in a UHS-I slot

UHS-II cards carry a second row of contacts behind the first. That second row is the entire reason the card costs what it does, and a UHS-I host cannot use it. The card still works, because UHS-II is backward compatible, but it falls back to the older bus.

I measured that directly. The same ProGrade card, minutes apart, on a UHS-I only reader: 95.0 MB/s read and 88.5 MB/s write, against 296.9 and 108.4 on the UHS-II reader. Reading, that is a factor of 3.1 from the host alone, with the card unchanged.

The practical version: if your camera body is UHS-I, a UHS-II card buys you nothing at all. It will perform like a good UHS-I card, because for the duration it is one. Check the slot before you buy the card. I went into what the reader alone is worth in the ProGrade PGM0.5 reader review.

Neither card has a cache cliff, and neither one stalled

Plenty of cards start fast and collapse once a small fast buffer fills. That collapse is the single most common way a card that benchmarks well fails in a camera, and it only shows up if you keep writing.

I wrote to both cards until they were full. The ProGrade began its sustained run at 104.5 MB/s and finished the entire 128 GB at 104.2. The Lexar began at 112.2 and ended at 111.1. Neither curve has a knee in it anywhere. Speed measured at the first, middle and last tenth of the card varied by 0.28 percent on the ProGrade and 0.83 percent on the Lexar, which is to say not at all.

Neither card produced a single write block longer than 500 milliseconds. Across nearly 30,000 blocks on the ProGrade the longest pause was 52 milliseconds, and across nearly 32,000 on the Lexar it was 56. For recording, that steadiness matters more than any peak figure.

How hot do they get, and why I had to reshoot it

Both cards run hot under a sustained write. Pulled from the reader the moment the write stopped, with the room at 76 F, the ProGrade measured 145.1 F and the Lexar 142.7 F on the card body.

Getting those numbers honestly took two attempts, and the first one was wrong in a way worth explaining. A full-size SD card disappears almost entirely into this reader. Shooting it in place, the only hot spot the thermal camera can see is through the slot opening, and those readings came in at 134.8 F and 132.3 F. That looked like a reasonable card temperature. It was not. It was the hottest thing visible through a gap, and the actual card was 10 to 14 F hotter than the gap suggested.

The fix is unglamorous: end the write, eject the card, and photograph the card itself. The lesson generalizes, and it is the same one a USB drive taught this bench earlier. A thermal camera reports the hottest point it can see, which is not the same as the hottest point there is.

Neither card lost any speed as it heated, which is the part that actually matters. Hot is only a problem if it costs you throughput, and across every run here it did not.

Are they really 128 GB? Every byte, verified

The capacity test fills the card with verifiable data, reads every byte back, and compares. It is slow and it is the only test that catches a fake.

Both cards passed with nothing to argue about. The ProGrade delivered 128.08 GB and the Lexar 128.01 GB, against a 128 GB label, with zero mismatched bytes and zero read errors on either. Both are marginally over their stated size rather than under.

Lexar is unusual in publishing its own definition, stating that 1 GB means 1 billion bytes and warning that usable storage may be less than stated. Measured, there is no shortfall at all. If your computer shows about 119 GB, that is the same 128 billion bytes counted in binary units, not a card that came up short.

Neither card is made by the company on the label

Every SD card carries a small identity register that names who actually built it. Reading both cards on a Raspberry Pi, with no USB bridge in the path, gives an answer that is normal for this industry and still a little surprising.

The ProGrade reports manufacturer ID 0x27, which is Phison, a contract manufacturer that builds for many brands. The Lexar reports 0xAD with an OEM code of LS, the Longsys signature, which is consistent with Lexar having passed from Micron to Longsys ownership. Neither is a red flag. Almost nobody who sells memory cards operates their own fab, and a brand is a specification and a warranty rather than a factory.

It is worth knowing because a mismatch between the brand and the manufacturer ID is also the signature of a counterfeit. The difference is coherence: these two report exactly what their real manufacturers should report, with manufacture dates that line up with the codes printed on the cards. A fake tries to report the brand it is pretending to be. I took one apart in this teardown of a counterfeit SanDisk.

Two scores, one recommendation

The bottom line on both cards

These are two good cards, and the bench separates them by less than their marketing does. Both sustain a video class above the one printed on them, both verified every byte of their capacity, neither has a cache cliff and neither stalled once. If you have a UHS-II camera and you want a card that will not surprise you in the middle of a take, either one does the job.

The Lexar edges it, on two grounds that have nothing to do with peak speed. It writes meaningfully faster in sustained use, and it kept the promise it made, where the ProGrade fell 17 percent short of a 130 MB/s write figure it chose to publish. The lifetime warranty does not hurt either. The ProGrade answers with the steadier random-write behavior of the two, by a factor of ten.

The bigger caution applies to both: check that your camera actually has a UHS-II slot before buying either one. Without it, both of these cards perform like ordinary UHS-I cards, and a good UHS-I card costs a great deal less per gigabyte.

Disclosure: I bought both of these cards myself, at retail, and no manufacturer had any involvement in the testing or saw this review before it published. This post contains affiliate links. As an Amazon Associate I earn from qualifying purchases, at no extra cost to you. Affiliate links like these and my YouTube channel are what keep the site running.