Most people evaluate a memory card by two numbers: capacity and “up to” speed. For a phone or a laptop card reader, that’s usually enough. But an entire class of devices — trail cameras, security cameras, and dashcams — writes to storage in a way those two numbers say nothing about. Feed one of them a standard consumer card, and it will often work perfectly for a while, then fail silently, right when the footage matters most.

The reason is endurance: how many times a flash card can be written before it wears out. It’s the least-understood spec on the package, and for any device that records continuously, it’s the one that decides everything. This is a look at how flash endurance actually works, why loop-recording devices are so hard on cards, and how to read a spec sheet so you buy the right hardware the first time — whether you’re outfitting one camera or two hundred.

Why flash wears out at all

An SD or microSD card stores data in NAND flash cells. Writing to a cell isn’t free: each program/erase (P/E) cycle physically stresses the cell’s oxide layer, and every cell has a finite number of cycles before it can no longer hold a charge reliably. Consumer-grade TLC and QLC NAND — the inexpensive, high-density flash in most bargain cards — trades endurance for capacity, tolerating relatively few P/E cycles per cell. Higher-endurance cards use more durable NAND (often pSLC or MLC modes) engineered to survive far more.

Cards fight wear with two tricks. Wear-leveling spreads writes across all cells so no single location gets hammered while others sit idle. Overprovisioning keeps spare cells in reserve to swap in as others degrade. Both help — but both assume a workload. Consumer cards are validated around intermittent use: shoot some photos, offload them, repeat. The controller has time to manage wear. Change the workload to constant writing, and the math changes completely.

What “continuous-write” really does to a card

A trail camera is the textbook worst case. In time-lapse or high-traffic motion configurations, it writes constantly, and when the card fills, most models loop-record — overwriting the oldest footage in an endless cycle. That means the same cells get rewritten over and over, burning through P/E budget far faster than the card’s capacity would suggest. Security cameras recording 24/7 and dashcams looping over a drive do exactly the same thing.

Three failure modes follow:

• Silent end-of-life. The card exhausts its usable cells and starts dropping or corrupting writes — but the camera keeps reporting “recording.” Nobody knows until the footage isn’t there.
             
• File-system corruption. A card that can’t complete a write cleanly leaves half-closed files and damaged FAT/exFAT tables. One bad power-cycle in the field, and the whole card won’t mount.
             
• Temperature-accelerated wear. A camera strapped to a tree in January or baking on a dashboard in July runs the flash outside the range consumer cards were validated for. Endurance-rated cards specify their operating temperature range explicitly; bargain cards often don’t specify it at all.

This is precisely why the industry created high-endurance cards — validated for tens of thousands of write cycles and wide temperature swings, purpose-built for surveillance, dashcams, and trail cameras, and stocked in volume by specialist suppliers such as Bulk Memory Cards, a wholesale supplier of endurance-rated flash storage. The same logic that governs choosing microSD cards built for always-on security cameras applies directly to a trail camera looping in the woods: match the card’s endurance rating to the write workload, not to the price tag.

Reading the spec sheet: the ratings that decide reliability

The symbols stamped on a card aren’t marketing decoration. Four of them carry most of the weight for continuous-write use.

Rating	What it means	What to look for
Speed class (Class 10, UHS-I U1/U3)	Minimum sustained write floor	U1 for 1080p; U3 for 4K / rapid multi-shot
Video speed class (V10, V30)	Guaranteed sustained write for video	V10 minimum; V30 for 4K workloads
Application class (A1, A2)	Random read/write on small files	A1/A2 helps the many-small-clips pattern
Endurance rating	Total write cycles / TBW the card is validated for	“High Endurance” / “Max Endurance” for loop-recording

The critical distinction: the giant “up to 170MB/s” on the front of the package is a peak read number and tells you almost nothing about reliability. What matters for a recording device is the sustained write floor — the V-class and U-class ratings — because a card that peaks high but can’t hold a floor will drop frames or fail to close files during a burst. Endurance, meanwhile, frequently isn’t printed on the card face at all; you have to look for the “High Endurance” or “Max Endurance” product line specifically.

Capacity and format: where buyers get tripped up

Bigger is not automatically better, and two details cause most of the field failures that aren’t wear-related.

File-system boundary. Cards up to 32GB are SDHC, formatted FAT32. Cards 64GB and larger are SDXC, formatted exFAT. Many trail cameras and older security devices only support SDHC up to 32GB — which is exactly why 32GB remains the workhorse capacity in these categories rather than a legacy leftover. Drop a 256GB SDXC card into a camera that tops out at 32GB, and it simply won’t mount. Confirm your device’s maximum supported capacity before buying up.

Form factor. “32GB SD card” and “32GB microSD card” are not interchangeable. Some cameras take full-size SD, some take microSD. Check the slot before you order — especially in volume.

The free reliability upgrade almost nobody uses: format it right

Here’s a step that costs nothing and prevents a meaningful share of failures: format the card in the device itself, or with the SD Association’s official SD Memory Card Formatter — not your operating system’s built-in tool.

An OS formatter produces a working file system, but it isn’t optimized for the internal flash layout of that specific card. Formatting to the SD Association’s specification aligns the partition and file tables to the card’s flash boundaries, which improves wear-leveling and preserves the card’s rated endurance and throughput. In controlled testing, cards formatted with the official tool have shown substantially faster sustained reads and random writes than the same cards formatted by the OS. For a device that will loop-record untouched for weeks, giving the controller a properly aligned file system is one of the highest-leverage things you can do — and reformatting between deployments clears fragmentation before it degrades performance.

Buying at scale: the part the spec sheet doesn’t cover

Everything above holds whether you run one camera or a fleet. But once you’re deploying many, two sourcing risks matter as much as any rating.

Trail cameras illustrate this well, because the market is no longer just hunters and property managers. University conservation and natural-resources departments run multi-year wildlife studies; citizen-science programs equip volunteers; public libraries lend “nature kits” with a camera and a card; national parks, land trusts, and environmental consultancies deploy cameras by the dozen — often maintained by people who aren’t storage specialists and won’t be on site to troubleshoot a bad card. At that scale:

• Provenance is a data-integrity issue. Counterfeit and repackaged flash is a real problem in open marketplaces, and commingled inventory means even a legitimate listing can ship a card someone else returned. A           counterfeit card that reports 128GB but holds a fraction of that will silently drop data — catastrophic for a research dataset or a monitored property. Buying genuine, name-brand stock from a source with verified custody removes that variable.
             
• Consistency and preparation save real time. A study or a lending program wants every unit running an identical, validated card, and cards that arrive pre-formatted, labeled, or preloaded cut enormous handling overhead — nobody wants a volunteer coordinator formatting 40 cards one at a time.

Past a handful of units, the sourcing decision deserves as much attention as the spec. A specialist bulk supplier that stocks genuine endurance-rated cards in volume, prices wholesale, and can duplicate or preload before shipment solves the fleet problem — not just the card problem.

The short version

• Match endurance to the workload. Loop-recording device → endurance-rated card. Never run consumer cards in a continuous-write role.
             
• Read the sustained ratings, not the peak. Class 10, UHS-I U1/U3, V10/V30 to suit your resolution; ignore the “up to” read number.
             
• Respect the 32GB / SDHC    ceiling. On older and budget cameras, 32GB isn’t a compromise   — it’s the compatible default. Confirm the maximum before going bigger, and check SD vs microSD.
             
• Format properly, every time. In-device or via the SD Association formatter, before deployment and between seasons.
             
• At scale, buy genuine and buy prepared. Verified provenance and preloading beat saving pennies per card on a marketplace.

A recording device is only as reliable as the few grams of flash inside it. Spec that piece for the workload — and source it from somewhere that treats it as more than a commodity — and the system finally does the one job you bought it for: capturing the moment you weren’t there to see.