USB drive, microSD card or UDP module — which one should I actually buy?
Choose by how the storage connects to the product, not by capacity or price. A USB drive is for products where the user plugs and removes it — promotional gifts, software delivery, portable storage. A microSD card is for products with a card slot where the media is expected to be replaceable or upgraded by the user — cameras, dashcams, phones, consumer devices. A UDP module is for soldered, permanent storage inside a sealed product — IoT devices, wearables, industrial controllers, anything that must not be opened. If the user never sees the storage, you want a module, not a card.
The one question that decides it
Ask: **does the end user ever touch this storage?**
If yes, and they plug it into a host computer, that is a USB drive. If yes, and they insert it into a slot in the device, that is a card. If no, and it lives inside the product until the product dies, that is a module.
Almost every downstream difference — cost, reliability, failure mode, certification — follows from that one answer.
| USB drive | microSD card | UDP module | |
|---|---|---|---|
| Connection | Plugs into a host port | Slot in the device | Soldered to the PCB |
| User removable | Yes | Yes | No |
| Typical use | Gifts, software, portable data | Cameras, dashcams, phones | IoT, wearables, industrial |
| Failure mode | Connector wear, loss | Slot contact, removal damage | Solder joint, thermal |
| Field replacement | Trivial | Easy | Not possible |
Why the failure modes differ so much
The USB drive's weak point is the connector and the fact that it leaves the building. A drive that is lost is a drive that failed, and no amount of testing prevents that. For a promotional product this is acceptable; for anything with operational data it is not.
The card's weak point is the slot. Contacts oxidise, cards get inserted at an angle, and users remove them while the device is writing. The SD Association design anticipates this, which is why cards and hosts implement power-fail handling — but that only mitigates, it does not remove, the risk.
The module removes both problems by removing the connector entirely. What it introduces instead is permanent attachment: if the module fails, the product fails, because there is nothing to swap. That is why module selection rewards specifying the memory grade and controller carefully, and why soldering conditions matter more than they seem.
Cost is the wrong first filter
Modules are often cheaper per gigabyte in volume than a finished drive, because there is no housing, no connector and no packaging. But if your design has a card slot, switching to a module means a PCB redesign, not a cheaper BOM line.
Conversely, if you are shipping a sealed device and using a card to save the cost of soldering, you are one connector failure away from a returned product that cannot be repaired in the field.
Start with the architecture, then price within it.
Sources
- SD Association — SD and microSD specifications
- USB.org — USB mass storage class and connector specifications
- JEDEC — memory package and interface standards
Figures quoted on this page are from the published reporting listed above. Where a number is an estimate rather than a published figure, it is described as approximate.
Frequently asked questions
Can I use a microSD card soldered permanently instead of a UDP module?
Not as a drop-in. A card and a module are different packages with different footprints, different thermal paths and different electrical characteristics, even when the memory inside is similar. Soldering a card is a common informal practice, but it forfeits the card standard's guarantees and is not a design you can specify to a manufacturer as a module. If you want permanent storage, specify a module.
Which is best for an IoT device that logs data for years?
A soldered module, usually. Logging is a continuous-write, long-duration workload, which stresses endurance and thermal behaviour rather than peak speed. A module lets you specify memory grade and wear behaviour as part of the design, and removes the connector and the possibility of a user removing the media mid-write. If the device must be serviceable, a card slot is a legitimate choice — but size the endurance for the write rate and decide in advance what happens when the media wears out.
Is a USB drive ever the right choice inside a product?
Yes, when the storage is meant to be handed over — a configuration file, a firmware update, a set of media that ships with the product. In that role the removability is the feature, not a risk. The mistake is using a USB drive as internal storage because it is easy to source, which turns a consumer accessory into a component with a connector that was never designed for permanent service.
Working on a specification?
Send us the product type, capacity, quantity and target date. We reply to business enquiries within 1 business day.
Request a quote