RMA (Return Merchandise Authorization)

RMA, short for Return Merchandise Authorization, is the formal process by which a buyer returns a product — in this world, usually a faulty ASIC miner or one of its major components — to the seller or manufacturer for repair, replacement, or refund. The defining word is "authorization": before shipping anything back, the buyer must obtain approval, typically a ticket or reference number, so the vendor can track the unit, confirm warranty status, and route it to the right service bench. Returning hardware without that authorization risks it being refused at the dock or lost in an unlabeled pile.

The typical ASIC RMA flow

Although miners come from many manufacturers, the RMA flow is nearly identical across the industry. You contact your supplier and describe the fault; a technician usually walks you through basic triage first — power cycling, checking the kernel log, reseating cables, testing with a known-good PSU — because a meaningful share of "dead miner" reports are cabling or power issues that never needed to ship anywhere. If the unit genuinely needs service, you open an RMA ticket detailing the symptom, serial number, purchase date, and shipping information. The vendor then authorizes the return and tells you whether to send the whole machine or just the faulty module — often a single hashboard or PSU, which cuts shipping cost and turnaround dramatically. Photograph everything before packing, keep the ticket number on the box, and use packaging that can survive a freight depot, because transit damage disputes are the ugliest part of any RMA.

Who actually handles the warranty

A key point for buyers: manufacturer warranties typically run to the manufacturer's direct customers. If you bought through a reseller or broker, your RMA usually goes back through that reseller, who then deals with the maker on your behalf — adding a hop, and time, to the process. Buyers commonly pay shipping to the service center while the vendor covers the return leg, though terms vary. Cross-border RMAs add customs paperwork and can stretch turnaround from weeks toward months, which is why many operators keep spare boards on the shelf or build a relationship with a local repair bench instead of shipping overseas for every fault. Hosting customers often have it easier: a good colocation provider manages RMAs in bulk on clients' behalf, with standing accounts at the manufacturers.

RMA, repair, or refurb: choosing the path

An RMA is the warranty path, and when a unit is in warranty it is usually the right call — the labor is free even if the shipping and downtime are not. Out of warranty, the economics flip: independent repair of a hashboard is routinely cheaper and faster than manufacturer service, and a board-level fix keeps a machine earning instead of waiting in a queue. That is the niche a repair-first shop like D-Central occupies: diagnose at the board level, fix what is fixable, and be straight with you when a board is not worth saving. (On the sales side, D-Central's own return policy is deliberately narrow — DOA-only returns, with warranty handled through the manufacturer or D-Central depending on the product — because build-to-order hardware is not restockable inventory.) Machines that fail RMA economics entirely feed the refurbished ASIC channel, where working pulls and repaired units get a second life. If you are weighing warranty return against local repair, start a repair intake and compare the quote against the RMA's true cost: shipping both ways, weeks of lost hashing, and the risk of a "no fault found" fee.