Dedicated capacity with a private ops channel and custom power planning. Pod is for fleets that need predictable expansion, written rate certainty before a multi-unit ship, and cooling that is not fighting a shared neighborhood. You pay for density and attention — use it when bay and rack math already proved the thesis and surprise dock arrivals would hurt.
← ALL GUIDESHOSTING
HOSTING IS NOT A RENTAL.
BAY · RACK · POD
MONTHLY COST IS
UPTIME IS A CONTRACT.
TWO LEGAL WAYS
REMOTE HANDS EXIST.
BEFORE ANY CARRIER
FACILITY QUESTIONS
RAR / HOSTING GUIDE · MID · ~13 MIN
HOSTING
BASICS.
Hosting means your ASIC lives in a Rent A Rig facility: you keep title to the steel, and RAR supplies the power plant, cooling, remote hands, and an uptime SLA. You are not buying hashrate by the day the way a rental works — you are paying for powered space and operations, usually from a $/kWh rate card that tightens as you move from bay to rack to pod. This guide is the facility blueprint: what hosting is and is not, how to size a tier against wattage, how to read the rate math, how intake shipping actually lands a machine on telemetry, and what day-2 ops look like before you schedule an exit ticket.
OWNERSHIP BRIEF
HOSTING IS NOT A RENTAL.
AND IT IS NOT A FREE-FOR-ALL.
On a rental you pay for allocation time on RAR infrastructure and you do not take title. On hosting you already own the machine — or you buy one and route it into intake — and you pay for the watts and the hands that keep it alive. That distinction matters when something breaks: residual value, resale on the marketplace, and the right to relocate later all stay with you, while facility downtime maps to SLA credits instead of “you bought a bad week of hashrate.” Hosting is also not self-colocation with no rules. Power envelopes, noise, and firmware policies exist so one tenant’s exotic flash does not take a row offline. If you are still deciding whether ownership is even the right move this quarter, read Rent vs Buy first; if you already know you want steel in a hall you do not have to staff, stay here and size the tier honestly.
FACILITY ELEVATION
BAY · RACK · POD
AS POWER CEILINGS.
Always match SKU wattage to the tier’s continuous power ceiling. A M160-class draw will not fit an underspec bay no matter how good the $/kWh looks on the card. Density without headroom is just a polite way to trip breakers.
MATCH WATTS FIRST
Multi-unit allocation with higher aggregate kilowatts and high flow cooling. Rack is the operator pick when you are past one or two HashCore-class machines and want density without negotiating a full pod. Lower $/kWh than bay usually assumes commitment and fill — ask for a written rate before you ship a half rack of someone else’s marketplace units.
Single-machine slot in a shared power neighborhood at the entry rate. Bay is correct for one or two units while you learn intake, remote hands, and how hosting bills feel against your hashrate. It is the wrong place to force a machine that already sits at the ceiling — leave headroom for PSU reality and startup, then graduate to rack when total wattage demands it.
POWER RATE MATH
MONTHLY COST IS
kW × HOURS × RATE.
Hosting cards show rates like FROM $0.068 / kWh. Your planning monthly power cost is roughly average kilowatt draw × 24 × days in the month × the contracted rate. Use the product page power rating as a ceiling for planning; live telemetry will sit near that when the unit is hashing hard. Lower $/kWh on rack and pod tiers usually assumes longer commitment and higher density — a teaser rate is not a purchase order. Get the number in writing from Partnerships or Support before you put ten machines on a truck.
AVG kW3.2
×HOURS720
×RATE$0.068
≈MONTH (ILLUST.)~$157
ILLUSTRATIVE MONTHLY POWER · SAME 3.2 kW DRAWNOT A QUOTE · RATE CARD MAY MOVE
The chart holds draw constant so you can see rate alone. Real fleets change both: denser tiers may host hotter SKUs, and contracted rates can include commitment floors. Re-run the spreadsheet with your actual nameplate watts and the written rate you were given — never with a homepage teaser alone.
SLA TARGETS
UPTIME IS A CONTRACT.
CREDITS ARE NOT MAGIC.
Hosting tiers publish uptime targets — 97% on bay, 99% on rack, 99.5% on pod. Facility-caused downtime outside maintenance windows maps to credits; misconfigured workers do not. Read Uptime & Credits for the claim logic, and keep monitoring yourself either way.
97%BAY
Entry SLA for shared neighborhoods. Credits cover facility misses, not pool typos. Good enough for a first hosted unit while you learn intake — upgrade the target when fleet value makes the tighter SLA economically obvious.
99%RACK
Operator-grade target for multi-unit allocations. Still does not cover self-inflicted offline events. Pair the SLA with dashboard alerts so you notice a dark worker before the monthly credit reconciliation does.
99.5%POD
Dedicated-channel expectation for serious fleets. Maintenance windows still exist — published windows are not “free downtime you ignore.” Credits realign incentives when the hall fails you; they never replace a written ops channel you actually use.
INTAKE CORRIDORS
TWO LEGAL WAYS
ONTO THE DOCK.
Never surprise the dock. Every machine that racks into RAR should arrive on a path the facility already expects — either store purchase with hosting delivery selected, or a inbound ship tied to an intake ticket ID on the label. Photograph serials before the carrier leaves, remove loose screws and rattling accessories, and declare accurate weight and value for insurance. An unlabeled box is not a hosting strategy.
PATH A · STORE
BUY AND ROUTE TO HOSTING
Purchase from the Rent A Rig store and select HOSTING delivery so logistics routes straight into intake. The SKU, wattage, and destination are already known to the hall, which shortens the surprise surface. You still confirm the hosting tier and written rate before the buy if you are filling more than a single bay — checkout convenience is not a substitute for power planning. After payment, keep the order ID next to the future asset tag so Support can stitch commerce to the rack without archaeology.
PATH B · BRING YOUR OWN
SHIP WITH AN INTAKE TICKET
Open a hosting request first. Get the intake address and RMA-style label instructions, then send tracking to Support before the truck moves. Marketplace units and previously self-hosted machines are welcome when they fit power and thermal rules — open approval with model, wattage, and photos before you spend on freight. On arrival the sequence is inspect → asset tag → rack → power → dashboard agent → your pool config. You should see telemetry before you celebrate or redirect payouts.
01INSPECT
02ASSET TAG
03RACK
04POWER
05AGENT
06POOL
07TELEMETRY
DAY-2 OPERATIONS
REMOTE HANDS EXIST.
POLICY STILL BINDS.
Remote hands can reseat power, swap PSUs when stocked, and check networking when a unit goes dark for reasons that are not your pool string. Firmware changes follow facility policy — ask before flashing anything exotic, because one bad image can become a row problem. Treat Support like a partner with a ticket ID, not like a chat that owes you instant miracles without serials, photos, or timestamps.
When you leave hosting, schedule an exit ticket. RAR will de-rack and stage for outbound shipping; you pay outbound logistics unless your contract says otherwise. Do not assume a marketplace listing you posted last night automatically empties a bay — the hall needs the exit workflow the same way it needed intake. Plan the calendar the way you planned the wattage: early, written, and without surprise freight on either end. Shipping policy and warranty pages cover carrier and defect edges; this guide only insists you never ghost the dock.
PRE-SHIP TAPE
BEFORE ANY CARRIER
PICKS UP STEEL.
TIER — Bay, rack, or pod is chosen against total continuous wattage with headroom, not against which card looks premium on the hosting page.
RATE — Written $/kWh and SLA are confirmed for the density you are actually shipping, especially before multi-unit freight.
TICKET — Intake ticket ID exists before pickup, and the same ID is on the label the carrier will scan.
PATH — Store-to-host or ship-with-tracking is chosen on purpose; the dock is never the first team to learn a machine is coming.
PROOF — Serial photos, accessory strip-down, and honest weight/value declarations are done before the box closes.
LIVE — Telemetry online is verified before you update pool payouts or tell anyone the fleet “is hashing.”
DESK FAQ
FACILITY QUESTIONS
BEFORE YOU FREIGHT.
CAN I HOST A NON-HASHCORE UNIT?
Often yes if it fits power and thermal rules. Open a hosting request with model, wattage, and photos for approval before shipping so the dock is not inventing a policy under a pallet. Approval is cheaper than a refused intake and a return leg you did not budget.
WHO OWNS THE MACHINE IN HOSTING?
You do. RAR provides facility services — power, cooling, remote hands, SLA credit logic. Title does not transfer unless you separately sell the unit on the marketplace or elsewhere. That is why hosting economics can look like ownership without the warehouse, and why exit shipping is your logistics problem unless the contract says otherwise.
HOW FAST IS RACK-IN AFTER ARRIVAL?
Typical intake is measured in hours to a couple of business days depending on queue depth and inspection findings. Damaged freight, missing ticket IDs, and undocumented modifications slow the line for everyone. Clean paperwork and honest photos are the fastest rack-in tools you control from outside the building.
WHAT IF MY HOME POWER LOOKS CHEAPER?
Compare all-in: your $/kWh, your uptime staffing, noise limits, and the capital you would spend on switchgear and cooling. Hosting rates include the plant you are not building. A lower residential sticker rate can still lose when the machine sits dark because nobody is on site at 3 a.m. Run the spreadsheet with honesty, then pick the path that matches who carries power risk.