Setup

The same steps for every coin. Pick yours, and the values in blue follow it.

Further down: Miner options · Troubleshooting

Coin

Requirements

System

Windows	64-bit
Linux	glibc 2.31 or newer (Ubuntu 20.04 and later, HiveOS)

NVIDIA

GeForce	RTX 30, 40 and 50 series, desktop and laptop
Workstation	RTX A2000 to A6000, RTX 2000 to 6000 Ada, RTX PRO Blackwell
Data center	A2, A10, A16, A40, L4, L40, L40S
Driver	NVIDIA 580 or newer

System

Unzip opminer-win64-1.0.1.zip.
Open pearl.bat in Notepad and put your PRL address on the WALLET line. Save.
Double-click pearl.bat.

If Windows Defender removes opminer.exe, see Troubleshooting.

Recommended: check the zip matches the SHA-256 on the download list. In PowerShell: Get-FileHash opminer-win64-1.0.1.zip

The GPU is shared with whatever else uses it, so the hashrate drops while you game or watch video on it. Keep Windows' Hardware-Accelerated GPU Scheduling on (Settings, System, Display, Graphics).

Miner options

The settings at the top of pearl.bat (pearl.sh on Linux). Leave one empty for its default.

Pool 4 settings

`WALLET`	Your PRL address; the pool pays it. The miner won't start until it's set
`WORKER`	A name for this PC; the pool shows it on its stats page
`POOL`	The pool to mine on, e.g. `stratum+tcp://prl-us.kryptex.network:7048`; any PRL pool works
`BACKUP_POOLS`	Tried in order while `POOL` is down; space-separated (in quotes on Linux), empty for none

Optional 4 settings

`DEVICES`	The GPUs to mine on, as `nvidia-smi` numbers them, e.g. `0,2` (empty: every one)
`TEMP_LIMIT`	Pause a GPU at this many °C; 10 °C below, it mines again
`MEM_TEMP_LIMIT`	The same for its memory (GPUs that report it)
`EXTRA`	Any other options, separated by spaces; the useful ones are under Extra options below

Overclock 7 settings

Needs "Run as administrator" (sudo ./pearl.sh on Linux), and is put back when the miner stops. One value for every GPU, or one per GPU: 1800,1750,- (- leaves a GPU alone).

`CORE_LOCK`	Lock the core clock, in MHz
`MEM_LOCK`	Lock the memory clock, in MHz
`CORE_OFFSET`	Raise (or, negative, lower) the core clock, in MHz
`MEM_OFFSET`	Raise (or, negative, lower) the memory clock, in twice the MHz
`POWER_LIMIT`	Power limit, in watts
`FAN`	Fan speed, in %
`OC_PERSIST`	`yes` leaves the overclock on (until a reboot, or `opminer --oc-reset`)

Extra options 6 options

Add them to EXTRA, separated by spaces: set EXTRA=--cpu-check --api-port 4079 (in quotes in pearl.sh; on HiveOS and mmpOS, in the flight sheet's Extra config arguments).

`--cpu-check`	Check each share on the CPU before sending it; use it while tuning an overclock
`--api-port PORT`	The stats API's port (default 4078); give a second miner on the same PC its own, `0` turns it off
`--pass PASS`	Pool password, for pools that read settings from it (e.g. suprnova's `d=N`)
`--no-cert-validation`	TLS pools with self-signed certificates
`--allow-sleep`	Windows: let the PC sleep while mining
`--temp-resume C`	Where a GPU paused by `TEMP_LIMIT` mines again (default 10 °C below)

Everything else: opminer --help (in a terminal in the miner's folder).

Troubleshooting

Windows Defender deletes opminer.exeThreats found

It does that with most miners. If the zip's SHA-256 matches the download list, it's a false positive: add an exclusion for the miner's folder (Windows Security › Virus & threat protection › Manage settings › Exclusions), then unzip again.

The miner won't startno wallet address yet

Put your address on the WALLET line of pearl.bat (pearl.sh on Linux), then save.

The driver is too olddriver too old (…; R580 or newer needed)

Update the NVIDIA driver to 580 or newer (HiveOS: nvidia-driver-update 580).

The pool refuses the login… refused the login

Check the wallet address and the worker name.

A GPU is skippedunsupported GPU

It isn't one the coin supports; the list is in the requirements above.

The overclock isn't appliedoverclocking needs administrator rights

Run it as administrator (sudo on Linux), or leave the overclock settings empty.

A GPU fails its self-testSELF-TEST FAILED

Lower the overclock, and check the temperatures.

An overclock is still onafter a crash or power cut

Run opminer --oc-reset as administrator (sudo on Linux). It puts every GPU's clocks, power limit and fans back, then exits.

Anything else: ask on Discord.