Miner is restarting…
reconnecting automatically
Miner offline
no response from /metrics — showing last known values
Miner
—
Mining is paused
The miner is disconnected from the pool and its threads are stopped.
This page keeps working. A restart of the miner resumes mining too.
Total hashrate
—
raw work of the card(s)
GPU hashrate
—
CPU hashrate
—
Efficiency
—kH/W
Hashrate — last 30 minutes
raw now —
avg — peak —
effective now —
collecting samples…
Log — last 100 lines
waiting for the first status line…
Shares
Accepted
Rejected
Stale
Accepted—
Rejected—
Stale—
Submitted—
GPU & thermals
GPU temperature—
GPU load—
CPU temperature—
Power draw—
Memory used—
GPUs
| GPU | Status | Hashrate | Share | Power | Temp | kH/W |
|---|
Session
Uptime—
Difficulty—
Total hashes—
Current job—
Rig
CPU—
Mining threads—
GPUs active—
Last update—
Settings
—
—
Change one setting at a time. Clock changes are tested for 10 minutes before being saved; if the card becomes unstable, the previous value comes back automatically.
Control tokenrequired to enable the controls below
Mining configuration
Change any of these, then press Save & restart at the
bottom of this box: it saves them in config.env
and restarts the miner to apply them. Every field is checked before
anything is written; the previous file is kept as
config.env.bak.
Walletwhere the pool pays out
Poolhost name and port, no stratum+tcp://
CPU threads—
—
Graphics cards—
GPU power limit—
Memory offset—
Locked while another test is running.
GPU intensity0–100. Applied on restart.
Advanced
For experienced users. These change how the card clocks itself; a wrong value can cost hashrate or stability. The power limit and memory offset above are enough for most tuning.
Core clock—
Locked while another test is running.
Memory clock—
Locked while another test is running.
Core offset—
Locked while another test is running.
Adopt current state
Save what the card is running now, without a trial. For a rig
tuned outside this miner. Offsets are always adopted — they can be
read back exactly. Clock locks cannot: NVML has no way to report them,
so tick only what you know is locked.
Dropped shares
Found by the card but never sent: they came too soon after the
previous share. Not an error. The total comes mostly from the first
minutes after a start; the last 10 minutes is what counts.
More
—
GPU candidates
Shares the card found in its batches, and how many of them came
second or later in the same batch. Before this version the kernel
reported one share per batch and the rest were lost.
More
—
Enter the control token to enable these controls.