Hashwarp

Open source · No fee

The official miner for Parallax

Hashwarp mines XHash on NVIDIA and AMD GPUs, on a pool or solo on your own node. It is open source, takes no fee, and watches every card's temperature while it works.

Download HashwarpView the source
Ready forHiveOSmmpOS

Linux and Windows · CUDA and OpenCL

Hashwarp's console mining on an RTX 4090: hashrate, accepted shares, temperature, fan, power and efficiency

01Quick start

Mining in one command

Run it without arguments and a short wizard sets it up. Or give it a pool and your address, and it starts.

  • A wizard on first run

    It asks whether you mine on a pool or solo, offers the Parallax pools that take an address, saves the answers to hashwarp.json and starts. Next time it starts straight away.

  • Every pool's stratum

    EthereumStratum 1.0, EthProxy and plain Stratum, detected from the pool, and EthereumStratum 2.0 when you ask for it. With TLS and failover pools. If a pool drops, it reconnects with backoff and never gives up the only pool it has.

  • One file for a whole farm

    Every option works on the command line or in a JSON file. Copy one hashwarp.json to every rig and override what differs on the command line.

./hashwarp

Starts the wizard. The answers are saved for next time.

Compare pools on the pool index

02Console

Everything a rig needs to tell you

Hashrate, temperatures and every share, on one screen you can read at a glance.

Hashrate you can trust

1 and 15 minute averages beside the effective hashrate the pool sees, so a drop shows up before the payout does.

A row for every GPU

Hashrate, temperatures, fan speed, power draw and efficiency for each card.

One line per share

Each share with its difficulty and latency, and when a pool rejects one, the reason it gave.

Tuned at start

On NVIDIA cards, Hashwarp times its two search kernels on each GPU at start and keeps the faster one.

Keys while it mines

h
Show these keys
s
Statistics
p / r
Pause or resume all GPUs
0–9
Pause or resume one GPU
c
Connection details
l
Debug output
q
Quit

03Safety

Built to run unattended

Rigs run overnight and in back rooms. Hashwarp protects the cards and tells your scripts what happened.

Temperature limits

Pause a GPU when its core, memory or hotspot reaches a limit, and resume it once it cools. One value for all cards, or one per card.

A last-resort stop

Stop mining entirely if any GPU reaches the temperature you set.

A watchdog

Restarts a GPU whose mining failed. If a card stops responding, it runs a script of yours, to reboot the rig for example, and exits with its own code.

Every share checked

Each solution is verified on the CPU before it's submitted, so an overclock pushed too far shows up as bad results on that GPU, not as rejected shares.

A self-test

hashwarp --selftest mines offline for up to 30 seconds and checks every GPU's results. Run it after changing drivers or an overclock.

Through every epoch

The DAG changes about every five days. If building it fails on a GPU, Hashwarp retries instead of leaving the card idle.

Exit codes for your scripts

CodeMeaning
0Normal exit, or --selftest passed
1Invalid options, or --selftest failed
2Runtime error, such as no usable GPU
5Watchdog: a GPU kept failing or stopped responding
6A GPU reached --shutdown-temp

04Monitoring

Watch every rig from one place

Turn on the API and each rig serves a dashboard, a JSON summary and Prometheus metrics.

  • A web dashboard

    Hashrate, shares and every GPU's temperatures in a browser, for this rig and any others you add with --rig.

  • A JSON API

    A summary at /api/v1/summary for your own scripts and tools.

  • Prometheus metrics

    /metrics for Grafana dashboards and alerts across a farm.

  • Closed unless you open it

    Add a token to require one on every request. Pausing, resuming and restarting from the API stay off until you allow them.

Hashwarp's web dashboard, showing four rigs and this rig's GPU
One rig's dashboard, watching three more rigs added with --rig.

Turn it on

hashwarp --api-bind 127.0.0.1:21550 --api-token <token>

  • http://127.0.0.1:21550
  • /api/v1/summary
  • /metrics

05Rig operating systems

Ready for HiveOS and mmpOS

Each Linux release has a package for both, reporting hashrate, shares, temperatures and fans to the rig OS.

HiveOS

Add it to a flight sheet as a custom miner:

Miner name
hashwarp
Installation URL
hashwarp-<version>.tar.gz
Hash algorithm
xhash
Wallet and worker
%WAL%.%WORKER_NAME%

mmpOS

Create a miner profile with the Custom miner type, and give it the download URL of the release's mmpOS package, the one ending in -mmpos.tgz, and a pool. Extra Hashwarp options go after the arguments.

Both rig OSes restart Hashwarp when it exits, which is what its watchdog relies on.

HiveOS and mmpOS setup

06Download

Get Hashwarp

Free and open source under GPL-3.0, for Linux and Windows.

Download Hashwarp

What it needs

NVIDIA
Driver 525 or newer
AMD
The AMD driver with OpenCL (ROCm or Adrenalin)
Linux
glibc 2.28 or newer: Ubuntu 20.04, Debian 10, RHEL 8 and later, current HiveOS and mmpOS
Windows
Windows 10 or newer; the Visual C++ runtime comes in the archive
Hashwarp's documentation, with tuning and troubleshooting

New to mining Parallax? Rewards, difficulty and pools, on one page.

Mining Parallax