3763996ae698c96be111c8711a17bc0914bf5be5
ci / fmt (pull_request) Successful in 20s
bench / build (pull_request) Successful in 1m16s
ci / clippy (pull_request) Successful in 1m44s
bench / measure (pull_request) Successful in 2m52s
ci / doc (pull_request) Successful in 2m1s
ci / test (pull_request) Successful in 6m59s
The closed-source qpow-cuda kernel runs 2x our hashrate on the same silicon with the same 1472 multiplies per hash. Its PTX shows why: every field operation is one self-contained inline-PTX block, the 128-bit product is reduced with mad.lo.cc/madc.hi.cc (one IMAD.WIDE with a carry-out) plus three ALU ops, the internal layer's row sum rides in the mad.wide partial products for free, and accumulators are 32-bit limb chains that ptxas turns into IADD/IADD.X with merged carries. Ours paid a compare-and-select reduction (~14 ops), a 64-bit add-with-carry per diagonal element, and five instructions per accumulator add on sm_120. This ports those primitives behind LAIR_FUSED_MUL (default 1): gf_mul, gf_sqr (three-product squaring), gf_mul_add with the addend folded into the partial products, acc_add/acc_add2 as limb chains, and a four-op acc_reduce. The round constant now joins the unreduced sum in the internal layer. LAIR_EXACT_REDUCE (default 0) keeps the reduction bit-exact for three more ops per multiply. With 0 the final borrow is dropped, as in their kernel: wrong only when the product's bits 64..95 are zero and its low 64 bits are below its top 32 bits, about 2^-64 per multiply (emulated against a reference: 0/3M random mismatches, 2^48*2^48 fails as predicted). Exactness measured at -15% on beast, so it is off. Measured, three interleaved rounds, parity 40/40 on each host: beast 2x5090 1356 -> 2146 MH/s (+58%) benjy 4090 460 -> 711 MH/s (+54%) quadbrat 3060 84 -> 140 MH/s (+66%) Static sm_120 31,280 -> 20,888 instructions, sm_86/89 39,208 -> 22,576, no spills. ncu: 22,913 instructions per nonce against their 22,233. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Ue5ZZm1Hiv5zPnucykKKuF
External Miner Service for Quantus Network
High-performance external mining service for Quantus Network with support for CPU, GPU, and hybrid CPU+GPU mining.
Building
# CPU-only build (default)
cargo build -p miner-cli --release
# With GPU support (recommended)
cargo build -p miner-cli --release
The binary will be available at target/release/quantus-miner.
Running
The node requires a shared auth token and TLS cert pin. Both live under the
node's chain config dir (<base-path>/chains/<chain>/):
miner-auth-token— shared secret (not logged by the node; read this file)miner-tls-cert-sha256— SHA-256 of the miner TLS cert (also printed in node logs)
# Preferred: mount/read the node's chain config files
./target/release/quantus-miner serve \
--node-addr 127.0.0.1:9833 \
--auth-token-file /path/to/miner-auth-token \
--tls-cert-sha256-file /path/to/miner-tls-cert-sha256 \
--cpu-workers 4
# Or pass values directly (token from miner-auth-token; fingerprint also in node logs)
./target/release/quantus-miner serve \
--node-addr 127.0.0.1:9833 \
--auth-token <TOKEN> \
--tls-cert-sha256 <FINGERPRINT> \
--gpu-devices 1
# Hybrid CPU+GPU mining
./target/release/quantus-miner serve \
--node-addr 127.0.0.1:9833 \
--auth-token-file /path/to/miner-auth-token \
--tls-cert-sha256-file /path/to/miner-tls-cert-sha256 \
--cpu-workers 4 \
--gpu-devices 1
Configuration
| Argument | Environment Variable | Description | Default |
|---|---|---|---|
--node-addr <ADDR> |
MINER_NODE_ADDR |
Node address to connect to | 127.0.0.1:9833 |
--auth-token <TOKEN> |
MINER_AUTH_TOKEN |
Shared secret from the node's miner-auth-token file (not logged) |
required |
--auth-token-file <PATH> |
MINER_AUTH_TOKEN_FILE |
Read the shared secret from a file (preferred) | — |
--tls-cert-sha256 <HEX> |
MINER_TLS_CERT_SHA256 |
SHA-256 of the node's miner TLS cert (miner-tls-cert-sha256 / node logs) |
required |
--tls-cert-sha256-file <PATH> |
MINER_TLS_CERT_SHA256_FILE |
Read the TLS cert fingerprint from a file | — |
--cpu-workers <N> |
MINER_CPU_WORKERS |
Number of CPU worker threads | Auto-detect |
--gpu-devices <N> |
MINER_GPU_DEVICES |
Number of GPU devices | Auto-detect |
--gpu-batch-size <N> |
MINER_GPU_BATCH_SIZE |
GPU batch size in nonces | 1000000 |
--cpu-batch-size <N> |
MINER_CPU_BATCH_SIZE |
CPU batch size in hashes | 10000 |
--gpu-throttle-ms <MS> |
MINER_GPU_THROTTLE_MS |
Sleep duration (ms) between GPU batches | 0 |
--metrics-port <PORT> |
MINER_METRICS_PORT |
Prometheus metrics port | 9900 |
GPU Mining
GPU support uses WGPU for cross-platform acceleration:
- macOS: Metal backend (Apple Silicon & Intel)
- Linux: Vulkan/OpenGL backends
- Windows: DirectX 12/Vulkan backends
Setup
Build with GPU support:
cargo build -p miner-cli --release
Platform requirements:
- macOS: Works out-of-the-box
- Linux: Install GPU drivers (
nvidia-driver,mesa-vulkan-drivers) - Windows: Ensure recent graphics drivers are installed
Performance Monitoring
- macOS:
sudo powermetrics --samplers gpu_power -i 1000 - Linux:
nvidia-smi(NVIDIA) orradeontop(AMD) - Windows: Task Manager GPU tab
Examples
All serve examples need the auth token and TLS pin (files or inline values).
# CPU mining with 8 workers
./target/release/quantus-miner serve \
--auth-token-file /path/to/miner-auth-token \
--tls-cert-sha256-file /path/to/miner-tls-cert-sha256 \
--cpu-workers 8
# Pure GPU mining
./target/release/quantus-miner serve \
--auth-token-file /path/to/miner-auth-token \
--tls-cert-sha256-file /path/to/miner-tls-cert-sha256 \
--gpu-devices 1
# GPU mining with throttle (reduce GPU utilization)
./target/release/quantus-miner serve \
--auth-token-file /path/to/miner-auth-token \
--tls-cert-sha256-file /path/to/miner-tls-cert-sha256 \
--gpu-devices 1 --gpu-throttle-ms 50
# Hybrid mining: 4 CPU + 1 GPU workers
./target/release/quantus-miner serve \
--auth-token-file /path/to/miner-auth-token \
--tls-cert-sha256-file /path/to/miner-tls-cert-sha256 \
--cpu-workers 4 --gpu-devices 1
# With verbose logging
RUST_LOG=debug ./target/release/quantus-miner serve \
--auth-token-file /path/to/miner-auth-token \
--tls-cert-sha256-file /path/to/miner-tls-cert-sha256 \
--cpu-workers 2 --gpu-devices 1
# Production setup with metrics
./target/release/quantus-miner serve \
--node-addr 127.0.0.1:9833 \
--auth-token-file /path/to/miner-auth-token \
--tls-cert-sha256-file /path/to/miner-tls-cert-sha256 \
--cpu-workers 6 \
--gpu-devices 1 \
--metrics-port 9900
Protocol
The miner uses a QUIC-based protocol for communication with the node:
- Transport: QUIC with TLS 1.3 (self-signed certificate, pinned by SHA-256)
- Auth:
Ready { token }must match the node'sminer-auth-token - ALPN:
quantus-miner/2 - Port: 9833 (default)
- Messages:
Ready(miner→node),NewJob(node→miner),JobResult(miner→node)
For full protocol specification, see the node's MINING.md.
Benchmarking
# Benchmark CPU performance
./target/release/quantus-miner benchmark --cpu-workers 8 --duration 30
# Benchmark GPU performance
./target/release/quantus-miner benchmark --gpu-devices 1 --duration 30
# Benchmark hybrid performance
./target/release/quantus-miner benchmark --cpu-workers 4 --gpu-devices 1 --duration 30
Releases
2
quantus-miner v4.0.2-lair.2
Latest
Languages
Rust
77.9%
WGSL
13%
Cuda
7%
Shell
2%