Cryptis Miner

Cryptis Miner is a performance-oriented miner designed for real multi-device rigs.
It focuses on broad hardware compatibility, stable operation, and practical tuning for production mining setups.

Website: https://cryptis-miner.cryptix-network.org/
Public Benchmarks: https://cryptis-miner.cryptix-network.org/bench
Benchmark API: https://cryptis-miner.cryptix-network.org/api/v1/overview

Documentation Guide

Installation/platform setup: INSTALL.md
Batch starter scripts (Windows): batch/README.md
API reference: API.md
Known issues: BUGS.md

Currently Supported Targets

ox8 1% DevFee | CPU + AMD + NVIDIA + INTEL (CUDA + OPENCL)
hoohash 2% DevFee | CPU + AMD + NVIDIA + INTEL (CUDA + OPENCL)
randomx 1% DevFee | CPU
autolykosv2 1% DevFee | CPU + AMD + NVIDIA + INTEL (CUDA + OPENCL)

CPU Mining (1 CPU Hash):

ox8
hoohash
randomX

Hybrid Mining (1 CPU Hash + 1 GPU Hash):

ox8 + ox8
ox8 + hoohash
randomX + ox8
randomX + hoohash
hoohash + hoohash
hoohash + ox8

GPU Mining (1 GPU Hash):

ox8
hoohash
autolykosv2

Multi GPU Mining (1 Core GPU Hash + 1 Memory GPU Hash parallel on the GPU):

ox8 + autolykosv2
hoohash + autolykosv2

Tripple Mining ( 1 CPU Hash + 1 Core GPU Hash + 1 Memory GPU Hash)

ox8 + ox8 + autolykosv2
ox8 + hoohash + autolykosv2
randomX + ox8 + autolykosv2
randomX + hoohash + autolykosv2
hoohash + hoohash + autolykosv2
hoohash + ox8 + autolykosv2

What It Supports

CPU-only mining
GPU-only mining
CPU + GPU hybrid mining
Dual GPU lane mining on the same GPU set (core lane + memory lane)
Triple mining (CPU + GPU core lane + GPU memory lane)
Hybrid target overrides (cpu-coin/cpu-hash and gpu-coin/gpu-hash)
Multi-CPU support
Multi-GPU support
Mixed GPU rigs (AMD + NVIDIA + Intel in one rig)

Supported Platforms

Operating systems:

Linux
Windows
HiveOS

Architectures:

x64
x86
ARM

macOS / Apple is not supported in official releases.

Browser Dashboard

Open http://127.0.0.1:8943/ in your browser to access the embedded dashboard.

Quick Start

Windows (CLI only):

cryptis-miner.exe --coin <COIN> --hash <HASH> --pool stratum+tcp://pool.example.com:3333 --wallet <WALLET>

Linux / HiveOS (CLI only):

./cryptis-miner --coin <COIN> --hash <HASH> --pool stratum+tcp://pool.example.com:3333 --wallet <WALLET>

Windows (config file):

cryptis-miner.exe --config configs/default.toml mine

Linux / HiveOS (config file):

./cryptis-miner --config configs/default.toml mine

Check available targets on your binary:

cryptis-miner list-coins
cryptis-miner list-algorithms

Start Examples (Modes + Hash Targets)

Replace placeholders first:

<POOL_URL> example: stratum+tcp://pool.example.com:3333
<CPU_POOL_URL> / <GPU_POOL_URL> for mixed-target hybrid mode
<WALLET> = your wallet/login for the selected pool
<CPU_WALLET> / <GPU_WALLET> for mixed-target hybrid mode
<WORKER> = rig label

CPU-only:

cryptis-miner.exe --pool <POOL_URL> --wallet <WALLET> --worker <WORKER> --password x --coin cryptix --hash ox8 --no-gpu
cryptis-miner.exe --pool <POOL_URL> --wallet <WALLET> --worker <WORKER> --password x --coin unknown --hash ox8 --no-gpu
cryptis-miner.exe --pool <POOL_URL> --wallet <WALLET> --worker <WORKER> --password x --coin hoosat --hash hoohash --no-gpu
cryptis-miner.exe --pool <POOL_URL> --wallet <WALLET> --worker <WORKER> --password x --coin pepepow --hash hoohash --no-gpu
cryptis-miner.exe --pool <POOL_URL> --wallet <WALLET> --worker <WORKER> --password x --coin unknown --hash hoohash --no-gpu
cryptis-miner.exe --pool <POOL_URL> --wallet <WALLET> --worker <WORKER> --password x --coin monero --hash randomx --no-gpu
cryptis-miner.exe --pool <POOL_URL> --wallet <WALLET> --worker <WORKER> --password x --coin zephyr --hash randomx --no-gpu
cryptis-miner.exe --pool <POOL_URL> --wallet <WALLET> --worker <WORKER> --password x --coin unknown --hash randomx --no-gpu
cryptis-miner.exe --pool <POOL_URL> --wallet <WALLET> --worker <WORKER> --password x --coin ergo --hash autolykosv2 --no-gpu
cryptis-miner.exe --pool <POOL_URL> --wallet <WALLET> --worker <WORKER> --password x --coin unknown --hash autolykosv2 --no-gpu

GPU-only:

cryptis-miner.exe --pool <POOL_URL> --wallet <WALLET> --worker <WORKER> --password x --coin cryptix --hash ox8 --no-cpu
cryptis-miner.exe --pool <POOL_URL> --wallet <WALLET> --worker <WORKER> --password x --coin unknown --hash ox8 --no-cpu
cryptis-miner.exe --pool <POOL_URL> --wallet <WALLET> --worker <WORKER> --password x --coin hoosat --hash hoohash --no-cpu
cryptis-miner.exe --pool <POOL_URL> --wallet <WALLET> --worker <WORKER> --password x --coin pepepow --hash hoohash --no-cpu
cryptis-miner.exe --pool <POOL_URL> --wallet <WALLET> --worker <WORKER> --password x --coin unknown --hash hoohash --no-cpu
cryptis-miner.exe --pool <POOL_URL> --wallet <WALLET> --worker <WORKER> --password x --coin ergo --hash autolykosv2 --no-cpu
cryptis-miner.exe --pool <POOL_URL> --wallet <WALLET> --worker <WORKER> --password x --coin unknown --hash autolykosv2 --no-cpu

CPU+GPU hybrid:

cryptis-miner.exe --pool <POOL_URL> --wallet <WALLET> --worker <WORKER> --password x --coin cryptix --hash ox8
cryptis-miner.exe --pool <CPU_POOL_URL> --wallet <CPU_WALLET> --worker <WORKER> --password x --coin monero --hash randomx --cpu-pool <CPU_POOL_URL> --cpu-wallet <CPU_WALLET> --gpu-pool <GPU_POOL_URL> --gpu-wallet <GPU_WALLET> --gpu-password x --gpu-coin unknown --gpu-hash ox8
cryptis-miner.exe --pool <CPU_POOL_URL> --wallet <CPU_WALLET> --worker <WORKER> --password x --coin cryptix --hash ox8 --cpu-pool <CPU_POOL_URL> --cpu-wallet <CPU_WALLET> --gpu-pool <GPU_POOL_URL> --gpu-wallet <GPU_WALLET> --gpu-password x --gpu-coin ergo --gpu-hash autolykosv2
cryptis-miner.exe --pool <CPU_POOL_URL> --wallet <CPU_WALLET> --worker <WORKER> --password x --coin hoosat --hash hoohash --cpu-pool <CPU_POOL_URL> --cpu-wallet <CPU_WALLET> --gpu-pool <GPU_POOL_URL> --gpu-wallet <GPU_WALLET> --gpu-password x --gpu-coin unknown --gpu-hash ox8
cryptis-miner.exe --pool <POOL_URL> --wallet <WALLET> --worker <WORKER> --password x --coin unknown --hash hoohash --gpu-coin unknown --gpu-hash hoohash

Dual GPU lanes (same GPUs, OpenCL): core + memory

cryptis-miner.exe --pool <GPU_CORE_POOL_URL> --wallet <GPU_CORE_WALLET> --worker <WORKER> --password x --coin unknown --hash ox8 --no-cpu --gpu-backend opencl --gpu-core-coin unknown --gpu-core-hash ox8 --gpu-memory-coin unknown --gpu-memory-hash autolykosv2 --gpu-core-pool <GPU_CORE_POOL_URL> --gpu-core-wallet <GPU_CORE_WALLET> --gpu-core-password x --gpu-memory-pool <GPU_MEMORY_POOL_URL> --gpu-memory-wallet <GPU_MEMORY_WALLET> --gpu-memory-password x --gpu-core-intensity 0.55 --gpu-memory-intensity 0.45
cryptis-miner.exe --pool <GPU_CORE_POOL_URL> --wallet <GPU_CORE_WALLET> --worker <WORKER> --password x --coin unknown --hash hoohash --no-cpu --gpu-backend opencl --gpu-core-coin unknown --gpu-core-hash hoohash --gpu-memory-coin unknown --gpu-memory-hash autolykosv2 --gpu-core-pool <GPU_CORE_POOL_URL> --gpu-core-wallet <GPU_CORE_WALLET> --gpu-core-password x --gpu-memory-pool <GPU_MEMORY_POOL_URL> --gpu-memory-wallet <GPU_MEMORY_WALLET> --gpu-memory-password x --gpu-core-intensity 0.55 --gpu-memory-intensity 0.45

Triple mode (CPU + GPU core + GPU memory):

cryptis-miner.exe --pool <CPU_POOL_URL> --wallet <CPU_WALLET> --worker <WORKER> --password x --coin monero --hash randomx --cpu-pool <CPU_POOL_URL> --cpu-wallet <CPU_WALLET> --gpu-backend opencl --gpu-core-coin unknown --gpu-core-hash ox8 --gpu-memory-coin unknown --gpu-memory-hash autolykosv2 --gpu-core-pool <GPU_CORE_POOL_URL> --gpu-core-wallet <GPU_CORE_WALLET> --gpu-core-password x --gpu-memory-pool <GPU_MEMORY_POOL_URL> --gpu-memory-wallet <GPU_MEMORY_WALLET> --gpu-memory-password x --gpu-core-intensity 0.55 --gpu-memory-intensity 0.45
cryptis-miner.exe --pool <CPU_POOL_URL> --wallet <CPU_WALLET> --worker <WORKER> --password x --coin monero --hash randomx --cpu-pool <CPU_POOL_URL> --cpu-wallet <CPU_WALLET> --gpu-backend opencl --gpu-core-coin unknown --gpu-core-hash hoohash --gpu-memory-coin unknown --gpu-memory-hash autolykosv2 --gpu-core-pool <GPU_CORE_POOL_URL> --gpu-core-wallet <GPU_CORE_WALLET> --gpu-core-password x --gpu-memory-pool <GPU_MEMORY_POOL_URL> --gpu-memory-wallet <GPU_MEMORY_WALLET> --gpu-memory-password x --gpu-core-intensity 0.55 --gpu-memory-intensity 0.45

Windows starter scripts are in batch/:

GPU script examples:
- batch/start-gpu-cryptix-ox8.bat
- batch/start-gpu-hoosat-hoohash.bat
- batch/start-gpu-pepepow-hoohash.bat
- batch/start-gpu-unknown-hoohash.bat
- batch/start-gpu-ergo-autolykosv2.bat
- batch/start-gpu-unknown-autolykosv2.bat
- batch/start-gpu-unknown-ox8.bat

Hybrid naming format:

start-hybrid-<cpu-coin>-<cpu-hash>__<gpu-coin>-<gpu-hash>.bat

Dual naming format:

start-dual-<gpu-core-coin>-<gpu-core-hash>__<gpu-memory-coin>-<gpu-memory-hash>.bat

Triple naming format:

start-triple-<cpu-coin>-<cpu-hash>__<gpu-core-coin>-<gpu-core-hash>__<gpu-memory-coin>-<gpu-memory-hash>.bat

Examples:

start-hybrid-cryptix-ox8__ergo-autolykosv2.bat
start-hybrid-monero-randomx__unknown-ox8.bat
start-hybrid-hoosat-hoohash__hoosat-hoohash.bat
start-hybrid-unknown-hoohash__unknown-hoohash.bat
start-hybrid-cryptix-ox8__unknown-ox8.bat
start-dual-unknown-ox8__unknown-autolykosv2.bat
start-triple-monero-randomx__unknown-ox8__unknown-autolykosv2.bat

Hybrid script note:

Same-target hybrid (ox8+ox8) can use one pool setup.
Mixed-target hybrid (for example CPU randomx + GPU ox8 or CPU ox8 + GPU autolykosv2) requires separate endpoint sets for CPU and GPU, including failovers.
Use --cpu-pool/--gpu-pool and optional --cpu-failover-pools/--gpu-failover-pools.

Runtime Notes

CLI options override values from --config.
GPU routing works for mixed rigs via --gpu-devices, --cuda-devices, and --opencl-devices.
Inspect detected devices with cryptis-miner device-inventory.
In auto mode, backend selection follows runtime availability:
- ox8: CUDA/OpenCL based on runtime/device routing.
- autolykosv2: NVIDIA prefers CUDA only when CUDA hash path is active (--cuda-experimental), otherwise OpenCL fallback is used; AMD/Intel use OpenCL.
OpenCL hashing is available and recommended for production hashrate right now.
CUDA hashing can be enabled for testing with --cuda-experimental (usually together with --cuda).
CUDA experimental mode is currently for validation/testing only; CUDA kernel acceleration/tuning is not finished yet.
RandomX accepts --stratum-protocol v2; this uses the RandomX compatibility login/job/submit workflow for pools that expose RandomX over a v2 profile.
Default CUDA experimental state is controlled by CUDA_HASHING_EXPERIMENTAL_ENABLED in src/mining/gpu/cuda.rs; --cuda-experimental overrides it at runtime.
Mixed rigs can run OpenCL + CUDA together when cuda_devices and opencl_devices are set to disjoint GPU id lists.
In hybrid mode with different targets, overlapping CPU/GPU pool endpoint sets are rejected at startup (must be separated by target).
Dual/triple lane mode is available via --gpu-core-* + --gpu-memory-* and currently runs in OpenCL lane mode (both lanes are forced to OpenCL in this release).
Current dual/triple hash pairs for production are gpu-core=ox8 + gpu-memory=autolykosv2 and gpu-core=hoohash + gpu-memory=autolykosv2 (the hoohash pair is OpenCL-only in this release; other pairs keep GPU core active and disable memory lane).
GPU launch autotune is globally serialized across lanes/process workers in one miner instance, so dual/triple startup tuning runs one-at-a-time instead of competing on the same GPUs.
In dual/triple mode with different targets, overlapping pool endpoint sets across active lanes are rejected at startup (CPU, GPU core, GPU memory).
Hybrid CPU core reservation is configurable in [mining.runtime]:
- hybrid_cpu_reserve_min_cores
- hybrid_cpu_reserve_max_cores
- hybrid_cpu_reserve_gpu_threshold
CPU-only system-core reservation is configurable in [mining.runtime]:
- cpu_only_reserve_system_core (default: true)
- cpu_only_reserved_cores (default: 1)
Pool reconnect budgets:
- pool.retry_count (default: 50)
- pool.failover_retry_count (default: 0 = unlimited)
Per-hash GPU share policy is configurable in [mining.runtime.gpu_hash_policies.<hash>]:
- cpu_verify
- opencl_mad_enable
- opencl_native_math_enable (experimental)
- opencl_fp_contract_disable (hoohash-only)
- opencl_accuracy_boost (experimental, hoohash-only, requires cpu_verify=true)
- cuda_strict_math_enable (hoohash-only)
- strict_kernel_verify (hoohash-only)
- strict_job
- recent_job_max_ids
- recent_job_max_age_ms
- Default policy:
  - ox8: cpu_verify=false, strict_job=false, window 64 / 8000ms
  - hoohash: cpu_verify=true, opencl_fp_contract_disable=true, cuda_strict_math_enable=true, strict_kernel_verify=false, strict_job=false, window 640 / 1000ms
  - autolykosv2: cpu_verify=false, strict_job=true
HooHash OpenCL tuning:
- opencl_native_math_enable=false by default; can increase speed but may reduce CPU-verified share accuracy.
- opencl_fp_contract_disable=true by default; reduces FMA/contraction variance between OpenCL compilers.
- opencl_accuracy_boost=false by default; can improve accepted-share quality near precision boundaries at the cost of extra CPU verify work.
HooHash CUDA tuning:
- cuda_strict_math_enable=true by default; uses stricter FP compile flags to reduce GPU/CPU drift.
- strict_kernel_verify=false by default; enabling it enforces strict GPU-vs-CPU startup parity checks and may reject unstable device/driver combos.
HiveOS wrapper keys for the same behavior:
- HYBRID_CPU_RESERVE_MIN_CORES
- HYBRID_CPU_RESERVE_MAX_CORES
- HYBRID_CPU_RESERVE_GPU_THRESHOLD
- CPU_ONLY_RESERVE_SYSTEM_CORE
- CPU_ONLY_RESERVED_CORES
- GPU_CPU_VERIFY (for example ox8=off,hoohash=on,autolykosv2=off)
- GPU_OPENCL_MAD_ENABLE (for example hoohash=off)
- GPU_OPENCL_NATIVE_MATH_ENABLE (for example hoohash=off)
- GPU_OPENCL_ACCURACY_BOOST (for example hoohash=on)
- GPU_STRICT_JOB (for example ox8=off,hoohash=off,autolykosv2=on)
- GPU_RECENT_JOB_MAX_IDS (for example hoohash=640)
- GPU_RECENT_JOB_MAX_AGE_MS (for example hoohash=1000)
- Pool/user input can come from the classic CUSTOM_URL CSV, dedicated HiveOS pool/template fields, or CUSTOM_USER_CONFIG CSV keys such as POOL, TEMPLATE, WALLET, WORKER, USER, and PASS
- Common HiveOS macros are resolved in wrapper inputs: %URL%, %URL_HOST%, %URL_PORT%, %WAL%, %WALLET%, %WORKER_NAME%
- Literal null/none in Wallet and worker template is treated as empty so the official HiveOS POOL:...,WALLET:%WAL%,WORKER:%WORKER_NAME% format works
- Raw CLI passthrough from CUSTOM_USER_CONFIG only happens for lines that start with -/--; CSV KEY:VALUE entries are parsed by the wrapper and not appended as unknown miner arguments
Optional benchmark telemetry uploads only performance/tuning metadata (hashrate, efficiency, temperatures, clocks, batch/autotune/backend/OC settings). No wallet/private-key secrets are sent.
--coin unknown disables coin-specific wallet validation while benchmark telemetry/insights remain available
CPU mining supports ox8, hoohash, randomx, and autolykosv2.
--gpu-hash autolykosv2: NVIDIA runs via CUDA, AMD/Intel runs via OpenCL
Autolykos block tuning can be pinned with --autolykos-block-size <N> or mining.runtime.autolykos_block_size (>=64, divisible by 8).

Frontend Runtime Notes

Embedded dashboard is enabled by default.
Default URL: http://127.0.0.1:8943/
Miner log pane uses logging.file and shows the latest 50 lines with a 5s refresh.
Frontend log endpoint: GET /api/logs (on frontend port)
Main options:
- --frontend-disabled
- --frontend-logs-enabled
- --frontend-logs-disabled
- --frontend-bind
- --frontend-port
- --frontend-password-enabled
- --frontend-password
- --frontend-rate-limit-per-minute
Config toggle: frontend.logs_enabled = true|false (default: true)
HiveOS wrapper toggle: FRONTEND_LOGS_DISABLED:1 (in CUSTOM_URL) or CUSTOM_FRONTEND_LOGS_DISABLED=1

CLI Reference (All Startup Arguments)

Usage:

cryptis-miner [OPTIONS] [COMMAND]

Always check your local binary help for the latest live output:

cryptis-miner --help
cryptis-miner <COMMAND> --help

Commands

mine - start mining (default when command is omitted)
benchmark - run benchmark mode
- --duration <SECONDS> - benchmark duration
- --coin <COIN> - benchmark coin target
- --hash <HASH> - benchmark hash target
- --algorithm <NAME> - hidden legacy selector
list-algorithms - print supported algorithms
list-coins - print supported coins
compatibility - print runtime compatibility matrix
device-inventory - print detected GPUs and CUDA/OpenCL visibility
config generate --output <FILE> - generate default config
config validate --file <FILE> - validate a config file
config show - print current effective config as JSON

Target and Pool

--coin <COIN> - coin target
--hash <HASH> - hash family target
--cpu-coin <COIN> - CPU target coin override
--cpu-hash <HASH> - CPU target hash override
--gpu-coin <COIN> - GPU target coin override
--gpu-hash <HASH> - GPU target hash override (ox8, hoohash, or autolykosv2)
--gpu-core-coin <COIN> - GPU core-lane coin override (dual/triple)
--gpu-core-hash <HASH> - GPU core-lane hash override (dual/triple supports ox8 and hoohash; hoohash core pair currently requires OpenCL)
--gpu-memory-coin <COIN> - GPU memory-lane coin override (dual/triple)
--gpu-memory-hash <HASH> - GPU memory-lane hash override (dual/triple currently expects autolykosv2)
-a, --algorithm <NAME> - hidden legacy selector (compatibility only)
-p, --pool <URL> - pool URL (stratum+tcp://... or stratum+ssl://...)
--cpu-pool <URL> - CPU pool URL override
--cpu-failover-pools <URLS> - CPU failover pools (comma-separated)
--cpu-stratum-protocol <v1|v2> - CPU Stratum protocol override
--cpu-user <USER> - CPU pool login override
--cpu-password <PASS> - CPU pool password override
--cpu-wallet <WALLET> - CPU wallet override
--gpu-pool <URL> - GPU pool URL override
--gpu-failover-pools <URLS> - GPU failover pools (comma-separated)
--gpu-stratum-protocol <v1|v2> - GPU Stratum protocol override
--gpu-user <USER> - GPU pool login override
--gpu-password <PASS> - GPU pool password override
--gpu-wallet <WALLET> - GPU wallet override
--gpu-core-pool <URL> - GPU core-lane pool URL override
--gpu-core-failover-pools <URLS> - GPU core-lane failover pools (comma-separated)
--gpu-core-stratum-protocol <v1|v2> - GPU core-lane Stratum protocol override
--gpu-core-user <USER> - GPU core-lane pool login override
--gpu-core-password <PASS> - GPU core-lane pool password override
--gpu-core-wallet <WALLET> - GPU core-lane wallet override
--gpu-memory-pool <URL> - GPU memory-lane pool URL override
--gpu-memory-failover-pools <URLS> - GPU memory-lane failover pools (comma-separated)
--gpu-memory-stratum-protocol <v1|v2> - GPU memory-lane Stratum protocol override
--gpu-memory-user <USER> - GPU memory-lane pool login override
--gpu-memory-password <PASS> - GPU memory-lane pool password override
--gpu-memory-wallet <WALLET> - GPU memory-lane wallet override
--stratum-protocol <v1|v2> - Stratum protocol version
--stratum-protocol-fallback / --no-stratum-protocol-fallback - try v1<->v2 fallback on protocol connect/authorize failure
--stratum-transport <auto|tcp|tls> - transport mode
--stratum-transport-fallback / --no-stratum-transport-fallback - try TCP<->TLS fallback when connect/handshake fails
--pool-retry-count <N> - reconnect attempts (0 = unlimited)
--pool-retry-delay-ms <MS> - reconnect base delay
--pool-connect-timeout-ms <MS> - TCP connect timeout
--pool-tls-timeout-ms <MS> - TLS handshake timeout
--pool-tls-verify-cert / --pool-tls-no-verify-cert - enable or disable TLS certificate+hostname verification
--pool-request-timeout-ms <MS> - Stratum request timeout
--pool-job-channel-size <N> - internal pool job queue size

Identity and Auth

-u, --user <USER> - pool login user
-P, --password <PASS> - pool password
-w, --wallet <WALLET> - wallet address
-n, --worker <NAME> - worker name
--rig-label <NAME> - alias for worker label

CPU and GPU Routing

-t, --threads <N> - CPU thread count
--cpu-only-reserve-system-core / --no-cpu-only-reserve-system-core - toggle CPU-only system-core reservation
--cpu-only-reserved-cores <N> - reserved logical core count in CPU-only mode
--cpu-affinity <IDS> - bind CPU workers to core IDs (comma-separated)
--no-cpu - disable CPU mining
--no-gpu - disable GPU mining
--gpu-devices <IDS> - selected GPU IDs (comma-separated)
--cuda-devices <IDS> - force selected GPU IDs to CUDA
--opencl-devices <IDS> - force selected GPU IDs to OpenCL
--disable-gpu-amd - skip AMD GPUs
--disable-gpu-nvidia - skip NVIDIA GPUs
--disable-gpu-intel - skip Intel GPUs

Intensity and Batch Sizing

--intensity <X> - global intensity fallback (CPU + GPU)
--intensity-min <X> - lower clamp for effective intensity
--intensity-max <X> - upper clamp for effective intensity
--cpu-intensity <X> - CPU intensity override
--gpu-intensity <X> - GPU intensity override
--gpu-core-intensity <X> - GPU core-lane intensity override (dual/triple)
--gpu-memory-intensity <X> - GPU memory-lane intensity override (dual/triple)
--cpu-batch-base <N> - CPU batch baseline
--cpu-batch-min <N> - CPU batch minimum
--cpu-batch-max <N> - CPU batch maximum
--cpu-batch-size <N> - fixed CPU batch override
--gpu-batch-base <N> - GPU batch baseline
--gpu-batch-min <N> - GPU batch minimum
--gpu-batch-max <N> - GPU batch maximum
--opencl-batch-size <N> - fixed OpenCL batch override
--opencl-local-work-size <N> - fixed OpenCL local work size
--autolykos-block-size <N> - Autolykos-only GPU block size override (OpenCL + CUDA)
--cuda-batch-size <N> - fixed CUDA batch override
--cuda-block-size <N> - fixed CUDA block size (threads per block)

Autotune

--opencl-autotune / --no-opencl-autotune - enable or disable OpenCL startup autotune
--cuda-autotune / --no-cuda-autotune - enable or disable CUDA startup autotune
--cpu-autotune / --no-cpu-autotune - enable or disable CPU startup autotune
--cpu-autotune-probe-ms <MS> - CPU autotune probe duration
--gpu-autotune-rounds <N> - rounds per GPU candidate (median scoring)

Autotune behavior:

CPU autotune ON: startup selects CPU batch size automatically.
CPU autotune OFF: cpu_batch_size is used if set; otherwise defaults are computed.
OpenCL autotune ON: startup selects OpenCL launch values automatically.
OpenCL autotune OFF: manual OpenCL launch values are used if set.
CUDA autotune ON: startup selects CUDA block size and batch size automatically.
CUDA autotune OFF: manual CUDA launch values are used if set.

Runtime and Queue Control

--share-submit-rate <N> - max submit rate (shares per second)
--share-submit-burst <N> - submit burst budget
--share-queue-capacity <N> - submit queue capacity
--recent-job-max-ids <N> - recent-job cache size
--recent-job-max-age-ms <MS> - recent-job cache max age
--gpu-cpu-verify <HASH=BOOL[,HASH=BOOL...]> - per-hash GPU CPU-verify toggle (ox8|hoohash|autolykosv2)
--gpu-opencl-mad-enable <HASH=BOOL[,HASH=BOOL...]> - per-hash OpenCL MAD toggle (ox8|hoohash|autolykosv2)
--gpu-opencl-native-math-enable <HASH=BOOL[,HASH=BOOL...]> - per-hash experimental native OpenCL math toggle (ox8|hoohash|autolykosv2)
--gpu-opencl-fp-contract-disable <HASH=BOOL[,HASH=BOOL...]> - hoohash-only OpenCL FP contraction control (hoohash)
--gpu-opencl-accuracy-boost <HASH=BOOL[,HASH=BOOL...]> - per-hash experimental OpenCL accuracy boost (hoohash only, requires CPU verify)
--gpu-cuda-strict-math-enable <HASH=BOOL[,HASH=BOOL...]> - hoohash-only CUDA strict FP compile toggle (hoohash)
--gpu-strict-kernel-verify <HASH=BOOL[,HASH=BOOL...]> - hoohash-only strict GPU kernel parity verification toggle (hoohash)
--gpu-strict-job <HASH=BOOL[,HASH=BOOL...]> - per-hash strict-job toggle (on = only latest job)
--gpu-recent-job-max-ids <HASH=N[,HASH=N...]> - per-hash recent-job id window (used when strict-job is off)
--gpu-recent-job-max-age-ms <HASH=MS[,HASH=MS...]> - per-hash recent-job age window in milliseconds (used when strict-job is off)
--job-recv-timeout-ms <MS> - timeout for incoming jobs
--stats-interval-ms <MS> - stats update interval
--gpu-status-board-interval-ms <MS> - GPU status board interval (0 disables)
--hybrid-cpu-reserve-min-cores <N> - hybrid reserved CPU cores when GPU count is at/below threshold
--hybrid-cpu-reserve-max-cores <N> - hybrid reserved CPU cores when GPU count is above threshold
--hybrid-cpu-reserve-gpu-threshold <N> - GPU-count threshold used by hybrid reserve policy
--task-drain-timeout-ms <MS> - graceful shutdown drain timeout
--shutdown-poll-ms <MS> - shutdown poll interval
--reconnect-min-delay-ms <MS> - reconnect delay floor
--reconnect-backoff-max-power <N> - reconnect backoff exponent cap
--worker-idle-sleep-ms <MS> - worker sleep while paused
--worker-recv-timeout-ms <MS> - worker receive timeout
--worker-max-slice-ms <MS> - max worker slice before yielding
--worker-slice-check-interval <N> - loop interval for slice checks
--worker-active-poll-interval <N> - loop interval for active-state checks
--worker-stats-flush-threshold <N> - hash threshold for stats flush
--worker-stats-flush-interval-ms <MS> - periodic stats flush interval

Rule examples:

--gpu-cpu-verify ox8=on,hoohash=on,autolykosv2=off
--gpu-opencl-mad-enable hoohash=off
--gpu-opencl-native-math-enable hoohash=off
--gpu-opencl-fp-contract-disable hoohash=on
--gpu-opencl-accuracy-boost hoohash=on
--gpu-cuda-strict-math-enable hoohash=on
--gpu-strict-kernel-verify hoohash=on
--gpu-strict-job ox8=off,hoohash=off,autolykosv2=on
--gpu-recent-job-max-ids hoohash=640
--gpu-recent-job-max-age-ms hoohash=1000

Backend Selection

--gpu-backend <auto|cuda|opencl> - preferred GPU backend
--cuda / --no-cuda - enable CUDA preference or disable CUDA
--cuda-experimental - enable experimental CUDA hash path at runtime (testing only; not performance-optimized yet)
--opencl / --no-opencl - enable OpenCL preference or disable OpenCL

OC Command Runner

--oc / --no-oc - enable or disable OC command phase
--oc-on-start / --no-oc-on-start - apply OC commands on startup
--oc-fail-on-error / --oc-ignore-errors - abort or continue on OC errors
--oc-dry-run / --no-oc-dry-run - print only or execute commands
--oc-timeout-ms <MS> - timeout per OC command
--oc-cmd "<CMD>" - generic OC command template (repeatable)
--oc-nvidia-cmd "<CMD>" - NVIDIA OC command template (repeatable)
--oc-amd-cmd "<CMD>" - AMD OC command template (repeatable)
--oc-intel-cmd "<CMD>" - Intel OC command template (repeatable)

Supported OC placeholders:

{id} - internal runtime device index
{device_id} - physical device index reported by backend
{vendor} - normalized vendor key (amd, nvidia, intel, ...)
{vendor_name} - full vendor string
{backend} - active backend (cuda or opencl)

Startup Extras

--window-title <TITLE> - set console window title
--startup-cmd "<CMD>" - run one startup command before mining
--startup-timeout-ms <MS> - timeout for startup command
--gpu-sensors / --no-gpu-sensors - enable or disable GPU sensor reads
--randomx-hugepages <auto|on|off> - RandomX huge pages mode (for all RandomX coins, e.g. Monero/Zephyr)
--randomx-msr <auto|on|off> - RandomX MSR tuning mode (best-effort; admin/root needed)

Algorithm and CPU Feature Switches

--ox8-avx2 / --no-ox8-avx2 - force AVX2 backend on or off (if supported)

Config, Logging, and Output

-c, --config <FILE> - config file path
-v, --verbose - increase log verbosity (repeatable)
--log-file <FILE> - write logs to file
--no-color - disable colored output

REST API

Default bind: 127.0.0.1:48673
Hive-compatible stats endpoint uses the same port (/hiveos/stats by default).
--api-live / --api-disabled - enable or disable REST API
--api-bind <ADDR> - API bind address
--api-port <PORT> - API port
--api-token <TOKEN> - require Bearer token for REST API endpoints
--hive-stats-disabled - disable Hive-compatible endpoint
--hive-stats-path <PATH> - custom Hive endpoint path

Frontend

--frontend-disabled - disable embedded frontend
--frontend-logs-enabled - force-enable frontend miner log pane
--frontend-logs-disabled - disable frontend miner log pane
--frontend-bind <ADDR> - frontend bind address
--frontend-port <PORT> - frontend port
--frontend-password-enabled - require frontend password
--frontend-password <PASS> - set frontend password
--frontend-rate-limit-per-minute <N> - request limit per IP

Benchmark Telemetry

--bench-report / --no-bench-report - enable or disable benchmark telemetry uploads
--bench-report-interval-sec <SECONDS> - upload interval
--bench-report-id-file <PATH> - reporter id persistence file
--bench-report-api-key <TOKEN> - optional X-Api-Key header for collector auth
--bench-insights / --no-bench-insights - enable or disable device-vs-network comparison insights

Notes:

Upload failures are ignored by design and never stop mining.
If id-file creation fails, a runtime-only reporter id is used automatically.
Collector URL is embedded in the miner binary and not exposed via config/CLI.
The miner accepts both secure and insecure transport fallback for this optional cloud feature.

Utility Flags

-h, --help - show help
-V, --version - show version

API Endpoints

When API is enabled:

GET /api/v1/health
GET /api/v1/stats
GET /api/v1/telemetry
GET /api/v1/system
GET /api/v1/devices
GET /api/v1/hive
GET /hiveos/stats (or custom hive_stats_path)

Full developer API reference:

API.md

Config File

Use the shipped config as your baseline:

configs/default.toml

Benchmark telemetry in config:

Section: [benchmark_telemetry]
Default: enabled
Intended use: optional benchmark/performance analytics only

Generate a fresh config from current defaults:

cryptis-miner config generate --output myconfig.toml

Validate a config:

cryptis-miner config validate --file myconfig.toml

Cryptis Miner does not allow mining on rplant.xyz (including subdomains/server IPs) due to ethical concerns regarding the operators. The software will terminate immediately upon startup; mining on the Rplant pool is not possible, but mining on all other pools is.

Support

Support in Cryptix Network Discord (Cryptis Miner Channel): https://discord.cryptix-network.org/

Name		Name	Last commit message	Last commit date
Latest commit History 36 Commits
configs		configs
hiveos/custom/cryptis-miner		hiveos/custom/cryptis-miner
mmpos/custom/cryptis-miner		mmpos/custom/cryptis-miner
tools		tools
.gitattributes		.gitattributes
API.md		API.md
BUGS.md		BUGS.md
INSTALL.md		INSTALL.md
README.md		README.md
TODO.md		TODO.md
cryptis-miner-1.png		cryptis-miner-1.png
cryptis-miner-2.png		cryptis-miner-2.png
cryptis-miner-3.png		cryptis-miner-3.png

Folders and files

Latest commit

History

Repository files navigation

Cryptis Miner

Documentation Guide

Table of Contents

Currently Supported Targets

What It Supports

Supported Platforms

Browser Dashboard

Quick Start

Start Examples (Modes + Hash Targets)

Runtime Notes

Frontend Runtime Notes

CLI Reference (All Startup Arguments)

Commands

Target and Pool

Identity and Auth

CPU and GPU Routing

Intensity and Batch Sizing

Autotune

Runtime and Queue Control

Backend Selection

OC Command Runner

Startup Extras

Algorithm and CPU Feature Switches

Config, Logging, and Output

REST API

Frontend

Benchmark Telemetry

Utility Flags

API Endpoints

Config File

Cryptis Miner does not allow mining on rplant.xyz (including subdomains/server IPs) due to ethical concerns regarding the operators. The software will terminate immediately upon startup; mining on the Rplant pool is not possible, but mining on all other pools is.

Support

About

Resources

Uh oh!

Stars

Watchers

Forks

Releases 1

Packages 0

Uh oh!

Contributors

Uh oh!

Languages

Packages