NOID-GPU v1.3.0

NOID-GPU 1.3.0 release notes

Release date: 27 August 2026

Summary

Version 1.3.0 adds the R4 GPU path and its new field representation. The
release focuses on higher hashrate, better measured energy efficiency on the
controlled RTX 5070 Ti comparison, and correctness across the share-target
range used by current pools.

The public miner developer fee is 5%. The separate Parano1d Pool fee is
3%.

Benchmark results

GPU / path	Gross wall rate	Board power	Efficiency
RTX 5090, R4	127.096 MH/s	492.977 W	0.257813 MH/s/W
RTX 4090, R4	88.951 MH/s	359.870 W	0.247175 MH/s/W
RTX 5070 Ti, previous path	41.315 MH/s	162.420 W	0.254371 MH/s/W
RTX 5070 Ti, R4	51.337 MH/s	188.733 W	0.272009 MH/s/W (+6.934%)

The RTX 5090 and RTX 4090 results come from three 300-second offline runs per
GPU. The RTX 5070 Ti result is a controlled comparison from 41.315 to 51.337
MH/s, a 24.258% increase in gross wall rate. Its board power changed from
162.420 to 188.733 W while measured MH/s/W improved by 6.934%.

All figures are gross offline measurements of the real GPU hashing path. They
are not pool-side hashrate, exclude network and fee-window effects, and do not
guarantee the same result on every card or rig.

Miner changes

• Added the new R4 field representation and optimized CUDA path.
• Tested the fast target path through 26 bits, including 24-bit share
  targets.
• Extended correctness checks around CPU/GPU digest agreement, target
  comparison, and nonce-search results.
• Kept CPU verification of GPU results before share submission.
• Preserved multi-GPU operation and device selection.
• Kept public help, status, errors, launchers, and documentation in English.

Packages

• noid-gpu-windows-1.3.0.zip
• noid-gpu-linux-1.3.0.tar.gz
• noid-gpu-hiveos-1.3.0.tar.gz
• SHA256SUMS.txt

The Windows and Linux packages support NVIDIA Ampere and newer GPUs: RTX 30,
RTX 40, RTX 50, and equivalent professional cards. RTX 20 and older NVIDIA
architectures are not supported. NVIDIA driver 580 or newer is required.

The HiveOS asset was built on Ubuntu 20.04/Focal with GLIBC 2.31. Its highest
required GLIBC symbol is GLIBC_2.30. The final archive passed extraction,
layout, and permission checks; its miner executable is byte-for-byte identical
to the Linux release executable and passed startup and self-test inside the
Focal build environment.

Pool selection

--pool accepts parano1d, ariapool, a compatible host:port, or an HTTP
URL. A custom endpoint can therefore point to a compatible local pool:

noid-gpu --gpu --pool 192.0.2.10:3784 --coinbase o1YOUR_PUBLIC_ADDRESS --worker rig1

The custom endpoint applies to the user's mining jobs. Scheduled developer-fee
jobs use the destinations embedded in the public executable.

Fees

• Public miner developer fee: 5% of scheduled mining time (30 seconds per
  600-second cycle, switching only between jobs).
• Parano1d Pool fee: 3%, separate from the miner developer fee.

Verification

Verify every asset against SHA256SUMS.txt before extraction or execution:

sha256sum -c SHA256SUMS.txt

On Windows, use Get-FileHash -Algorithm SHA256 and compare the value with
the matching checksum entry.

NOID-GPU 1.2.0

Lower fees. The pool fee is now 3% and the miner developer fee is 5% — down from 7.5% and 7.5%. Nothing else about how they work has changed: the dev fee is still a scheduled window, now 30 seconds out of every 600, announced in the header on every run with the address in full, and it still never touches the coinbase of a block.

you have	take
Windows	noid-gpu-windows-1.2.0.zip
Linux	noid-gpu-linux-1.2.0.tar.gz
HiveOS	noid-gpu-hiveos-1.2.0.tar.gz — paste its URL as the custom miner Installation URL

noid-gpu --gpu --coinbase o1YOURADDRESS --worker rig1

Also in this release:

• the --help text was quoting a pool fee that had already changed. It now reads the real one, and a test checks the link rather than repeating the number, so it cannot drift again;
• the miner points at https://parano1d-pool.fun for your statistics instead of a bare IP.

Same as before: NVIDIA driver 580 or newer, Ampere or later (RTX 30xx/40xx/50xx, A-series, H100). The Linux and HiveOS builds need no GLIBC newer than 2.30, so they run on HiveOS Focal, jammy and noble alike.

Pool dashboard: https://parano1d-pool.fun

Verify your download against SHA256SUMS.txt.

NOID-GPU 1.1.0

NOID-GPU 1.1.0 release notes

Release date: 24 August 2026

Highlights

• 18.99% higher measured wall-rate on RTX 5070 Ti in a controlled B-A-B
  comparison.
• 70.492 MH/s at 280.14 W on RTX 4090.
• 104.526 MH/s at 453.69 W on RTX 5090.
• RTX 4090 median board power remained 169.86 W below its 450 W limit.
• RTX 5090 median board power remained 46.31 W below its 500 W limit.
• Public miner developer fee reduced from 10% to 7.5%.
• Parano1d Pool fee changed from 10% to 7.5%.
• Compact English startup panel and English user-visible errors.
• Windows, Linux, and HiveOS packages; Linux/HiveOS remain compatible with
  the Ubuntu 20.04 GLIBC 2.31 baseline.

Measured performance and power

Stabilized RTX 4090 and RTX 5090 runs

GPU	NOID-GPU wall rate	Board power	MH/s/W	Configured power limit	Unused headroom
RTX 4090	70.492 MH/s	280.14 W	0.25163	450 W	169.86 W (37.75%)
RTX 5090	104.526 MH/s	453.69 W	0.23039	500 W	46.31 W (9.26%)

The measured hashrate therefore did not require either card to sit at its
configured maximum board power throughout the stabilized run.

The normal 16,777,216-nonce batch produced:

GPU	Median kernel rate	Median wall rate	Short live dry-run active rate
RTX 4090	70.433 MH/s	70.314 MH/s	70.23 MH/s
RTX 5090	104.710 MH/s	104.298 MH/s	103.78 MH/s

Power data came from the longer 268,435,456-nonce stabilization run, using
two discarded warm-ups, five measured samples, and NVIDIA telemetry sampled
every 100 ms. The normal batch used five warm-ups and fifteen samples.

These were separate dedicated cards. Results are measured evidence from one
card of each model, not a promise that every rig will reproduce the same
watts.

RTX 5070 Ti controlled speed comparison

Build	Median kernel rate	Median wall rate
Previous frozen kernel	34.679 MH/s	34.566 MH/s
NOID-GPU 1.1.0	41.290 MH/s	41.130 MH/s
Improvement	+19.06%	+18.99%

Power was not resampled during this controlled B-A-B. Accordingly, this
release claims a measured speed increase on RTX 5070 Ti, but does not
claim a controlled watt reduction versus the previous kernel.

A separate Windows field run with three RTX 5070 Ti cards produced about
120 MH/s combined while instantaneous NVIDIA-SMI readings were roughly
132-180 W per card, despite configured 300 W limits. This shows that the
observed hashrate did not require the cards to reach the power cap. It is a
field observation rather than a controlled before/after power benchmark.

CUDA optimization

• Reduced the full Poseidon2b path from 502 to 448 GF(2^128)
  multiplications per permutation, a 10.76% reduction in arithmetic
  work.
• Added a lazy most-significant-bit-first comparison directly from the flat
  GPU state, avoiding full digest conversion on the normal reject path.
• Replaced the hot MDS calculation with algebraically equivalent factorized
  full and partial circuits.
• Shortened the Karatsuba XOR dependency chain through clmad.lo accumulation.
• Added a full-grid, one-nonce-per-thread kernel for pool work and the offline
  benchmark.
• Retained the persistent kernel for the rare block-target path where early
  exit is valuable.
• Kept the final sm_120 hot kernels at 64 registers per thread with no static
  stack, local-memory spill, or shared-memory allocation.

The performance gain comes from doing less work per hash and improving the
GPU execution path. It is not presented as a result of simply raising the
power limit.

User-visible changes

• Replaced the verbose normal GPU startup with a compact NOID-GPU-WORKER
  panel showing payout, selected GPU indices, hardware-test result, and
  mining status.
• Removed the obsolete allow-custom-coinbase startup advice.
• Public CLI names, help, status, benchmark labels, warnings, CUDA errors,
  launchers, and HiveOS documentation are now in English.
• Operating-system I/O failures no longer inherit localized Windows text;
  they use an English category and retain the numeric OS error code.
• Detailed CPU, batch, and arithmetic diagnostics remain available through
  explicit self-test and benchmark modes.

Correctness and release gates

• Added a deterministic offline benchmark of the real GPU PoW kernel with
  separate CUDA-event and wall-clock timing.
• Compared 16,384,000 optimized GF(2^128) products against emulated and
  independent schoolbook paths.
• Compared 40,960 fast permutations with the generic matrix path.
• Verified 64 complete CPU/GPU digests, 101,029 target comparisons, both
  positive search paths, and exact attempted-nonce counts.
• Every nonce returned by CUDA is still verified by the CPU; a mismatch stops
  the miner instead of submitting unverified work.
• Preserved stale-job detection and bounded concurrent share submission.

The miner/worker and Poseidon2b suites pass 113 tests with zero failures
(90 miner/worker tests and 23 arithmetic tests). Exact packaged executables
also passed offline GPU correctness and benchmark gates.

Packages and compatibility

Public assets:

• noid-gpu-windows-1.1.0.zip
• noid-gpu-linux-1.1.0.tar.gz
• noid-gpu-hiveos-1.1.0.tar.gz

Requirements:

• NVIDIA Ampere or newer: sm_80, sm_86, sm_89, sm_90, or sm_120.
• NVIDIA driver 580 or newer.
• CUDA runtime included; no CUDA toolkit required on the mining rig.
• Linux/HiveOS maximum required GLIBC symbol: GLIBC_2.30, compatible with
  HiveOS Focal/Ubuntu 20.04 GLIBC 2.31.

The public noid-gpu build supports its built-in pools and does not accept an
arbitrary node. The unrestricted open-pool, no-developer-fee noid-worker is a
separate private operator/test build, not a public 1.1.0 asset.

Developer fee

The public miner fee is 7.5% of scheduled mining time, implemented as a
45-second window in every 600-second cycle. Switching occurs only between
jobs, so a short sample can vary slightly at a job boundary. The full fee
address is printed at startup, and the miner never changes a block coinbase.
Pool fees remain separate.

Power-limit guidance

A configured power limit is a ceiling, not a target. Raising it alone does
not make a GPU consume more power or hash faster when the card is already
below the cap. Tune core clock and power one change at a time and compare both
MH/s and MH/s/W. Reject settings that introduce verification errors, rejected
shares, instability, or thermal throttling.

Pool fee

The Parano1d Pool fee is now 7.5%, reduced from 10%. The miner developer
fee is separate and is also 7.5% in this release.

noid-gpu v1.0

GPU miner for Parano1d (NOID).

Windows — noid-gpu-1.0-windows-x64.zip
Linux — noid-gpu-1.0-linux-x64.tar.gz
HiveOS — noid-gpu-1.0.tar.gz (use its URL as the custom miner Installation URL)

The bare noid-gpu.exe and noid-gpu are here too, for anyone who wants just the binary.

noid-gpu --gpu --coinbase o1YOURADDRESS --worker rig1

Measured per card, from the pool's own share accounting — the rate implied by shares actually delivered, not what the miner claims about itself:

card	per card
RTX 5090	62 MH/s
RTX 4090	42 MH/s
RTX 5080	32 MH/s
RTX 5070 Ti	26 MH/s

One process drives every card in the machine.

The miner reaches the Parano1d pool through parano1d-pool.fun and falls back through several addresses, so the pool can move without breaking miners already installed. If it loses the pool while running, it goes looking for it again instead of hanging. --pool ariapool mines on AriaPool instead.

This miner carries a disclosed 10% dev fee — 60 seconds out of every 600, under o1efnn3pn9c3p7etdv0addk9pply64cveujmkks7vt4rx0vjy2zufssg84zj, announced on every run with the address in full, and it never touches the coinbase of a block. If the fee cannot be delivered, the miner stops instead of mining. Details in the README.

NVIDIA driver 580 or newer. On HiveOS use a jammy or noble image.

Pool dashboard: https://parano1d-pool.fun

Verify your download against SHA256SUMS.txt.