ForgeMiner 2.0.0 cuts the Pearl, Quantus and NOID fee to 1% and adds Zano and CSD

ForgeMiner 2.0.0 was released on 13 November 2026. The developer fee on Pearl, Quantus and Parano1d (NOID) falls from 2% to 1% on NVIDIA and AMD. The same build adds Zano and Compute Substrate (CSD), and the author reports higher hashrates on several cards. The figures below are the developer's own tests, from the ForgeMiner 2.0.0 release notes, not independent measurements.

Fee cut and reported speed

• Pearl, Quantus and NOID: developer fee 1% (was 2%), on NVIDIA and AMD.
• Pearl on the RX 9000 series (RDNA 4): +14.7% at the same power, measured on an RX 9060 XT.
• Pearl on the RTX 30, 40 and 50 series: up to +1.5% at the same power (RTX 5060 Ti +1.5%, RTX 4070 Ti +0.5%), with fewer invalid shares and less work lost at a block change.
• Quantus: +0.9% on the RTX 30 series, +0.7% on the RTX 20 series and CMP 40HX / 50HX, and up to +0.2% on AMD RX 6000 / 7000 / 9000.
• NOID: about +39% on the RTX 20 series (RTX 2080 Ti) and +1.3% on the RTX 40 series.

Pearl is mined on the pearl-pow algorithm. On RTX 50 cards the fastest Pearl kernel is now the default (--kernel 1). --kernel 2 uses about 2% less board power for about 1% less hashrate. If you set --kernel 2 on an RTX 50 card earlier to go faster, remove it. Both kernels need NVIDIA driver 580 or newer. On the RTX 30 and 40 series there is now only one Pearl kernel, so --kernel no longer changes anything there.

New coins and other changes

Zano is launched with --algorithm zano on NVIDIA and AMD, with a 0.7% developer fee. On profit-mine.com it is mined with ProgPowZ. CSD uses --algorithm csd, also on NVIDIA and AMD, with a 1% developer fee; we do not yet have a CSD page. The release also includes ready-made launchers for LuckyPool.

• Separate memory temperature limits: --mem-temp-limit and --mem-temp-resume on Windows cards that report memory temperature (GDDR6X / GDDR7). A card with hot memory no longer stays paused because of a core limit. --temp-limit and --temp-resume still follow the core.
• On CMP 70HX, --mclk 5000, 810 and 405 now hold, as on the CMP 90HX and 50HX, with NVIDIA driver 610.43.03. The author reports that Pearl at 810 MHz uses about 29% less board power for 0.3% less hashrate, and at 405 MHz about 32% less power for 1.3% less hashrate.
• Pearl shares are sent compressed on pools that support it. On Linux, AMD cards now show live core voltage in the Volt column and in the API.

The RTX 50 series still needs NVIDIA driver 580 or newer. For HiveOS, use the versioned custom-miner URL from the release, so the flight sheet actually moves to 2.0.0.