OverXelis v1.1.1 · Final release

Your CPU.
Your XELIS.
Your control.

OverXelis v1.1.1 is the current release of the XELIS CPU miner. Beta testing is complete. Build it on your Linux computer, set your payout address and mine with your chosen pool.

Linux x86-64 · Build from source.

Linux x86-64CPU-only source edition
1% disclosedScheduled developer mining time
Your pool. Your wallet.No private keys required
Current releaseOverXelis v1.1.1
01 / The approach

Know what your
miner is doing.

Clear settings, visible fees, and correctness checks before hashing. No promises of guaranteed returns or a universal speed advantage.

01

Choose your destination

Enter your own XELIS payout address, compatible Stratum pool, and worker name. No wallet seed, custody, or deposit.

02

See both sides of the fee

Separate USER and FEE modes show where work is going. The terminal reports accepted shares and active-time accounting.

03

Test without a hidden service

Run the miner in the foreground and stop with Ctrl+C. It does not install an autostart service or change your clocks.

OverXelis v1.1.1 · Release evidence

Recorded runtime and hashrate

Four computers logged 609.24 machine-hours on real Kryptex XELIS jobs. The two 24-thread rigs contributed 290.89 machine-hours combined. Each system used huge pages for all workers.

Beta.5 testing of the code promoted to OverXelis v1.1.1, through the final beta shutdowns on 18 November 2026 at 15:37 UTC.
ComputerThreadsRecorded hoursAll sessions (kH/s)Last session (kH/s)
DesktopAMD Threadripper PRO 7995WX192171.15122.698123.114
CPU Rig 132147.2012.46912.462
CPU Rig 224143.978.5618.558
CPU Rig 324146.929.1529.152

“All sessions” weights each miner-displayed session average by its recorded duration. “Last session” is the average displayed when each computer ended its final beta run. The rates include user and developer-fee work. Runtime includes warmup and reconnects. Gaps between runs and starts that never reported a rate are excluded. The fleet total uses unrounded durations.

Kryptex accepted 86,134 shares, including 85,276 user shares and 858 developer-fee shares. The miners recorded 9 rejections and 195 submissions with an unresolved outcome. Sustained 140 kH/s and matched CPU package power remain unverified. Download the final runtime data ↗ · Earlier 15:20 UTC snapshot

Development update · 9 November 2026

Beta.5 Kryptex trial

Archived result from the beta.5 testing period.

In a trial that ended November 9, 2026, a beta.5 development build ran on Kryptex for 33 hours, 31 minutes, 18 seconds on an AMD Threadripper PRO 7995WX with 192 huge-page workers. The miner displayed a session average of 124.060 kH/s. The run recorded 13,911 accepted user shares, 137 accepted fee shares, zero rejected shares and 13 unresolved submissions during connection interruptions. The CPU peaked at 83.875 °C, below the 85 °C guard. All workers closed normally after a requested stop.

The rate comes from the miner’s session display. The run did not capture a completed-hash counter window or CPU package power. It does not establish a matched speed gain or qualify the 140 kH/s target.

The source preview is based on the tested development kernel and warm-handoff runtime. The public package was rebuilt with its own version metadata. The 33-hour trial predates that public rebuild.

The separate Stage 3 scratchpad patch remains in development. Its offline checks passed, and desktop qualification and a matched Kryptex test are pending.

Read the validation scope · View the sanitized trial evidence · Read the public validation notes.

02 / Earlier beta.4 hotfix · 6 November 2026

Measured work.
Clear conditions.

The beta.4 logged hour, short GPU-on counter measurements and unpatched hour use separate evidence and methods.

Beta.4 preview 2 crash fix. The earlier beta.4 source preview 1 can stop with an integer divide-by-zero fault, shown by the shell as “floating point exception.” Preview 2 checks the selected production kernel before that division, consumes the failed nonce, and excludes it from completed work and share submission. The first failing input is retained in the native runtime for diagnostics; the public launcher does not automatically save it to disk. The cause of the original zero intermediate value remains unverified because that input was not saved. Read the fix, qualification and limits.

Development hotfix logged hour · 6 November 2026

60 minutes 27.49 seconds after warmup · 192 huge-page workers · 119.739756 kH/s mean of 361 rounded local displays; sampled minimum 119.517, maximum 119.978 kH/s. The measured interval accepted 444 USER shares; the full session accepted 450, with zero rejected or unresolved displayed. All native workers closed normally after a requested stop.

This is logged continuity evidence from the development hotfix, whose native library is byte-identical to the qualified public preview-2 build. The launcher retained its earlier preview-1 banner. Raw hash-counter boundaries and a monotonic interval were not saved; native drops/undefined skips, GPU activity, cooling and temperatures were not recorded for this session. This is not an exact-counter hour, a full fee cycle or a 120 kH/s guarantee.

Read the dated method, identity and limits · Inspect sanitized interval data.

Earlier short hotfix counter measurements

6 November 2026 · 7995WX, 192 hugetlb workers, GCC 15.2 non-PGO, GPU mining running. Actual windows follow at least 60 seconds of settling.
Hotfix runCompleted-hash local rateMeasured intervalAccepted USER shares
First run119.302920 kH/s191.063907 seconds25
Repeat run118.955001 kH/s190.778115 seconds28
CHECKED FAILURE PATH

No invented hash

A zero round modulus is rejected before division. That nonce is consumed, is not counted as completed work, and cannot enter the submission queue.

REAL POOL WORK

25 and 28 accepted shares

Two approximately 191-second measured windows after warmup, with zero rejected, unresolved or queue-dropped work. These are short observations, not a soak test.

CORRECTNESS CHECKS

1,031 inputs · four C paths

Saved Rust answers replayed, complete scratchpads compared, and 192 actual huge-page workers verified. Forced-zero tests exercised 1, 24, 32 and 192 workers.

Scope: zero rejected, unresolved or queue-dropped work was reported in these short measurements. Saved Rust answers were replayed through four C implementations for 1,031 inputs, full scratchpads were compared, and 192 actual huge-page workers were verified. Forced-zero guard tests exercised 1, 24, 32 and 192 workers. Rust was not newly executed. These checks support the observed selected-kernel repair and normal hash behavior; they do not establish long-term stability, broad hardware performance, an exact-counter hotfix hour, a fee-cycle pass or a general 120 kH/s guarantee.

The production selected kernel uses the checked path. Reference and experimental kernels retain their earlier zero-modulus behavior; the legacy void selected API deliberately traps on failure. The hotfix does not define a new consensus hash for an undefined input. Capturing and replaying a naturally failing input remains necessary to distinguish a reference edge case, an optimization discrepancy or a transient computation fault.

Historical, unpatched beta.4 profiles

Source preview 1 is superseded because of the recorded crash. Its GPU-paused hour averaged 121.157224 kH/s across 366 rounded Local displays, minimum 120.908 kH/s, over 3,671.545 active USER seconds after settling, with 393 accepted shares. This is unpatched-build evidence, not hour-long validation of the hotfix. The shorter desktop and rig records below also belong to that earlier source. Read the preserved methods and conditions.

Short desktop profiles: 24 and 32 threads

These separate Kryptex runs used the same 7995WX desktop build, huge pages, paused GPU mining and Mining100 cooling. Each followed 60 seconds of settling. Rates below use gross completed hashes divided by the measured interval; they are not the hour’s rounded-display metric or results from separate physical rigs.

6 November 2026 · short real-pool measurements on the 7995WX desktop; zero rejected or unresolved shares.
Desktop profileGross completed rateMeasured window and accepted work
24 threads23.844 kH/s311.380 seconds · 7,424,693 completed hashes · 3 accepted USER shares.
32 threads31.786 kH/s311.512 seconds · 9,901,797 completed hashes · 13 accepted USER shares.

Three physical rigs: local builds and startup checks

These are separate computers, not the desktop thread-count profiles above. Each rig compiled the source locally, passed the full 1,031-input/four-C hash and full-scratchpad qualification, verified all requested huge-page workers, and submitted fresh Kryptex USER work. The desktop native library was not copied to them.

Startup qualification only · different durations; no matched or sustained rig hashrate comparison is claimed.
Physical rigCompilerObserved startup result
Rig 1 · 32 threadsGCC 13.3.032 huge-page workers · 1 accepted USER share at first success · 0 rejected / unresolved.
Rig 2 · 24 threadsGCC 13.3.024 huge-page workers · 1 accepted USER share at first success · 0 rejected / unresolved.
Rig 3 · 24 threadsGCC 11.4.024 huge-page workers · 1 accepted USER share at first success · 0 rejected / unresolved.

Preserved beta.3 versus beta.2 results · 1 November 2026

The earlier matched native benchmark below is historical. It used GCC 14 and a different measurement method from the historical unpatched beta.4 live-pool hour; it is not a matched beta.4 comparison.

Beta.3 matched benchmark111.168 kH/s

192 threads · hugetlb · GCC 14 · ordinary non-PGO build. Four separate 60-second candidate samples, with 15 seconds of warmup before every sample.

1 November 2026 UTC · same desktop, compiler, memory mode and ordered CPU selection. Each sample used one native library in a fresh process.
Software / buildWeighted local rateDifferenceMeasurement
OverXelis beta.2public-source-preview.2 · GCC 14110.234 kH/sReferenceFour 60-second control samples interleaved with the candidate. Roles alternated order between pairs.
OverXelis beta.3public-source-preview.1 · GCC 14111.168 kH/s+0.85%Aggregate weighted gainFour separate 60-second candidate samples: 110.877–111.611 kH/s. The exact public build reproduced the measured native binary.

Other XELIS CPU miners — recorded hashrates

These are the same recorded miner rates shown on the beta.2 website. They come from historical sessions with different dates, settings and measurement windows. The matched beta.3-versus-beta.2 benchmark above is a separate test.

Historical observations published 29 October 2026. Original software versions and test context are retained.
Software / buildRecorded local rateTest context
BzMinerv100.46106.360 kH/sHistorical mean: 1,984 reports after warmup in a roughly 34.6-hour run. Different dates and measurement method.
SRBMiner-MULTIv3.7.5105.867 kH/sFive short historical trials; ten later displays and 55 accepted-share events. Startup exclusions are disclosed below.
TNN-Minerv0.8.0101.970 kH/s40 live displays; 83 user shares accepted, zero user rejected. Snapshot at 04:05 UTC; tuning excluded.
cpuminer-opt-suprAVX2; banner v1.40 / folder v1.4795.786 kH/sEight displays and three accepted shares in a 48-second startup record. Version discrepancy remains unresolved.
hoo_cpu / HTNv1.4.397.018 kH/sFive later rolling displays in a short trial; 19 accepted USER shares. Not the newer v1.4.24.
HydraX Minerv3.4.519.242 kH/s10 min 19 sec; 30 displays and 11 accepted shares. 4 KB pages vs OXM’s 2 MiB huge pages: settings differ.

Read the original measurements, exclusions and historical arithmetic comparisons ↗

Preserved beta.3 functional checks

BETA.3 LIVE POOL CHECK

110.870 kH/s

Local session average over a finite 300-second USER pool run, including connection and job waits. This is the application’s local counter display, not a pool-side rate estimate.

ACCEPTED WORK

38 accepted · 0 rejected

Zero unresolved, locally stale, duplicate or invalid shares and zero native queue drops. All 192 workers used huge pages; the run exited normally and closed its workers.

CORRECTNESS CHECKS

124 + 15 tests

124 production tests and 15 tuning-tool tests passed, alongside the 1,031-input hash and full-scratchpad qualification, public self-test and actual 192-worker layout check.

The 110.615 kH/s development target was exceeded in these tests. All four candidate benchmark windows and the final five-minute pool-run local average exceeded it. This is evidence from one desktop with the other CPU miner stopped, not a guarantee on every machine. A matched 5% lead over BzMiner, a full live beta.3 fee cycle, deliberate reconnect testing and a long soak remain unverified.

The benchmark is synthetic native work; the separate pool run produced accepted USER shares. Weighted averages use total completed hashes divided by total measured seconds. The four candidate samples are not one continuous four-minute run. Saved Rust reference answers were replayed during qualification; Rust was not newly executed.

03 / Current release

Download OverXelis v1.1.1

Download the Linux source and verify it before building on the computer that will mine.

Check your setup

  • Operating systemLinux x86-64
  • CPU supportAES-NI required
  • Build toolsPython 3.10+, GCC, GNU binutils
  • Build modeOrdinary non-PGO
  • Memory modeOrdinary or preconfigured huge pages
  • User accountNormal user
  • Pool protocolXELIS-compatible Stratum

The 192-worker huge-page profile belongs to the tested workstation. Verify your own compiler, CPU and memory mode.

Read the requirements ↗

04 / Get started

From source to
your first share.

Download and verify first. Follow the complete guide for dependencies, build checks, memory modes, and troubleshooting.

01

Verify, then extract

Download the archive and its checksum file. Stop if the checksum fails. A checksum is not a publisher signature.

sha256sum --check SHA256SUMS-v1.1.1.txt && \
tar -xzf OverXelis-Miner-1.1.1-source.tar.gz && \
cd OverXelis-Miner-1.1.1
02

Build and qualify

Stop other CPU miners normally. The build command uses GCC from your PATH. The miner fetches no compiler. Do not bypass a failed hash check.

python3 -I -B build_public.py --compiler gcc && \
python3 -I -B qualify_public.py && \
bash run.sh --self-test
03

Add your settings

The terminal wizard asks for your public payout address, pool, worker name, threads, and memory mode. No private keys.

bash run.sh --configure && \
bash run.sh --check
04

Start deliberately. Stop anytime.

The flag acknowledges the disclosed 1% scheduled developer fee. Keep the terminal open. Ctrl+C stops this miner.

bash run.sh --config config.json --accept-dev-fee

New to source builds? Start with the complete installation guide, not a copied command alone.

05 / Optional configuration helper

Your address stays
in your browser.

Create config.json locally. This form sends no data to a server and stores nothing in browser storage. The terminal wizard is also available.

Generate your local configuration

Public address only. Never enter your seed phrase or private key. Basic syntax checking is not full address validation.

Use the endpoint published by your pool. TCP is unencrypted; TLS requires the pool’s TLS port.

8 is an example, not hardware detection.

Not an exchange or wallet password.

This helper needs JavaScript. Use bash run.sh --configure from the installation guide when scripts are unavailable.

Public address. Private control.

Move the downloaded file into the extracted miner folder. Review it before mining. Do not upload it in a feedback report.

chmod 600 config.json && \
bash run.sh --config config.json --check
A configuration file is not a miner installation. You must still build and pass qualification. The public launcher refuses root and never changes system huge-page settings.

The website does not validate ownership or the XELIS address checksum. Double-check your payout address before use. Your pool must accept it.

06 / A fee you can see

99 minutes for you.
Up to 1 for development.

No wallet withdrawals. No seed phrases. The fee is a scheduled allocation of mining time, not a promise of exactly 1% of coins.

99 minactive USER work
≤60 sFEE window

Paused or offline user time earns no fee credit. Failed developer connections are skipped; no catch-up debt. Restarting the miner resets the fee schedule. Network handovers add overhead, and operating-system timers are not hard real-time.

Developer destination

Kryptex North America · XELIS

stratum+tcp://xel-us.kryptex.network:7019
xel:pqu4e39zln0g7lx9grdhj3tavuy6e45sjglp9ss9n6lx2zc0w9fqqxhxxvt

The terminal shows separate USER/FEE workers and share counts. Local hashrate includes both roles; it is not your net payout rate. Pool fees are separate.

07 / What to expect

Honest answers.
Useful testing.

Is OverXelis officially part of XELIS?

No. OverXelis Miner (OXM) is an independent community project. It is not affiliated with or endorsed by the XELIS project, Slixe, or Kryptex. XELIS is the network this software mines.

With gratitude: thank you to Slixe (GitHub profile, opens in a new tab) for creating XELIS and for his development work, and to the wider XELIS contributor community. We deeply value the project’s work on privacy, open participation, and programmable digital money. This acknowledgment expresses our appreciation—not an endorsement of OXM.

What do our recorded tests show?

See recorded runtime and hashrate for 192, 32 and 24 threads ↗

The earlier result below is one session within that total.

The Kryptex trial lasted 33 hours, 31 minutes, 18 seconds on an AMD Threadripper PRO 7995WX with 192 huge-page workers. The miner displayed a session average of 124.060 kH/s. Kryptex accepted 13,911 user shares and 137 fee shares. The miner recorded zero rejected shares and 13 unresolved submissions during connection interruptions. The CPU peaked at 83.875 °C, below the 85 °C trial guard.

The trial recorded a session-display rate. A completed-hash counter window and CPU package power were unavailable. A matched speed gain and the 140 kH/s target remain unverified. Read the beta.5 trial and the earlier beta.4 validation.

Can I use Windows, a GPU, or an ARM computer?

This download is a Linux x86-64 CPU source release. It does not include a Windows executable, a macOS build, an ARM build, or a GPU miner. Do not assume compatibility with untested environments.

Will opening this page use my CPU to mine?

No. This is a static project website. Mining begins only when you deliberately run the installed miner on your own computer. There is no browser miner, remote activation, or automatic service installation.

What still needs validation?

The separate Stage 3 scratchpad patch remains in development. Its offline checks passed, and desktop qualification and a matched Kryptex test are pending. A sustained 140 kH/s result, package-power comparison and broader hardware qualification remain open. Read the current scope.

Why do local and pool hashrates differ?

The console counts local completed hashes. The pool estimates useful work from submitted shares. Short samples, connection gaps, stale shares, and fee allocation can affect the comparison. Report the measurement window and accepted/rejected counts.

What happens if I see an error or a security warning?

Stop on invalid hashes or checksum errors. Preserve redacted logs. Do not blindly disable security software. The native workers retain bounded nonce tracking instead of beta.2’s 32,768-entry cache stop. Duplicate and lower nonce protection persists across jobs, reconnects and USER/FEE changes. This correction does not establish long-soak reliability.