[GUIDE] Mining PCoin (PCN) on an ordinary CPU - RandomX, no separate miner neede

[pars5555]

Short version: PCoin is a Bitcoin Core v29.4 fork with RandomX proof-of-work. The
miner is built into the node, so there is no xmrig to configure, no
stratum URL to get wrong and no config file. One command and your CPU is mining.

Full announcement and the honest disclosures are in the ANN thread:
topic 5591113. This
post is just the practical part.


Start mining in one command

Linux (Debian/Ubuntu). The installer verifies its download against the
release checksums and runs a setup wizard that asks for your payout address:

Code:

curl -fsSL https://pc.am/dl/install.sh | sudo sh

Windows - portable zip, node + CLI + a tray miner:
github.com/pars5555/pcoin/releases
Mining is off by default on purpose. Turn it on from the tray once the node has
synced - a node that mines before it has verified the chain is building on
something it has not checked, and anything it finds is thrown away.

Android - a miner APK is in the same releases page. A phone does perhaps
50 H/s, which is not much, but it works and it costs you nothing.

Manual control on Linux if you skip the wizard:

Code:

sudo -u pcoin pcoin-cli -datadir=/var/lib/pcoin getcpuminerinfo
sudo -u pcoin pcoin-cli -datadir=/var/lib/pcoin startmining "<your pc1q... address>" 4

The last number is threads; 0 means every core.


What to actually expect from a CPU

Measured on my own machines, light-mode RandomX:

CPU	threads	H/s
i9-10900X (20 threads)	4	264
i9-10900X (20 threads)	10	595
i9-9900K (16 threads)	3	235
6-core desktop	1	74
Android phone	-	~50

More threads is not proportionally more hashrate. Light mode shares a
256 MiB cache, so past roughly a third of your cores you are mostly buying cache
contention. One machine went from 13 threads to 3 and lost only 58% of its
hashrate - per-thread rate nearly doubled. Try a few settings before assuming
100% is correct.


The arithmetic, including the part you will not like

Do not trust a hashrate number in a forum post, including this one. When I
first wrote this the network was around 17,000 H/s; within a day it had risen
several times over, and difficulty with it. Work it out yourself instead - it
takes ten seconds:

Code:

https://pool.pc.am/api/pools

Read networkStats.networkHashrate. The chain emits 50 PCN per block at a
600 second target, so roughly 7,200 PCN a day is split across everyone mining.

So: (your H/s / network H/s) x 7,200 = roughly your PCN per day.

That number falls as more people join - it is a share of a fixed emission,
not a rate. It has already fallen sharply once. If the network doubles, your share
halves, and nothing about your machine changed.

And the caveat that matters more than any of the arithmetic: PCN is not traded
anywhere. There is no exchange, no orderbook and no way to sell it. What it
currently does is buy credit at five services I run (an uptime checker, a Telegram
site builder, an AI control panel and two 3D-model generators) at a rate I set.
That is a use, not a price, and you should treat it as worth zero dollars until a
real market exists.

If you want money, mine something else - genuinely. If you want to point idle
cores at a young chain and hold the coins in case it becomes something, this is
that.


Solo variance, and the pool

Solo mining is a lottery: you find a whole 50 PCN block or you find nothing.
At 500 H/s you would average a block every ~7 hours, but that is an average -
plenty of days you will see nothing and conclude it is broken.

The pool is open. It pays miners directly from each block's coinbase,
so there is no pool wallet, no private key on the server, no withdrawal to
request, and nothing an operator could run off with. PPLNS, 2% fee taken off the
block reward, minimum 0.00000294 PCN (the network dust limit). Your login is your
payout address - there is no registration.

Code:

pool.pc.am:3333

Live stats and the exact command for your OS: pool.pc.am

You need v1.3.0 or newer, and both the node and bitcoin-cli must be that
version - an older CLI sends the thread count as text and the call fails with a
type error. Solo still works exactly as before if you would rather keep the whole
block reward and the variance that comes with it.


Check everything yourself

Code:

Explorer API   https://explorer.pc.am/api/status
Pool API       https://pool.pc.am/api/pools
Supply         https://pc.am/supply/circulating.txt  (bare number, from a full node)
Source         https://github.com/pars5555/pcoin
Releases       published with SHA-256 checksums

Binaries are unsigned, so Windows SmartScreen will warn on first run. I do
not have a code-signing certificate. Say so plainly rather than have you find out.

Questions here or in the ANN thread.

[bogdan198]

At first glance, everything looks pretty well done—it’s clear you took your project seriously. Guess I should connect a rig to you, who knows, maybe something will actually come of it. 🤝

[pars5555]

Update: PCoin can now be mined with xmrig (xmrigCC), and one miner now holds a majority of hashrate

When I posted this guide the only way to mine PCN was the miner built into the node. That is still the easiest path, but it was also what stopped anyone already running xmrig from pointing spare cores at this chain.

1. rx/pcoin

There is now an rx/pcoin implementation for xmrigCC. It is submitted upstream and has not been reviewed or merged, so treat it accordingly:
https://github.com/Bendr0id/xmrigCC/pull/435

Linux x86-64 build with SHA256SUMS:
https://github.com/pars5555/xmrigCC/releases/tag/rx-pcoin-v1

Code:

./xmrigMiner -o pool.pc.am:3333 -u <your pc1q... address> -a rx/pcoin -t 4

No Windows build from me. Cross-compiling from Linux fails because the prebuilt xmrig deps are UCRT-linked and Ubuntu ships no UCRT mingw compiler. On Windows, either build the branch with MSVC, or just use the node's own miner (pcoin-win64-miner.zip in the PCoin releases) — that is the supported path and needs no trust in a binary from a stranger's fork.

2. If you tried xmrig against the pool before yesterday and got nothing, that was my bug

The pool rejected 100% of shares from any standard xmrig with "share above target", while the miner hashed perfectly. It rebuilt the submitted header with the nonce bytes in the opposite order from the xmrig convention, so it was hashing a header nobody had mined. My node's built-in miner and my pool were written from the same wrong assumption, so they agreed with each other and the bug stayed invisible until a third-party miner existed. Fixed — both byte orders are accepted now. Worth retrying if you gave up.

3. A disclosure I would rather make than have you find

Over the last 60 blocks, one single solo address found 34 of them — 56.7%. That is a majority of block production by one party, and it is not me. Two hours earlier the same measurement read 50%, so it is rising.

A single miner above 50% can reorganise the chain and double-spend. Nothing bad has happened — the chain is healthy, supply is exact, no reorgs — but you should know the number before you decide to point hardware at this. It is checkable: every coinbase is on https://explorer.pc.am

The honest reason I want more independent miners is that one, not price.

4. Current numbers

Height 3945. Blocks landing about every 535 s over the last 300, so roughly 8,100 PCN/day emitted in total. 192,250 PCN circulating. The pool has 22 miners.

Work out your own share rather than trusting any figure in a forum post, including this one. Read networkStats.networkHashrate from https://pool.pc.am/api/pools and compute (your H/s / network H/s) x 8,100. Network hashrate has roughly tripled in three days, so a number I write today is wrong by the weekend.

5. Unchanged, and still the important part

PCN is not traded on any exchange. No orderbook, no way to sell it. Treat it as worth zero dollars until a real market exists. It is a share of a fixed emission, so what you earn falls every time someone else joins. If you want money, mine something else — genuinely.

Everything else in the original post still stands.

[pars5555]

Correction to my own post: the 56.7% figure was a 60-block sample. Over 300 blocks it is 45.0%.

I told you to work out your own numbers rather than trust a figure in a forum post, including mine. Then I posted one that did not survive its own advice.

The fix

I said one solo address found 34 of the last 60 blocks — 56.7% — and that it was rising.

Over the last 300 blocks that address found 135: 45.0%, 95% CI 39.4-50.6%.

Block discovery is Poisson. At p≈0.5 across only 60 blocks the standard error is ~6.5 percentage points, so one reading moves about ±13 points at 95% confidence before anything real has changed. My four consecutive 60-block readings were 50.0, 56.7, 41.7, 53.3. I described the rise from the first to the second as a trend and it was not one — it was the same number, sampled four times.

What this changes for you, and what it does not

Does not change: one party is at roughly half of block production, and 45.0% with an interval touching 50.6% cannot tell you which side of 50 it is on. Know that before you point hardware here.

Does change: I cannot say it is rising, and I cannot say it is above half. I said both.

And a number I now have to put next to it

Over the same 300 blocks my pool found 51.0% (95% CI 45.3-56.7%). Having disclosed someone else's 45% I am not going to leave my own 51% out. 27 independent miners sit behind it and can leave in one config line, which makes it a softer kind of concentration than a single solo miner — but it is still a majority behind one endpoint I run, and a second independent pool would fix it properly. Pool code is MIT; I will help anyone who wants to stand one up.

How to check it yourself, correctly

Read coinbases from https://explorer.pc.am over at least 200-300 blocks. A pool block pays many addresses; a solo block pays one. Anything shorter and you will measure the sampling noise instead of the chain — which is exactly the mistake I made and published.

Current numbers

Height 4120, about 494 s/block over the last 300, so roughly 8,737 PCN/day emitted in total; 201,000 PCN circulating; 27 miners on the pool. Work out your share as (your H/s / networkHashrate from https://pool.pc.am/api/pools) x 8,737.

Unchanged, and still the important part

PCN is not traded on any exchange. No orderbook, no way to sell it. Treat it as worth zero dollars. It is a share of a fixed emission, so what you earn falls every time someone else joins. If you want money, mine something else.

[pars5555]

Follow-up: a different miner is now at 71%

I corrected a concentration figure a few hours ago. Re-measuring properly turned up something worse, so here it is straight away rather than tomorrow.

What changed

The address I have been reporting on is not the problem any more. A different address, which had never mined a block before height 4114, has found roughly seven blocks in ten ever since:

Code:

blocks 4069-4118    8%
blocks 4119-4168   76%
blocks 4169-4218   73%
blocks 4219-4268   67%
blocks 4125-4274   71%   (95% CI 64-78%)

That is one party, solo mining — every one of its blocks is a single-output coinbase, not a pool payout — holding a sustained majority of block production for about a day. It is not mine, and it is not on my pool.

What that means

A miner above 50% can reorganise recent blocks and reverse a payment that already looked settled. This is the plain 51% condition, and it is live right now.

What has actually happened: nothing. The explorer reports reorg_count 0, no blocks unwound, and supply exact against height. I have no evidence of hostile intent and the likeliest explanation is dull — someone with real CPU capacity pointed it at a chain that is free to mine. Capability is still capability, which is why this is a disclosure and not a shrug.

What I changed

Against a party holding 70%, no confirmation depth is provably safe — waiting longer buys time and cost, not certainty. The services that credit PCN deposits therefore pair their crediting depth with watching for reorgs, rather than trusting depth alone.

I am deliberately not publishing what that depth is. Printing the number a service credits at, directly beneath a 71% concentration figure, tells anyone who wanted to reverse a deposit exactly how deep to dig — and it is an operational setting, not something you need in order to judge the chain. The concentration figure is the disclosure; that number is not.

Edited: this paragraph originally named a specific confirmation count. The count has since changed, and publishing it was a mistake on its own terms.

pc.am carries the same disclosure as of tonight.

Check it yourself

Every coinbase is public on https://explorer.pc.am . A pool block pays many addresses; a solo block pays one. Use a window of several hundred blocks — 60 is mostly sampling noise, which is the mistake I made this morning and corrected this afternoon.

The fix is the same as it has always been

More independent miners, and a second pool run by somebody who is not me. That is the entire ask. There is still no exchange, no orderbook and no way to sell PCN, so I am not asking you to buy anything.