SOST Protocol
Gold-backed Proof-of-Work protocol. Native chain. No bridges. No smart contracts. No governance tokens.
Genesis Block: March 15, 2026 — 18:00:00 UTC
Gold-backed Proof-of-Work protocol. Native chain. No bridges. No smart contracts. No governance tokens.
Genesis Block: March 15, 2026 — 18:00:00 UTC
| Parameter | Value |
|---|---|
| Algorithm | ConvergenceX v2.0 (memory-hard, CPU-friendly) |
| Block time | 600s (10 min) |
| Max supply | 4,669,201 SOST (hard cap by construction) |
| Emission | Exponential decay, q = e(-1/4) ≈ 0.7788 |
| Coinbase split | 50% miner / 25% Gold Funding Vault / 25% PoPC Pool |
| Difficulty | cASERT unified (12h bitsQ half-life + equalizer profiles) |
| Encoding | SOSTCompact Q16.16 fixed-point |
| Coinbase maturity | 1,000 blocks (~7 days) |
| Premine | NONE |
| ICO / IEO | NONE |
| Dev tax | NONE |
| Source | GitHub (public at launch) |
Gold has 5,000 years of monetary history. It is the most trusted store of value across civilizations. But gold has zero programmability — you cannot send it over TCP/IP, you cannot split it into eight decimal places, and you cannot verify its custody without a trusted third party.
Existing gold-backed tokens (XAUT, PAXG) attempt to bridge this gap. They are ERC-20 tokens on Ethereum — centralized custodians hold the gold, the tokens can be frozen by the issuer, and the entire scheme depends on counterparty trust in a single entity. If Paxos or Tether Gold decides your tokens are problematic, they disappear. This is not sound money.
The trust model is the fundamental failure. XAUT holders trust Tether to maintain 1:1 reserves in a Swiss vault. PAXG holders trust Paxos to not freeze their tokens. Both trust their respective custodians to survive regulatory action, remain solvent, and not selectively enforce terms of service. The gold exists in a vault you cannot visit, managed by a company you cannot audit, under a jurisdiction that may change its rules. Every gold-backed ERC-20 is a bearer instrument for counterparty risk dressed up as sound money.
These tokens also inherit every weakness of the host chain. XAUT and PAXG are ERC-20 contracts — upgradeable, pausable, and subject to Ethereum gas economics. A congested Ethereum network means your gold is stuck. A regulatory order means your gold is frozen. The issuer can blacklist addresses, block transfers, and claw back tokens. None of this is hypothetical — Tether has frozen USDT on hundreds of addresses, and the same legal mechanism applies to XAUT. The custody model is broken because it replaces one trusted third party (a bank) with another (a token issuer).
There is also no mechanism for individuals to prove they hold gold and be rewarded for it. The entire gold tokenization ecosystem flows in one direction: custodian holds gold, user holds a claim. There is no protocol-level incentive for self-custody of physical or tokenized gold. The custodian model concentrates gold in fewer hands, which is the opposite of what sound money requires.
Bitcoin solved digital scarcity through proof-of-work. But Bitcoin has no structural connection to physical reserves. SOST connects PoW scarcity to physical gold accumulation at the protocol level — not through promises or governance votes, but through consensus rules that are immutable at genesis.
Every SOST block automatically allocates 25% of the coinbase reward to a Gold Funding Vault address. This is not a governance decision. It is not a DAO vote. It is a consensus rule — any block that does not include the correct 50/25/25 split to the correct constitutional addresses is invalid and rejected by every node on the network. The split is hardcoded at genesis and cannot be modified.
The Gold Funding Vault accumulates SOST, which is periodically converted to XAUT and PAXG via a rule-based TWAP execution policy (rate-limited, anti-momentum, publicly auditable). Purchased gold tokens are deposited into a Heritage Reserve on Ethereum mainnet — sealed by default. The reserve is intended to remain perpetual. It cannot be moved except under an Emergency Catastrophe procedure requiring both 75% PoW miner signaling and a Foundation Execution Order.
The remaining 25% flows to a PoPC Pool powering Proof of Personal Custody — a mechanism that rewards holders who self-custody gold and prove it. PoPC operates in two models:
Model A — Bond with Autocustody. For crypto natives and miners. The participant declares an Ethereum wallet holding XAUT or PAXG, locks a SOST bond (12-30% of gold value, dynamically scaled), and commits to maintain custody for 1 to 12 months. ConvergenceX block entropy schedules random audits. A script verifies the XAUT/PAXG balance via Ethereum RPC. On successful completion, the participant recovers their full bond plus a reward from the PoPC Pool (up to 22% APR for 12-month commitments). On failure — if the gold balance drops below the committed amount — the bond is slashed automatically: 50% to the PoPC Pool, 50% to the Gold Funding Vault. The gold never leaves the user's wallet. Only the SOST bond is at risk.
Model B — Timelocked Escrow. For gold holders who want zero slash risk. The participant deposits XAUT or PAXG into an immutable escrow contract on Ethereum — no admin key, no upgrade proxy, no pause function. The SOST reward is paid immediately at deposit (up to 7% APR for 12-month terms). At contract expiry, the depositor withdraws 100% of their gold. No audits. No bond. No slash. Custody is enforced by the escrow contract, not by trust.
The two models serve complementary populations. Model A attracts crypto-native participants who already hold SOST and want yield on their gold without giving up custody. Model B attracts traditional gold holders who want exposure to SOST without selling their metal. Together, they create a two-sided flywheel: Model A generates buy pressure for SOST (participants must acquire SOST for bonds), while Model B distributes SOST to gold holders who may never have interacted with cryptocurrency. Anti-whale caps limit per-wallet participation to prevent pool depletion — 50 oz per wallet per epoch for Model A, 10 oz for Model B.
Both models use prices frozen at contract creation from a public Price Bulletin. No on-chain oracle is required for consensus. PoPC launches post-genesis as Ethereum smart contracts using ConvergenceX block entropy for audit randomness.
SOST is not pegged to gold. Holders have no redemption right against the reserve. The reserve ratio (reserve_value / circulating_supply) is an observable metric that markets price as they see fit. There is no stablecoin mechanic, no oracle dependency for consensus, and no algorithmic peg.
ConvergenceX is not a hash lottery. Each mining attempt requires solving a 32×32 linear system A(x) = b through 100,000 sequential rounds of integer-only gradient descent, then proving the solution sits in a stable basin of attraction. The proof is a verifiable mathematical certificate, not a lucky hash.
Memory-hard dataset: A 4GB persistent dataset (512M uint64_t entries) is generated once per block from the previous block hash. Combined with the 4GB scratchpad, total mining memory is ~8GB. It is reused across all nonce attempts for that block. CPU miners benefit from L3 cache reuse. ASICs cannot reduce the memory requirement — checkpoints every 6,250 rounds commit intermediate state to a Merkle tree, catching any attempt to shortcut memory reads. Node validation requires only ~500MB RAM (no dataset or scratchpad needed).
Per-block program: Each block generates a unique 256-operation program derived deterministically from the block hash. Operations: MUL, XOR, ADD, ROT, AND, OR, NOT, SUB (8 types). The program changes every block. No fixed circuit can be optimized for a program that does not yet exist. An ASIC for ConvergenceX v2.0 would require 8GB LPDDR5 on-package (4GB dataset + 4GB scratchpad) and a general-purpose ALU — economically equivalent to building a CPU.
Unified cASERT difficulty: The bitsQ primary controller handles long-term hashrate calibration with a 12-hour half-life and per-block adjustment. The equalizer is a unidirectional component that hardens ConvergenceX stability parameters when the chain runs ahead of schedule (emission protection). The equalizer uses profiles (E3–H6) that progressively tighten stability requirements. If no block is found for an extended period (max(ahead 00D7 10min, 2 hours) 2014 proportional to blocks ahead, no ceiling), a decay mechanism gradually reduces the equalizer level to prevent permanent chain stall. Decay affects mining only, never block validation.
Decay tiers: L8+ drops 1 level every 10 min (fast), L4–L7 every 20 min (medium), L2–L3 every 30 min (cautious). Cost uses the current effective level as it drops. Example — L10 (450 ahead): Activation at 75h. Then: L10→L8 FAST (30min), L7→L4 MEDIUM (80min), L3→L1 SLOW (60min). Total decay: 170 min. Total from start: ~78 hours.
All consensus arithmetic is integer-only. Zero floating-point operations exist in any consensus path. Difficulty is encoded as SOSTCompact Q16.16 fixed-point (uint32_t), eliminating cross-platform rounding divergence.
The emission formula: reward(height) = floor(R0 × qepoch), where epoch = height / 131,553. All arithmetic uses __int128 floor-division. The maximum supply is a mathematical consequence of the geometric series — not a tracked variable.
Of the total supply, approximately 50% goes to miners, 25% to the Gold Funding Vault, and 25% to the PoPC Pool. These ratios hold for every block from genesis to the final emission. There is no founder allocation, no team vesting schedule, and no treasury. The Foundation operates exclusively from protocol fees charged on PoPC contracts (5% on Model A, 10% on Model B) — never from minted coins.
End of emission: The last block with a nonzero subsidy occurs around epoch 81 (~year 2228, ~202 years after genesis). After that, block rewards are exactly 0 — the integer arithmetic truncates sub-stock values to zero. There is no tail emission (unlike Monero). The hard cap is not a counter that can be changed — it is a mathematical consequence of the geometric series converging. No function in the code can mint new coins. After emission ends, miners are compensated entirely by transaction fees. The same 50/25/25 split applies to fees: 50% to the miner, 25% to the Gold Funding Vault, 25% to the PoPC Pool. The minimum relay fee is 1 stock/byte (~0.00000001 SOST/byte), ensuring every transaction always pays something. This is the same economic model as Bitcoin — long-term security funded by fee revenue — but with the structural advantage that 25% of all fees continue accumulating gold reserves in perpetuity.
| Phase | Date | Milestone |
|---|---|---|
| 1 | March 15, 2026 | GENESIS BLOCK — 18:00:00 UTC. PoW chain live. Solo CPU mining. C++ miner released. GitHub public. |
| 2 | Q4 2026 | PoPC smart contracts deployed on Ethereum Sepolia testnet. Model A (bond + audit) and Model B (escrow timelock) open for public testing. |
| 3 | Q1 2027 | PoPC mainnet launch on Ethereum. First custody contracts active. |
| 4 | Q2 2027 | XAUT/PAXG reserve purchases begin (8 months post-public-launch per G4, subject to sufficient SOST market liquidity). Heritage Reserve deployed on Ethereum mainnet. Public reserve dashboard. |
| 5 | 2027+ | Multi-metal tokenization roadmap (silver, platinum). Physical gold proof extensions. Mobile audit app. ZK research. |
Requirements:
Build:
git clone https://github.com/SOST-Protocol/sost-core.git cd sost-core && mkdir build && cd build cmake .. -DCMAKE_BUILD_TYPE=Release && make -j$(nproc)
Run node (terminal 1):
./sost-node --genesis genesis_block.json --chain chain.json \
--rpc-user YOUR_USER --rpc-pass YOUR_PASS --profile mainnet
Run miner (terminal 2):
./sost-miner --address sost1YOUR_ADDRESS_HERE \
--genesis genesis_block.json --chain chain.json \
--rpc 127.0.0.1:18232 \
--rpc-user YOUR_USER --rpc-pass YOUR_PASS
The miner connects to your local node via RPC, fetches mempool transactions via getblocktemplate, and submits solved blocks via submitblock. First solo miner to solve a valid ConvergenceX proof gets the block reward.
| Property | Bitcoin (SHA-256d) | Monero (RandomX) | SOST (ConvergenceX) |
|---|---|---|---|
| PoW verify cost | ~1μs | ~50ms | ~5-30s |
| Memory (mining) | Negligible | 256MB-2GB | 8GB (4GB scratchpad + 4GB dataset) |
| Memory (node) | Negligible | ~256MB | ~500MB (no scratchpad/dataset) |
| ASIC resistance | None | High | Very high |
| Difficulty adjust | Every 2016 blocks | Every block (LWMA) | Every block (cASERT exp.) |
| Block time | 600s | 120s | 600s |
| Anti-stall | None | LWMA | cASERT Decay (4h/8h tiered) |
| Emission | Halving / 210K blocks | Tail emission | Smooth exp. decay (q=e^-¼) |
| Max supply | 21M BTC | Infinite | ~4,669,201 SOST |
| Constitutional reserve | None | None | 25% gold + 25% PoPC |
SOST uses an 8-layer verification pipeline:
Fast sync: Layers 1-7 always run. Layer 8 (CX recompute) is skipped for checkpointed blocks. Initial sync takes minutes, not days.
Full verification: With --full-verify, every block is fully recomputed. Takes weeks but provides maximum security.
SOST is experimental software. Mine and hold at your own risk. The gold reserve mechanism depends on market availability of XAUT/PAXG. The reserve is not a redemption right — SOST is not pegged to gold. This is not financial advice.