Alpenglow
Alpenglow (aSOLX) Whitepaper v1.1
Solana‑compatible experimental mainnet (hard‑fork) • Parallel execution
network • Low‑latency finality
This document extends v0.9 with formalized sections on system/threat
models, protocol sketches, economic functions, governance processes,
benchmarking, and a risk matrix. If there is any inconsistency with live
parameters, on‑chain records and official announcements prevail.
Table of Contents
- Executive Summary
- Background & Motivation
- System Assumptions & Threat Model
-
Protocol Design (Votor Consensus & Rotor Propagation)
-
Execution & Runtime (Sealevel, Token‑2022)
-
Performance Goals & Benchmark Methodology
-
Token Economics (Supply, Allocation, Emissions)
-
Sale & Distribution (Public Sale, Airdrop, Compliance)
- Governance
- Security
- Cross‑Chain & Ecosystem
- Roadmap
- Transparency & Status Page
- Risk Factors & Disclosures
- Glossary
- Appendix (Params, Pseudocode, TBF)
1. Executive Summary
Alpenglow aims to remain fully compatible with the Solana ecosystem
while pursuing ~100–150 ms p95 transaction finality via
Votor (a fast‑finality path with fallback) and
Rotor (one‑hop‑first dissemination with congestion
control). Under adverse conditions (up to ~20% byzantine validators plus
~20% offline, engineering target), the chain is designed to preserve
liveness. Target use‑cases include high‑frequency finance, real‑time
gaming, and AI/DePIN/IoT micropayments.
2. Background & Motivation
-
Pace differentiation: Solana mainnet optimizes for steady‑state
safety; Alpenglow pursues a parallel experimental path for measurable
low‑latency improvements.
-
Compatibility & optionality: SDK/Anchor/SPL compatibility; no
forced migration; applications choose to integrate as needed.
-
Evidence‑driven: publish weekly performance reports, open dashboards,
audit artifacts, and formal analysis to justify trade‑offs.
3. System Assumptions & Threat Model
3.1 System & Network Model
-
Time/slots and leader rotation with partial synchrony. The network
delay upper bound Δ drifts slowly under healthy conditions; messages
may be delayed, dropped, reordered, or duplicated.
-
Cryptography: Ed25519 signatures; VRF/random beacon for leader
selection and sampling; hash collision resistance.
3.2 Actors
- Validators / Leaders / Delegators / Relayers & Observers.
3.3 Adversarial Capabilities
-
Up to ~20% byzantine validators plus up to ~20% temporary offline
(engineering target).
-
Tactics: double‑signing, malicious voting, delay injection, partial
partitions, spam/flooding, leader snatching, penalty‑free replay
(mitigated).
-
Long‑range attacks and key leakage are considered; MEV manipulation
and stake centralization risks are addressed economically.
3.4 Goals
- Safety: vanishing probability of irreversible forks.
-
Liveness: progress to finality given bounded Δ and bounded adversary.
4. Protocol Design
4.1 Votor: Low‑Latency Finality (Design Sketch)
Idea: combine a fast‑path with single/double voting rounds and a
slow‑path with timeouts/leader changes. Validators maintain a lock on
the highest observed QC to prevent disordered rollbacks.
Objects
Block(b)
: candidate block.
-
Vote(v)
: validator’s vote for a block at a given round.
-
QC
(Quorum Certificate): aggregated signature proving ≥
2/3 voting power acceptance.
-
Lock
: validator’s lock on the highest QC (monotonic).
-
Commit(b*)
: commit of b*
and ancestors when
QC adjacency and lock rules are satisfied.
Fast‑Path (illustrative)
- Leader proposes
b_t
and broadcasts.
-
Validators vote
v_t
if not conflicting with their lock.
- Aggregated
QC_t
(≥ 2/3) is broadcast by leader.
-
If
QC_{t-1}
and QC_t
are adjacent and lock
rules are met, commit Parent(b_t)
(or
b_{t-1}
).
-
Otherwise or on timeout, the pacemaker elevates the round and the
leader rotates (slow‑path).
Safety Intuition (Informal)
Let N be participants, B a byzantine subset, and G the honest subset. If
|B|/|N| < 1/3 and lock monotonicity holds, two conflicting commit
chains cannot both obtain admissible QC sequences. Formal proofs and
state‑machine definitions will be released with audits.
4.2 Rotor: One‑Hop‑First Dissemination
-
Priority neighbor set of size K (low RTT / high reliability); fallback
neighbors for redundancy.
-
One‑hop‑first fanout: Leader → priority neighbors → remaining
neighbors; limit hierarchical fanout depth.
-
Backpressure & shaping: prioritize block > QC > vote >
gossip; drop or downsample under queue pressure.
- Adaptive K = f(BW, RTT, Loss, CPU) over observation windows.
-
Packet budgets cap bytes/frequency per message type to avoid
avalanches.
Message complexity: traditional multi‑layer fanout ~ O(n log n);
one‑hop‑first with stable topology approaches ~ O(n), reducing
duplicates.
4.3 Execution & Runtime
-
Sealevel parallelism with account/ISR scheduling based on read‑write
conflict graphs.
- Resource model: Compute Units (CU).
-
Token‑2022 extensions: confidential transfers, freeze controls,
compliance hooks (optional).
-
AI/data interfaces: standardized oracle hookups and off‑chain compute
proofs (future versions).
5. Performance Goals & Benchmark Methodology
5.1 Metrics
- Finality latency: p50/p95 (target p95 ≈ 100–150 ms).
-
Throughput: TPS, effective TPS (deduped), instructions per block.
- Fork rate: conflicting branches per time window.
- Resource utilization: CPU/memory/bandwidth/disk.
-
Robustness: graceful degradation under loss/jitter/partition/byzantine
injections.
5.2 Testbed & Methods
-
Geodistributed topologies across AMER/EU/APAC/OCE; RTT tiers 10/50/100
ms.
-
Workloads: payments (small, frequent), DEX (match/clear mix), NFT
(bursty), AI micropayments (steady drizzle).
- Ablation: Rotor off vs. on; Votor fast vs. slow path.
-
Weekly open reports with raw data, scripts, and reproducible images.
6. Token Economics (aSOLX)
6.1 Basics
-
Utility: gas, staking/security, governance, ecosystem settlement.
-
Total supply: 10,000,000,000 (inferred from 20% staking emission
totals; subject to TGE announcement).
6.2 Allocation & Vesting
Bucket |
Share |
Vesting / Lockup |
Notes |
Public Sale |
8% |
Per announcement; liquid by TGE rules |
Self‑custody; contract delivery |
DEX Liquidity |
2% |
LP lock ≥ 12 months |
Initial pool & price discovery |
Treasury/Reserve |
10% |
Multisig governed |
R&D/marketing/partnerships |
Strategic |
10% |
6‑month cliff + 24‑month linear |
|
Team & Advisors |
15% |
12‑month cliff + 36‑month linear |
|
Staking Rewards |
20% |
4‑year emissions |
Network security & delegation |
Ecosystem/Foundation |
20% |
≥ 12‑month linear |
Developers & incubation |
Community Airdrop |
15% |
Programmatic distribution |
Anti‑Sybil & task‑linked |
6.3 Staking Emission Function (4 years)
Total emissions R = 2,000,000,000
over four years with
coefficients {0.35, 0.30, 0.20, 0.15}. Year‑t allocation:
R_t = R × c_t
; weekly split
E_week = R_t / W_t
; per‑block
E_block = R_t / B_t
.
6.4 Validator Reward Weights
Effective weight w_i = s_i^β · q_i
, β ∈ (0,1], where
s_i
is stake and q_i
is a quality factor (0–1)
based on uptime, double‑sign penalties, and vote latency percentiles.
Period reward: reward_i = E_period · w_i / Σ w_j
.
6.5 Fee Adjustment (EIP‑1559‑like, Proposal)
BaseFee_{k+1} = BaseFee_k · exp[ α · (ρ_k − ρ*) ]
, where ρ*
is target utilization and α tunes responsiveness. Excess fees may be
burned to offset inflation.
6.6 Slashing & Penalties
-
Double‑signing: slash
min(0.05 · s_i, s_i^{crit})
; extend
unbonding windows.
- Offline: tiered penalties by windowed uptime.
- Late voting: reduce
q_i
(lowers future rewards).
7. Sale & Distribution
7.1 Public Sale
-
Self‑custody wallets (Phantom/Solflare, etc.) send to the official
sale address; contracts deliver aSOLX back to the sender.
- Accepted assets: SOL/USDT/USDC (per announcement and contract).
-
Two coexisting price wordings (to be unified): A) unit price 0.01
USDT/USDC; min 0.5 SOL or 100 USDT/USDC. B) 1 SOL = 1,000 aSOLX; min 1
SOL; max 50 SOL. Final public wording must match the contract’s
exchange ratio.
-
Warning: do not send from centralized exchanges
directly to the sale address.
7.2 Community Airdrop S1 (Anti‑Sybil)
-
Daily check‑in: 10 aSOLX; referral: +20 aSOLX to both inviter and
invitee.
-
Redemption threshold: ≥ 100 aSOLX; upon approval, T+1 settlement to
the bound address.
-
Abuse controls: device fingerprinting, IP/ASN rate limits, behavioral
anomaly scoring, and on‑chain clustering—one‑strike disqualification.
7.3 Compliance & Regions
- Enable KYC/AML as required by partners/jurisdictions.
-
Publish restricted regions and refund/appeal procedures on the
compliance page.
8. Governance (On‑chain)
8.1 Proposal & Voting
-
Stages: Deposit → Off‑chain discussion/temperature check → On‑chain
voting (with timelock) → Execution.
-
Thresholds: minimum deposit D_min; quorum q and passing threshold θ
(relative/absolute majority).
-
Voting power:
VP_i = s_i^γ
with γ ∈ [0.7, 1] to mitigate
centralization.
8.2 Emergency Mechanics
-
Emergency Brake: if monitored metrics exceed bounds (fork rate,
latency, offline rate) and the Security Council multisig confirms,
throttle rate, enable read‑only mode, or freeze proposals.
-
Timelock: ≥ 48 h for major treasury moves or parameter increases.
9. Security
-
Audit pathway: contracts → client critical paths → network layer.
-
Formal methods: Votor locking/commit safety proofs; Rotor
bounded‑delay and congestion analysis.
-
Bug bounty: tiered bounties (P0/P1/P2) requiring PoC and impact
assessment.
-
Key management & custody: validators are advised to use
HSM/threshold‑signing; multisig 3/5 or 4/7.
10. Cross‑Chain & Ecosystem
-
Bridges: integrate leading message/verification bridges (light‑client
or guardian‑network backed). Governance caps daily limits and rates.
-
Oracles: dual‑home with median aggregation and outlier rejection.
-
AI/DePIN: micropayment settlement, data‑availability proofs, task
marketplaces, and reputation systems.
11. Roadmap (Measurable Milestones)
-
2025‑07: Devnet “Sunrise”. SDK/Explorer/Faucet. Public p95 latency,
TPS, and fork‑rate reporting.
-
2025‑08: Incentivized testnet “Dawn”. Votor‑P2, Rotor‑P1. Validator
program & points; reservation‑only sale (no selling).
-
2025‑09: Public sale launch + Airdrop S1. Compliance & risk pages;
weekly performance reports.
-
2025‑10: Mainnet Beta + TGE + initial DEX LP (lock ≥ 12 months).
Governance live: multisig + timelock.
-
2025‑11: Ecosystem integrations (stables, oracles). Advance CEX
listings and PoR engagements.
-
2025‑Q4: Rotor GA (v1.x). Target p95 ≈ 150 ms. Cross‑chain bridge MVP.
-
2026‑Q1: Global nodes & hackathons. Enterprise pilots
(RWA/payments/IoT).
12. Transparency & Status Page
-
Publish treasury/lock addresses, vesting calendar, LP lock
transactions, audits & PoR, weekly raw data and scripts.
-
Flag anomalies on the status page (fork rate, latency, offline rate).
13. Risk Factors & Disclosures
- Market risk: price volatility and liquidity constraints.
-
Technical risk: implementation defects, unexpected network behavior,
bridge security incidents.
- Compliance risk: regulatory changes across jurisdictions.
-
Operational risk: validator concentration, governance apathy causing
parameter imbalance.
Disclaimer: This is not investment advice. Participants must evaluate
risks and comply with local laws.
14. Glossary (Selected)
- QC: Quorum Certificate (≥ 2/3 voting power aggregated).
- Lock: lock on the highest QC to avoid unsafe rollbacks.
- Pacemaker: timeout & round‑advance mechanism.
- CU: Compute Units, the metered unit of on‑chain computation.
15. Appendix
15.1 Parameter Sheet (Draft)
- Target p95 finality: 100–150 ms.
- Neighbor set size K: 4–8 (adaptive).
- Fee responsiveness α: 1e‑3–1e‑2.
- Staking weight β: 0.8–0.9; voting‑power exponent γ: 0.7–1.0.
- Timelock: ≥ 48 h. Multisig: 3/5 or 4/7.
15.2 Votor Pseudocode (Snippet)
OnReceiveBlock(b_t):
if not ConflictsWithLock(b_t):
vote ← Sign(b_t, round = t)
SendToLeader(vote)
OnQC(QC_t):
if Adjacent(QC_{t-1}, QC_t) and LockedOn(QC_{t-1}):
Commit(Parent(b_t))
else:
UpdateLock(QC_t)
15.3 TBF (To‑Be‑Finalized) List
-
Unify public sale pricing/threshold wording across contract, FAQ, and
frontend.
-
Explicitly state total supply (10B) and publish verifiable on‑chain
lock/cliff lists.
- Publish audit and formal‑verification links.
- Launch status page and weekly performance portal.
-
Compliance page: restricted regions, refund and appeal procedures.