mirror of
https://github.com/lux-apps/papers.git
synced 2026-11-02 11:39:52 +00:00
zeekay
98343d238e
Quasar Horizon (consensus composition over Pulsar / ML-DSA / BLS lanes
under Prism + LSS lifecycle), Warp 2.0 (Pulse-backed cross-chain envelope
with forward-only v1->v2 migration and lattigo ReadUint64Slice DoS fix),
benchmark report (microbench, reshare, activation, Warp 2.0 verify, Prism
composition, GPU headroom, cross-WAN latency), and NIST IR 8214C-style
submission package (spec, parameter sets, domain separation, serialization,
reference-impl pointers, evaluation report, security analysis, side-channel
considerations).
All four deliverables compile cleanly via pdflatex+bibtex and reference
the canonical proof bucket files at proofs/{definitions,pulsar,lss,quasar}.
No duplication of proof bodies; cite by file path.
40 lines
1.9 KiB
TeX
40 lines
1.9 KiB
TeX
\section{Conclusion}
|
|
|
|
\label{sec:conclusion}
|
|
%% ============================================================================
|
|
|
|
QuasarSTM~3.0 is the third generation of Block-STM and the first to
|
|
treat the GPU as a primary execution surface rather than a faster
|
|
CPU. Block-STM~1.0 proved that a known canonical order is a
|
|
validation primitive, not a serialisation bottleneck. Block-STM~2.0
|
|
ported that idea to the GPU. QuasarSTM~3.0 re-architects the entire
|
|
fabric around the lane abstraction: contention is predicted by
|
|
hint, prevented by Prism refraction, validated in three tiers
|
|
(lane-clock, key-MVCC, semantic), repaired incrementally against
|
|
fiber checkpoints, managed deterministically by a six-policy
|
|
contention manager with explicit hard limits, and finalised in
|
|
batches by commit horizons.
|
|
|
|
The headline result is that Block-STM stops being a repair loop and
|
|
becomes a serializability fabric. Production targets of
|
|
$\mathit{repair\_amplification} < 1.01$ on disjoint-lane workloads
|
|
and controlled degradation under adversarial Zipfian contention are
|
|
achieved by structure, not by tuning.
|
|
|
|
The four invariants specified in \S\ref{sec:correctness} ---
|
|
determinism, ordered serializability, repair monotonicity, and
|
|
horizon safety --- are the contract this fabric meets with the
|
|
Quasar consensus engine. The 1024-tx end-to-end stress in
|
|
${\sim}150\,\text{ms}$ on M1~Max and the textbook 16-same-key
|
|
contention test (120 conflicts, 120 repairs, 16 commits) demonstrate
|
|
the contract holds at production scale on commodity hardware.
|
|
|
|
The QuasarSTM~3.0 specification was frozen on December~15, 2025
|
|
and activated on the Quasar~3.0 mainnet on December~25, 2025.
|
|
The adaptive-evolution path (3.1 through 4.0) is documented in the
|
|
living LP-010~\cite{lp010-quasarstm} as ongoing research. The 3.0
|
|
specification stands on its own.
|
|
|
|
|
|
%% ============================================================================
|