mirror of
https://github.com/hanzoai/papers.git
synced 2026-10-22 17:20:17 +00:00
Add section 23 with literature-backed comparison scaffolds for: - AI agent frameworks: vs LangGraph, AutoGen, CrewAI, OpenAI Agents SDK - LLM inference engines: vs vLLM, TGI, TensorRT-LLM Hanzo column entries marked [TBD-measure] until a controlled re-run on the competitor's published harness is completed. Competitor cells sourced to repos, papers, and third-party benchmark articles. Reproducibility spec (hardware, models, software pins, decision rule) documents what is needed to fill the [TBD] cells honestly. No invented numbers; existing internal claims in sec 17/16 retained but flagged in section 23 as not-yet-apples-to-apples vs literature.
129 lines
5.8 KiB
TeX
129 lines
5.8 KiB
TeX
\documentclass[11pt,twocolumn]{article}
|
|
\usepackage{../shared/hanzocover}
|
|
\usepackage[utf8]{inputenc}
|
|
\usepackage[T1]{fontenc}
|
|
\usepackage{amsmath,amssymb}
|
|
\usepackage{graphicx}
|
|
\usepackage{booktabs}
|
|
\usepackage{hyperref}
|
|
\usepackage{xcolor}
|
|
\usepackage{listings}
|
|
\input{../shared/lstlang}
|
|
\usepackage{tikz}
|
|
\usetikzlibrary{shapes,arrows,positioning,fit,calc}
|
|
\usepackage{multicol}
|
|
\usepackage{enumitem}
|
|
\usepackage{tabularx}
|
|
\usepackage{caption}
|
|
\usepackage{float}
|
|
\usepackage{microtype}
|
|
\usepackage{fancyhdr}
|
|
\usepackage{lastpage}
|
|
|
|
\definecolor{codegreen}{rgb}{0,0.6,0}
|
|
\definecolor{codegray}{rgb}{0.5,0.5,0.5}
|
|
\definecolor{codepurple}{rgb}{0.58,0,0.82}
|
|
\definecolor{backcolour}{rgb}{0.95,0.95,0.92}
|
|
\definecolor{hanzoaccent}{HTML}{6366F1}
|
|
\definecolor{luxgold}{HTML}{D4A843}
|
|
\definecolor{wincolor}{HTML}{16A34A}
|
|
\definecolor{losecolor}{HTML}{DC2626}
|
|
\definecolor{tiecolor}{HTML}{D97706}
|
|
|
|
\lstdefinestyle{mystyle}{
|
|
backgroundcolor=\color{backcolour},
|
|
commentstyle=\color{codegreen},
|
|
keywordstyle=\color{codepurple},
|
|
numberstyle=\tiny\color{codegray},
|
|
stringstyle=\color{codegreen},
|
|
basicstyle=\ttfamily\footnotesize,
|
|
breakatwhitespace=false,
|
|
breaklines=true,
|
|
captionpos=b,
|
|
keepspaces=true,
|
|
numbers=left,
|
|
numbersep=5pt,
|
|
showspaces=false,
|
|
showstringspaces=false,
|
|
showtabs=false,
|
|
tabsize=2
|
|
}
|
|
\lstset{style=mystyle}
|
|
|
|
% Compact lists
|
|
\setlist{nosep,leftmargin=*}
|
|
|
|
% Custom commands for verdicts
|
|
\newcommand{\win}[1]{\textcolor{wincolor}{\textbf{#1}}}
|
|
\newcommand{\lose}[1]{\textcolor{losecolor}{#1}}
|
|
\newcommand{\tie}[1]{\textcolor{tiecolor}{#1}}
|
|
\newcommand{\yes}{\textcolor{wincolor}{\checkmark}}
|
|
\newcommand{\no}{\textcolor{losecolor}{$\times$}}
|
|
\newcommand{\pmark}{\textcolor{tiecolor}{$\sim$}}
|
|
|
|
% Header/footer
|
|
\pagestyle{fancy}
|
|
\fancyhf{}
|
|
\fancyhead[L]{\footnotesize\textit{Hanzo + Lux Stack Comparison}}
|
|
\fancyhead[R]{\footnotesize\textit{April 2026}}
|
|
\fancyfoot[C]{\footnotesize\thepage\ / \pageref{LastPage}}
|
|
\renewcommand{\headrulewidth}{0.4pt}
|
|
\renewcommand{\footrulewidth}{0pt}
|
|
|
|
\title{The Hanzo + Lux Stack: A Comparative Analysis of AI-Native Blockchain Infrastructure}
|
|
\author{Zach Kelling\thanks{[email protected]} \\ \textit{Hanzo Industries (Techstars '17)} \\ \texttt{[email protected]}}
|
|
\date{March 2026}
|
|
|
|
\begin{document}
|
|
|
|
\hanzocoverpage
|
|
|
|
\twocolumn[
|
|
\begin{@twocolumnfalse}
|
|
\maketitle
|
|
\begin{abstract}
|
|
We present a comprehensive comparative analysis of the Hanzo + Lux infrastructure stack against incumbent and emerging alternatives across sixteen technology categories spanning identity, API routing, application runtime, secrets management, object storage, vector search, analytics, payments, consensus, post-quantum cryptography, multi-party computation, fully homomorphic encryption, GPU acceleration, wire protocols, AI inference, and AI agent orchestration. For each category, we provide feature matrices, performance benchmarks, integration cost estimates, and total cost of ownership projections. The Hanzo + Lux stack achieves a unique position as the only fully integrated, open-source, post-quantum-ready platform for building AI-native blockchain applications. Where individual components may match or exceed specialized alternatives on isolated metrics, the integrated stack eliminates approximately 340 integration points, reduces operational surface area by an estimated 72\%, and provides cryptographic capabilities---including FIPS~203/204/205 compliance, TFHE-based confidential computing, and CGGMP21+FROST threshold signatures---that no competing stack offers as a unified product. The ZAP wire protocol demonstrates 4.2~million calls per second with zero heap allocations, and the Quasar consensus engine achieves 180,000 transactions per second with 420ms finality under adversarial conditions. We conclude that for organizations building at the intersection of AI and blockchain, the integrated Hanzo + Lux stack represents a 3--5$\times$ reduction in total engineering effort compared to assembling equivalent functionality from point solutions.
|
|
\end{abstract}
|
|
\vspace{1em}
|
|
\end{@twocolumnfalse}
|
|
]
|
|
|
|
\tableofcontents
|
|
|
|
%% ============================================================================
|
|
|
|
\input{sections/01-introduction}
|
|
\input{sections/02-identity-hanzo-iam}
|
|
\input{sections/03-api-gateway-hanzo-gateway}
|
|
\input{sections/04-runtime-hanzo-base}
|
|
\input{sections/05-secrets-hanzo-kms}
|
|
\input{sections/06-object-storage-hanzo-storage}
|
|
\input{sections/07-vector-search-hanzo-vector}
|
|
\input{sections/08-analytics-hanzo-insights}
|
|
\input{sections/09-payments-hanzo-commerce}
|
|
\input{sections/10-consensus-quasar}
|
|
\input{sections/11-cryptography-lux-crypto}
|
|
\input{sections/12-mpc-lux-mpc}
|
|
\input{sections/13-fhe-lux-fhe}
|
|
\input{sections/14-gpu-acceleration-lux-accel}
|
|
\input{sections/15-wire-protocol-zap}
|
|
\input{sections/16-ai-inference-hanzo-engine}
|
|
\input{sections/17-ai-agents-hanzo-agent-mcp}
|
|
\input{sections/18-the-integration-advantage}
|
|
\input{sections/19-unified-deployment-model}
|
|
\input{sections/20-unique-capabilities}
|
|
\input{sections/21-limitations-and-honest-assessment}
|
|
\input{sections/22-decision-framework}
|
|
\input{sections/23-named-competitor-scaffold}
|
|
\input{sections/99-conclusion}
|
|
|
|
\onecolumn
|
|
\section*{Reproducibility}
|
|
\textbf{Source code.} \texttt{github.com/hanzoai/platform} (commit \texttt{TBD}); component repos: \texttt{hanzoai/iam}, \texttt{hanzoai/gateway}, \texttt{hanzoai/base}, \texttt{hanzoai/kms}, \texttt{luxfi/crypto}, \texttt{luxfi/mpc}, \texttt{luxfi/fhe}.\\
|
|
\textbf{Build.} Per-component \texttt{make build}; integrated K8s deploy via \texttt{kubectl kustomize universe/infra/k8s/ | kubectl apply -f -}.\\
|
|
\textbf{Hardware tested.} Linux amd64/arm64; GPU acceleration on Apple Metal (M2/M3) and NVIDIA CUDA (A100/H100).\\
|
|
\textbf{Standards referenced.} OIDC, W3C DID, FIPS 203 (ML-KEM), FIPS 204 (ML-DSA), S3 API, OpenAI API.\\
|
|
\textbf{Contact.} [email protected].
|
|
|
|
\end{document}
|