Published April 26, 2026 | Version Verum Node OS: A Proof of Concept for Extreme NVMe I/O Optimization (Kernel 7.1)
Computational notebook Open

Verum Node OS: A Proof of Concept for Extreme NVMe I/O Optimization (Kernel 7.1) plus Kernel 8.0.0 (Verum Node OS Edition)

  • 1. ROR icon The European Institute for Crime Prevention and Control, affiliated with the United Nations

Description

Title: Verum Node Wrecked Machines: A Comprehensive Proof of Concept in Extreme NVMe I/O Optimization, Zero-Telemetry Architecture, and Data Sovereignty (Kernel 7.1) Version: 1.0 (Kernel 7.1-verum-wrecked) + Kernel 7.1-verum-wrecked8 (Resource Type: Software / Operating System Language: English

Description:

1. Executive Summary & Project Philosophy The Verum Node Operating System represents a paradigm shift in custom Linux distribution design, prioritizing absolute data sovereignty, uncompromised privacy by design, and extreme hardware optimization. Born from the necessity to overcome the inherent bloat, generalized scheduling, and pervasive telemetry found in upstream, commercially oriented operating systems, Verum Node acts as a highly specialized computational engine. It proves that by strictly tailoring the Linux kernel to specific modern hardware architectures (specifically AMD Ryzen processors and high-speed NVMe storage), a system can achieve an unparalleled leap in Input/Output Operations Per Second (IOPS) while simultaneously closing off background data-harvesting vectors.

2. Kernel-Level Engineering & The "Wrecked" Architecture At the heart of the Verum Node OS runs the 7.1-verum-wrecked and 8.0.0 kernel, a heavily modified, compiled-from-source Linux tree. Unlike generic kernels that ship with broad hardware support and heavy background monitoring, the Verum Wrecked configuration utilizes a highly aggressive, minimalist Kconfig profile.

Key architectural modifications include:

  • Eradication of Monitoring Overhead: Subsystems such as AUDIT, TASKSTATS, and FTRACE_SYSCALLS have been completely disabled. This eliminates the CPU cycles traditionally wasted on logging, tracing, and reporting system calls, returning that processing power directly to the user space.

  • CPU and Scheduling Optimizations: The kernel forces RCU_NOCB (Read-Copy-Update No Callbacks) to offload RCU callbacks from performance-critical CPUs, drastically reducing latency spikes and operating system jitter. Combined with a customized PREEMPT model, the kernel achieves a highly responsive state tailored for real-time data throughput.

  • I/O Stack Bypass: The storage stack is engineered to leverage modern asynchronous I/O frameworks, primarily io_uring, bypassing legacy block layer bottlenecks and allowing the NVMe controller to communicate almost directly with the processor's threads without blocking.

3. Empirical Validation: Bare Metal Phoronix Benchmarks The architectural choices within the Verum Wrecked kernel have been rigorously validated on bare metal using the industry-standard Phoronix Test Suite (Flexible IO Tester). Tested on an AMD Ryzen 5 7520U architecture paired with a Kingston SNV3S500G NVMe SSD, the system delivered unprecedented stability and speed compared to standard distributions on identical hardware:

  • Random Read (4KB, io_uring): Sustained throughput of 932 MB/s pushing an extraordinary 238,333 IOPS. The true triumph of this architecture is its mathematical consistency, maintaining a standard deviation of only 0.9% under severe stress.

  • Random Write (4KB, io_uring): Sustained throughput of 788 MB/s achieving 201,667 IOPS, with a standard deviation of just 1.1%.

These metrics effectively demonstrate that stripping the kernel of generic telemetry and optimizing the I/O scheduler can yield up to double the IOPS output of standard debian-based generic distributions.

4. Data Sovereignty and Enterprise Applicability Beyond sheer performance, Verum Node OS is fundamentally engineered for privacy. By severing ties with conventional Big Tech telemetry frameworks at the kernel level, it offers a "Privacy by Design" foundation. This makes the underlying architecture highly relevant for sectors requiring strict data compliance (such as legal tech, healthcare, and corporate environments adhering to rigid data protection laws), ensuring that localized data processing remains strictly local.

 

 

Files

Avctoris.pdf

Files (8.4 GB)

Name Size Download all
md5:edb2562bb47217f42b94bd82d2614c79
165.7 kB Preview Download
md5:541aca3dff02cb92c46a1d75f1bad8c3
1.8 MB Preview Download
md5:d3b7d0c1332c430d62535bf318ec8435
4.2 GB Download
md5:98826585c5ce29e08c1343251fcd326a
4.2 GB Download

Additional details

Funding

Carlsberg Foundation
Security and Privacy of Distributed Software Systems CF23-0791

Dates

Available
2025-04-26
"A performance-driven Linux distribution featuring a custom 7.1 kernel optimized for extreme NVMe I/O throughput (~238k IOPS) and zero-telemetry sovereignty."

Software

Repository URL
https://doi.org/10.1051/0004-6361/202451959
Programming language
Linux Kernel Module
Development Status
Active

References

  • NASA Astrophysics Data System