MESI-LCG bc8j: Zero-OS-Dependency ECDSA for Hardware Security Modules, Hardware Wallets, and IoT — +21% vs libsecp256k1 Aggregate Throughput, 84x Less Memory
Fuente:
Zenodo
Salvato in:
| Autore principale: | |
|---|---|
| Natura: | Recurso digital |
| Lingua: | inglese |
| Pubblicazione: |
Zenodo
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866901672296448000 |
|---|---|
| author | Pirolo, Andrés Sebastián |
| author_facet | Pirolo, Andrés Sebastián |
| contents | <p> </p> <h2><strong>Description</strong></h2> <p><strong>The Last Bottleneck in Trustless Blockchain Infrastructure — Solved.</strong></p> <p>The industry standard for ECDSA signing (<strong>libsecp256k1</strong>) relies heavily on architecture-specific <em>Assembly</em> and large precomputed tables (<strong>86KB</strong>) to achieve its speed. <strong>bc8j takes the opposite approach:</strong> absolute portability, physical L1-cache residency, and zero Operating System trust, while actually beating the standard in aggregate throughput.</p> <p>If you are building <em>Hardware Wallets, HSMs, IoT infrastructure,</em> or <em>Layer-2 Sequencers</em>, <strong>bc8j</strong> delivers enterprise-scale cryptographic execution without the side-channel attack surfaces inherent to RAM access and OS entropy dependencies.</p> <h2><strong>The Benchmark: bc8j vs libsecp256k1</strong></h2> <p><em>(Measured on identical Snapdragon ARM hardware)</em></p> <ul> <li> <p><strong>The Aggregate Knockout (+21% Speed):</strong><br>bc8j natively scales across all cores, achieving <strong>58,196 sigs/sec (8 threads)</strong>.<br>libsecp256k1 is single-threaded by design and maxes out at <strong>48,077 sigs/sec</strong>.</p> </li> <li> <p><strong>The Micro-Footprint (84x Less Memory):</strong><br>In a strict <em>1-thread vs 1-thread</em> comparison, bc8j achieves <strong>12,054 sigs/sec</strong> (<em>25% the speed of libsecp256k1</em>) but does so using only <strong>1KB of memory</strong> versus <strong>86KB</strong> precomputed tables.</p> </li> <li> <p><strong>Zero OS Trust:</strong><br>Nonces are generated via physical <em>MESI cache-coherence jitter</em>.<br>No <em>/dev/urandom</em>. No kernel calls. <em>CVE-2013-7372</em> is architecturally impossible here.</p> </li> <li> <p><strong>Zero Assembly:</strong><br>Written in pure <strong>C++17</strong>. It compiles and runs at full speed on <em>ARM, x86-64, RISC-V</em>, or any future 64-bit architecture without a single line of porting effort.</p> </li> </ul> <h2><strong>Downloads & Licensing</strong></h2> <p>This repository contains the <em>executive summary</em>, <em>full security paper</em>, and <em>benchmark reports</em>.</p> <p>Implementation is licensed under <strong>PolyForm Noncommercial License 1.0.0</strong>.</p> <p><em>For commercial licensing (HSM, Hardware Wallet, IoT, and L2 infrastructure integration):</em><br><strong><a href="mailto:lctrnc1@gmail.com">lctrnc1@gmail.com</a></strong></p> <p> </p> |
| format | Recurso digital |
| id | zenodo_https___doi_org_10_5281_zenodo_19687701 |
| institution | Zenodo |
| language | eng |
| publishDate | 2026 |
| publisher | Zenodo |
| record_format | zenodo |
| spellingShingle | MESI-LCG bc8j: Zero-OS-Dependency ECDSA for Hardware Security Modules, Hardware Wallets, and IoT — +21% vs libsecp256k1 Aggregate Throughput, 84x Less Memory Pirolo, Andrés Sebastián secp256k1 ECDSA hardware wallet HSM IoT embedded cryptography Lightning Network Bitcoin libsecp256k1 <p> </p> <h2><strong>Description</strong></h2> <p><strong>The Last Bottleneck in Trustless Blockchain Infrastructure — Solved.</strong></p> <p>The industry standard for ECDSA signing (<strong>libsecp256k1</strong>) relies heavily on architecture-specific <em>Assembly</em> and large precomputed tables (<strong>86KB</strong>) to achieve its speed. <strong>bc8j takes the opposite approach:</strong> absolute portability, physical L1-cache residency, and zero Operating System trust, while actually beating the standard in aggregate throughput.</p> <p>If you are building <em>Hardware Wallets, HSMs, IoT infrastructure,</em> or <em>Layer-2 Sequencers</em>, <strong>bc8j</strong> delivers enterprise-scale cryptographic execution without the side-channel attack surfaces inherent to RAM access and OS entropy dependencies.</p> <h2><strong>The Benchmark: bc8j vs libsecp256k1</strong></h2> <p><em>(Measured on identical Snapdragon ARM hardware)</em></p> <ul> <li> <p><strong>The Aggregate Knockout (+21% Speed):</strong><br>bc8j natively scales across all cores, achieving <strong>58,196 sigs/sec (8 threads)</strong>.<br>libsecp256k1 is single-threaded by design and maxes out at <strong>48,077 sigs/sec</strong>.</p> </li> <li> <p><strong>The Micro-Footprint (84x Less Memory):</strong><br>In a strict <em>1-thread vs 1-thread</em> comparison, bc8j achieves <strong>12,054 sigs/sec</strong> (<em>25% the speed of libsecp256k1</em>) but does so using only <strong>1KB of memory</strong> versus <strong>86KB</strong> precomputed tables.</p> </li> <li> <p><strong>Zero OS Trust:</strong><br>Nonces are generated via physical <em>MESI cache-coherence jitter</em>.<br>No <em>/dev/urandom</em>. No kernel calls. <em>CVE-2013-7372</em> is architecturally impossible here.</p> </li> <li> <p><strong>Zero Assembly:</strong><br>Written in pure <strong>C++17</strong>. It compiles and runs at full speed on <em>ARM, x86-64, RISC-V</em>, or any future 64-bit architecture without a single line of porting effort.</p> </li> </ul> <h2><strong>Downloads & Licensing</strong></h2> <p>This repository contains the <em>executive summary</em>, <em>full security paper</em>, and <em>benchmark reports</em>.</p> <p>Implementation is licensed under <strong>PolyForm Noncommercial License 1.0.0</strong>.</p> <p><em>For commercial licensing (HSM, Hardware Wallet, IoT, and L2 infrastructure integration):</em><br><strong><a href="mailto:lctrnc1@gmail.com">lctrnc1@gmail.com</a></strong></p> <p> </p> |
| title | MESI-LCG bc8j: Zero-OS-Dependency ECDSA for Hardware Security Modules, Hardware Wallets, and IoT — +21% vs libsecp256k1 Aggregate Throughput, 84x Less Memory |
| topic | secp256k1 ECDSA hardware wallet HSM IoT embedded cryptography Lightning Network Bitcoin libsecp256k1 |
| url | https://doi.org/10.5281/zenodo.19687701 |