Machine learning spots RISC-V hardware attacks from processor performance counters
Researchers built the first public RISC-V dataset combining hardware performance counters with instruction traces and showed that simple tree-based models ...
Researchers built the first public RISC-V dataset combining hardware performance counters with instruction traces and showed that simple tree-based models ...
An experimental graphene optical system uses red, green and blue light challenges, transformer-assisted image analysis and binary codes to create ...
A preliminary AES experiment found that bioinformatics-inspired alignment recovered keys with up to 44% fewer power traces than elastic alignment, ...
A new review argues that memcapacitors and meminductors, the energy-storing cousins of the memristor, could provide fundamentally stronger hardware security ...
Rice University physicists have theoretically shown that autferroic materials with a 'seesaw' magnetoelectric mechanism can generate true random numbers over ...
George Mason University researchers are designing and defending against microarchitectural GPU attacks that exploit power oversubscription to destabilize multi-tenant AI ...
Researchers have developed SeqCube-HT, a framework that uses trace-backed state-cube graphs to generate verifiably executable sequential hardware Trojan benchmarks far ...
Researchers at the Universitat Politècnica de Catalunya show that on-chip hardware performance counters combined with machine learning can detect microarchitectural ...
© 2025 Scienmag - Science Magazine
© 2025 Scienmag - Science Magazine