arXiv subjects

Fabian Schwarz

Publications and source records attributed to Fabian Schwarz.

5 recordsLinked to original sources

arXiv2026

WarpGuard: Towards Control-Flow Attestation for Heterogeneous CPU-GPU Execution

Heterogeneous CPU-GPU workloads are increasingly used in safety-critical embedded systems, yet no existing approach provides joint attestation of their execution. Prior Control-Flow Attestation (CFA) techniques focus on CPU-side CFA, while GPU attestation is limited to static, load-time verification and does not provide runtime guarantees. As a result, runtime attacks on GPU kernels and violations of the CPU-GPU interaction contract remain unaddressed. We present WarpGuard, the first composite CFA framework for heterogeneous CPU-GPU workloads. WarpGuard verifies execution against a unified control-flow graph (CFG) that captures both CPU and GPU components. It extends prior CFA techniques in two ways: it enables runtime CFA of GPU kernels by tracing their execution against kernel-specific CFGs, and it monitors kernel launch events and enforces per-call site policies to detect violations at the CPU-GPU boundary. These extensions address challenges arising from GPU parallelism and cross-device interactions. We implement WarpGuard using software-based instrumentation, requiring no specialized hardware or binary modifications. Our evaluation on an NVIDIA Jetson Orin Nano shows that WarpGuard detects GPU-side control-flow and cross-boundary attacks. Across microbenchmarks, SPECAccel, and eight TensorRT inference workloads, WarpGuard incurs moderate overheads, suggesting practicality for embedded safety-critical settings.

cs.CR↗

arXiv2025

pylimer-tools: A Python Package for Generating and Analyzing Bead-Spring Polymer Networks

The Python package pylimer-tools is a comprehensive toolkit for computational studies of polymer networks, particularly bead-spring networks. The package provides functionality to generate polymer networks using Monte Carlo (MC) procedures and analyze their structural and mechanical properties. Key features include detection of loops, reduction of the network to its ground state energy both with and without entanglements by the Force Balance procedure, and thereafter computing the soluble and dangling fractions of network strands, as well as the equilibrium shear modulus. The toolkit supports analysis of structures generated both internally and by external simulation software such as LAMMPS. The package implements theoretical frameworks including Miller-Macosko theory and provides a dissipative particle dynamics (DPD) simulator with slip-spring entanglement modeling. Built with C++ for performance and exposed through Python bindings, pylimer-tools addresses the need for specialized tools in computational polymer science.

cond-mat.soft↗

arXiv2025

Composition Effects on Ni/Al Reactive Multilayers: A Comprehensive Study of Mechanical Properties, Reaction Dynamics and Phase Evolution

Ni/Al reactive multilayers are promising materials for applications requiring controlled local energy release and superior mechanical performance. This study systematically investigates the impact of compositional variations, ranging from 30 to 70 at.% Ni, and bilayer thicknesses (30 nm and 50 nm) on the mechanical properties and reaction dynamics of Ni/Al multilayers. Multilayers with varying Ni-to-Al ratios were fabricated and subjected to instrumented nanoindentation testing to evaluate hardness and elastic modulus. Combustion experiments, conducted on dogbone-shaped multilayers deposited onto silicon wafers with thermal barrier coatings, characterized the reaction front's speed, temperature, and the resulting phases. The findings revealed that composition variations within this range enable precise tuning of reaction speed and temperature without significant changes in mechanical properties, while deviations in modulus and hardness at higher nickel concentrations suggest microstructural influences. Notably, phase formation in Al-rich samples deviated from equilibrium predictions, highlighting the role of kinetic factors, such as diffusion and rapid quenching, in driving non-adiabatic processes during phase evolution. Molecular dynamics simulations provided complementary atomistic insights into mechanical responses and reaction kinetics, bridging experimental observations with theoretical predictions. This integrated approach advances the understanding of Ni/Al multilayers, offering a framework for optimizing their composition and structural design to achieve tailored performance for application-specific requirements.

cond-mat.mtrl-sci↗

arXiv2024

KeyVisor -- A Lightweight ISA Extension for Protected Key Handles with CPU-enforced Usage Policies

The confidentiality of cryptographic keys is essential for the security of protection schemes used for communication, file encryption, and outsourced computation. Beyond cryptanalytic attacks, adversaries can steal keys from memory via software exploits or side channels, enabling them to, e.g., tamper with secrets or impersonate key owners. Therefore, existing defenses protect keys in dedicated devices or isolated memory, or store them only in encrypted form. However, these designs often provide unfavorable tradeoffs, sacrificing performance, fine-grained access control, or deployability. In this paper, we present KeyVisor, a lightweight ISA extension that securely offloads the handling of cryptographic keys to the CPU. KeyVisor provides CPU instructions that enable applications to request protected key handles and perform AEAD cipher operations on them. The underlying keys are accessible only by KeyVisor, and thus never leak to memory. KeyVisor's direct CPU integration enables fast crypto operations and hardware-enforced key usage restrictions, e.g., keys usable only for de-/encryption, with a limited lifetime, or with a process binding. Furthermore, privileged software, e.g., the monitor firmware of TEEs, can revoke keys or bind them to a specific process/TEE. We implement KeyVisor for RISC-V based on Rocket Chip, evaluate its performance, and demonstrate real-world use cases, including key-value databases, automotive feature licensing, and a read-only network middlebox.

cs.CR↗

arXiv2022

Deformation-induced topological transitions in mechanical metamaterials and their application to tunable non-linear stiffening

Mechanical metamaterials are periodic lattice structures with complex unit cell architectures that can achieve extraordinary mechanical properties beyond the capability of bulk materials. A new class of metamaterials is proposed, whose mechanical properties rely on deformation-induced transitions in nodal-topology by formation of internal self-contact. The universal nature of the principle presented, is demonstrated for tension, compression, shear and torsion. In particular, it is shown that by frustration of soft deformation modes, large highly non-linear stiffening effects can be generated. Tunable non-linear elasticity can be exploited to design materials mimicking the complex mechanical response of biological tissue.

physics.app-ph↗