Showing 1–50 of 111 results for author: Scherer, M

xTern: Energy-Efficient Ternary Neural Network Inference on RISC-V-Based Edge Systems

Authors: Georg Rutishauser, Joan Mihali, Moritz Scherer, Luca Benini

Abstract: Ternary neural networks (TNNs) offer a superior accuracy-energy trade-off compared to binary neural networks. However, until now, they have required specialized accelerators to realize their efficiency potential, which has hindered widespread adoption. To address this, we present xTern, a lightweight extension of the RISC-V instruction set architecture (ISA) targeted at accelerating TNN inference… ▽ More Ternary neural networks (TNNs) offer a superior accuracy-energy trade-off compared to binary neural networks. However, until now, they have required specialized accelerators to realize their efficiency potential, which has hindered widespread adoption. To address this, we present xTern, a lightweight extension of the RISC-V instruction set architecture (ISA) targeted at accelerating TNN inference on general-purpose cores. To complement the ISA extension, we developed a set of optimized kernels leveraging xTern, achieving 67% higher throughput than their 2-bit equivalents. Power consumption is only marginally increased by 5.2%, resulting in an energy efficiency improvement by 57.1%. We demonstrate that the proposed xTern extension, integrated into an octa-core compute cluster, incurs a minimal silicon area overhead of 0.9% with no impact on timing. In end-to-end benchmarks, we demonstrate that xTern enables the deployment of TNNs achieving up to 1.6 percentage points higher CIFAR-10 classification accuracy than 2-bit networks at equal inference latency. Our results show that xTern enables RISC-V-based ultra-low-power edge AI platforms to benefit from the efficiency potential of TNNs. △ Less

Submitted 29 May, 2024; originally announced May 2024.

Comments: Accepted for publication at IEEE ASAP 2024

Testing trajectory-based determinism via time probability distributions

Authors: Matheus V. Scherer, Alexandre D. Ribeiro, Renato M. Angelo

Abstract: It is notorious that quantum mechanics (QM) cannot predict well-defined values for all physical quantities. Less well-known, however, is the fact that QM is unable to furnish probabilistic predictions even in emblematic scenarios such as the double-slit experiment. In contrast, equipped with postulate trajectories, Bohmian mechanics (BM) has inherited more predictive power. It follows that, contra… ▽ More It is notorious that quantum mechanics (QM) cannot predict well-defined values for all physical quantities. Less well-known, however, is the fact that QM is unable to furnish probabilistic predictions even in emblematic scenarios such as the double-slit experiment. In contrast, equipped with postulate trajectories, Bohmian mechanics (BM) has inherited more predictive power. It follows that, contrary to common belief, QM and BM are not just different interpretations but distinct theories. This work formalizes the aforementioned assertions and illustrates them through three case studies: (i) free particle, (ii) free fall under a uniform gravitational field, and (iii) the double-slit experiment. Specifically, we introduce a prescription for constructing a flight-time probability distribution within generic trajectory-equipped theories. We then apply our formalism to BM and derive probability distributions that are unreachable by QM. Our results can, in principle, be tested against real experiments, thereby assessing the validity of Bohmian trajectories. △ Less

Submitted 15 April, 2024; originally announced April 2024.

Comments: 8 pages, 4 figures. Comments are welcome

Optimizing the Deployment of Tiny Transformers on Low-Power MCUs

Authors: Victor J. B. Jung, Alessio Burrello, Moritz Scherer, Francesco Conti, Luca Benini

Abstract: Transformer networks are rapidly becoming SotA in many fields, such as NLP and CV. Similarly to CNN, there is a strong push for deploying Transformer models at the extreme edge, ultimately fitting the tiny power budget and memory footprint of MCUs. However, the early approaches in this direction are mostly ad-hoc, platform, and model-specific. This work aims to enable and optimize the flexible, mu… ▽ More Transformer networks are rapidly becoming SotA in many fields, such as NLP and CV. Similarly to CNN, there is a strong push for deploying Transformer models at the extreme edge, ultimately fitting the tiny power budget and memory footprint of MCUs. However, the early approaches in this direction are mostly ad-hoc, platform, and model-specific. This work aims to enable and optimize the flexible, multi-platform deployment of encoder Tiny Transformers on commercial MCUs. We propose a complete framework to perform end-to-end deployment of Transformer models onto single and multi-core MCUs. Our framework provides an optimized library of kernels to maximize data reuse and avoid unnecessary data marshaling operations into the crucial attention block. A novel MHSA inference schedule, named Fused-Weight Self-Attention, is introduced, fusing the linear projection weights offline to further reduce the number of operations and parameters. Furthermore, to mitigate the memory peak reached by the computation of the attention map, we present a Depth-First Tiling scheme for MHSA. We evaluate our framework on three different MCU classes exploiting ARM and RISC-V ISA, namely the STM32H7, the STM32L4, and GAP9 (RV32IMC-XpulpV2). We reach an average of 4.79x and 2.0x lower latency compared to SotA libraries CMSIS-NN (ARM) and PULP-NN (RISC-V), respectively. Moreover, we show that our MHSA depth-first tiling scheme reduces the memory peak by up to 6.19x, while the fused-weight attention can reduce the runtime by 1.53x, and number of parameters by 25%. We report significant improvements across several Tiny Transformers: for instance, when executing a transformer block for the task of radar-based hand-gesture recognition on GAP9, we achieve a latency of 0.14ms and energy consumption of 4.92 micro-joules, 2.32x lower than the SotA PULP-NN library on the same platform. △ Less

Submitted 3 April, 2024; originally announced April 2024.

Comments: Pre-print manuscript submitted for review to the IEEE Transactions on Computers

Siracusa: A 16 nm Heterogenous RISC-V SoC for Extended Reality with At-MRAM Neural Engine

Authors: Arpan Suravi Prasad, Moritz Scherer, Francesco Conti, Davide Rossi, Alfio Di Mauro, Manuel Eggimann, Jorge Tómas Gómez, Ziyun Li, Syed Shakib Sarwar, Zhao Wang, Barbara De Salvo, Luca Benini

Abstract: Extended reality (XR) applications are Machine Learning (ML)-intensive, featuring deep neural networks (DNNs) with millions of weights, tightly latency-bound (10-20 ms end-to-end), and power-constrained (low tens of mW average power). While ML performance and efficiency can be achieved by introducing neural engines within low-power systems-on-chip (SoCs), system-level power for nontrivial DNNs dep… ▽ More Extended reality (XR) applications are Machine Learning (ML)-intensive, featuring deep neural networks (DNNs) with millions of weights, tightly latency-bound (10-20 ms end-to-end), and power-constrained (low tens of mW average power). While ML performance and efficiency can be achieved by introducing neural engines within low-power systems-on-chip (SoCs), system-level power for nontrivial DNNs depends strongly on the energy of non-volatile memory (NVM) access for network weights. This work introduces Siracusa, a near-sensor heterogeneous SoC for next-generation XR devices manufactured in 16 nm CMOS. Siracusa couples an octa-core cluster of RISC-V digital signal processing cores with a novel tightly-coupled "At-Memory" integration between a state-of-the-art digital neural engine called N-EUREKA and an on-chip NVM based on magnetoresistive memory(MRAM), achieving 1.7x higher throughput and 3x better energy efficiency than XR SoCs using NVM as background memory. The fabricated SoC prototype achieves an area efficiency of 65.2 GOp/s/mm2 and a peak energy efficiency of 8.84 TOp/J for DNN inference while supporting complex heterogeneous application workloads, which combine ML with conventional signal processing and control. △ Less

Submitted 14 April, 2024; v1 submitted 22 December, 2023; originally announced December 2023.

Comments: Final accepted manuscript pre-print submitted to the IEEE Journal of Solid-State Circuits

The Nordic-walking mechanism and its explanation of deconfined pseudocriticality from Wess-Zumino-Witten theory

Authors: Bilal Hawashin, Astrid Eichhorn, Lukas Janssen, Michael M. Scherer, Shouryya Ray

Abstract: The understanding of phenomena falling outside the Ginzburg-Landau paradigm of phase transitions represents a key challenge in condensed matter physics. A famous class of examples is constituted by the putative deconfined quantum critical points between two symmetry-broken phases in layered quantum magnets, such as pressurised SrCu$_2$(BO$_3$)$_2$. Experiments find a weak first-order transition, w… ▽ More The understanding of phenomena falling outside the Ginzburg-Landau paradigm of phase transitions represents a key challenge in condensed matter physics. A famous class of examples is constituted by the putative deconfined quantum critical points between two symmetry-broken phases in layered quantum magnets, such as pressurised SrCu$_2$(BO$_3$)$_2$. Experiments find a weak first-order transition, which simulations of relevant microscopic models can reproduce. The origin of this behaviour has been a matter of considerable debate for several years. In this work, we demonstrate that the nature of the deconfined quantum critical point can be best understood in terms of a novel dynamical mechanism, termed Nordic walking. Nordic walking denotes a renormalisation group flow arising from a beta function that is flat over a range of couplings. This gives rise to a logarithmic flow that is faster than the well-known walking behaviour, associated with the annihilation and complexification of fixed points, but still significantly slower than the generic running of couplings. The Nordic-walking mechanism can thus explain weak first-order transitions, but may also play a role in high-energy physics, where it could solve hierarchy problems. We analyse the Wess-Zumino-Witten field theory pertinent to deconfined quantum critical points with a topological term in 2+1 dimensions. To this end, we construct an advanced functional renormalisation group approach based on higher-order regulators. We thereby calculate the beta function directly in 2+1 dimensions and provide evidence for Nordic walking. △ Less

Submitted 18 December, 2023; originally announced December 2023.

Comments: 12 + 11 pages (typeset with sn-jnl.cls), 4 + 1 figures; comments welcome

Spectra of typical Hilbert space operators

Authors: Marcel Scherer

Abstract: Let $\mathcal{B}(H)$ be the bounded, linear operators on a separable Hilbert space equipped with the norm topology. A property is called typical if the set of operators fulfilling the property is co-meager. We show that having non-empty continuous spectrum is a typical property and that the operators with non-empty point spectrum form a nowhere dense set. In addition we show that the set of operat… ▽ More Let $\mathcal{B}(H)$ be the bounded, linear operators on a separable Hilbert space equipped with the norm topology. A property is called typical if the set of operators fulfilling the property is co-meager. We show that having non-empty continuous spectrum is a typical property and that the operators with non-empty point spectrum form a nowhere dense set. In addition we show that the set of operators with empty point spectrum is dense and characterize the closure of the set of those operators for which the spectrum and the point spectrum coincide. △ Less

Submitted 27 October, 2023; originally announced October 2023.

Comments: 17 pages

MSC Class: 47A11 (Primary). 47A53 (Secondary)

Closed Loop Molecular Communication Testbed: Setup, Interference Analysis, and Experimental Results

Authors: Lukas Brand, Maike Scherer, Teena tom Dieck, Sebastian Lotter, Maximilian Schäfer, Andreas Burkovski, Heinrich Sticht, Kathrin Castiglione, Robert Schober

Abstract: In this paper, we present a fluid-based experimental molecular communication (MC) testbed that, similar to the human cardiovascular system, operates in a closed circuit tube system. The proposed system is designed to be biocompatible, resource-efficient, and controllable from outside the tube. As signaling molecule, the testbed employs the green fluorescent protein variant "Dreiklang" (GFPD). GFPD… ▽ More In this paper, we present a fluid-based experimental molecular communication (MC) testbed that, similar to the human cardiovascular system, operates in a closed circuit tube system. The proposed system is designed to be biocompatible, resource-efficient, and controllable from outside the tube. As signaling molecule, the testbed employs the green fluorescent protein variant "Dreiklang" (GFPD). GFPDs can be reversibly switched via light of different wavelengths between a bright fluorescent state and a less fluorescent state. Hence, this property allows for writing and erasing information encoded in the state of the GFPDs already present in the fluid via radiation from outside the tube. The concept of modulating the GFPDs existing in the channel at the transmitter for information transmission, instead of releasing new molecules, is a form of media modulation. In our testbed, due to the closed loop setup and the long experiment durations of up to 250 min, we observe new forms of inter-symbol interferences (ISI), which do not occur in short experiments and open loop systems. In particular, up to four different forms of ISI, namely channel ISI, inter-loop ISI, offset ISI, and permanent ISI, occur in the considered system. To mitigate inter-loop ISI and offset ISI, we propose a light based eraser unit. We experimentally demonstrate reliable information transmission in our testbed achieving error-free transmission of 500 bit at a data rate of 6 bit/min based on a sub-optimal low-complexity detection scheme. △ Less

Submitted 9 November, 2023; v1 submitted 24 October, 2023; originally announced October 2023.

Comments: 7 pages, 5 figures, 1 table. This work has been submitted for possible publication to the IEEE International Conference on Communications (ICC) 2024

Absence of SO(4) quantum criticality in Dirac semimetals at two-loop order

Authors: Max Uetrecht, Igor F. Herbut, Emmanuel Stamou, Michael M. Scherer

Abstract: Evidence for relativistic quantum criticality of antiferromagnetism and superconductivity in two-dimensional Dirac fermion systems has been found in large-scale quantum Monte Carlo simulations. However, the corresponding ($2+1$)-dimensional Gross--Neveu--Yukawa field theory with $N_f=2$ four-component Dirac fermions coupled to two triplets of order parameters does not exhibit a renormalization gro… ▽ More Evidence for relativistic quantum criticality of antiferromagnetism and superconductivity in two-dimensional Dirac fermion systems has been found in large-scale quantum Monte Carlo simulations. However, the corresponding ($2+1$)-dimensional Gross--Neveu--Yukawa field theory with $N_f=2$ four-component Dirac fermions coupled to two triplets of order parameters does not exhibit a renormalization group fixed point at one-loop order. Instead, the theory only features a critical point for a large or very small fractional number of fermion flavors $N_f$, which disappears for a broad range of flavor numbers around the physical case, $N_f=2$, due to fixed-point annihilation. This raises the question on how to explain the observed scaling collapse in the quantum Monte Carlo data. Here, we extend previous renormalization-group analyses by studying a generalized model at two-loop order in $4-ε$ spacetime dimensions. We determine the $ε$ correction to the upper and lower critical flavor numbers for the fixed-point annihilation and find that they both go towards the physical case $N_f=2$. However, this only happens very slowly, such that an extrapolation to $ε=1$ still suggests the absence of criticality in $2+1$ dimensions. Thereby, we consolidate the finding that the continuum field theory does not feature a stable renormalization-group fixed point and no true quantum criticality would be expected for the considered system. We briefly discuss a possible reconciliation in terms of a complex conformal field theory. Further, we also explore the fixed-point structure in an enlarged theory space and identify a candidate stable fixed-point solution. △ Less

Submitted 23 August, 2023; originally announced August 2023.

Comments: 13 pages, 6 figures, 2 tables

Report number: DO-TH 23/13

Journal ref: Physical Review B, vol. 108, 245130 (2023)

Analog of cosmological particle production in Dirac materials

Authors: Mireia Tolosa-Simeón, Michael M. Scherer, Stefan Floerchinger

Abstract: Two-dimensional van der Waals materials have recently been established experimentally as a highly-tunable condensed matter platform, facilitating the controlled manipulation of band structures and interactions. In several of these materials, Dirac cones are present in the low-energy regime near the Fermi level. Thus, fermionic excitations emerging in these materials close to the Dirac cones have a… ▽ More Two-dimensional van der Waals materials have recently been established experimentally as a highly-tunable condensed matter platform, facilitating the controlled manipulation of band structures and interactions. In several of these materials, Dirac cones are present in the low-energy regime near the Fermi level. Thus, fermionic excitations emerging in these materials close to the Dirac cones have a linear dispersion relation near the Fermi surface as massless relativistic Dirac fermions. Here, we study low-energy fermionic excitations of such Dirac materials in the presence of a mass gap that may be generated by symmetry breaking. Introducing a dynamical Fermi velocity and/or time-dependent mass gap for the Dirac quasiparticles, we exhibit the emergence of an analog of cosmological fermion pair production in terms of observables such as the expected occupation number or two-point correlation functions. We find that it is necessary and sufficient for quasiparticle production that only the ratio between the mass gap and the Fermi velocity is time-dependent. In this way, we establish that highly-tunable Dirac materials can serve as analog models for cosmological spacetime geometries, in particular, for Friedmann-Lemaître-Robertson-Walker expanding cosmologies. We briefly discuss possibilities for experimental realization △ Less

Submitted 20 May, 2024; v1 submitted 18 July, 2023; originally announced July 2023.

Comments: 15 pages, 4 figures

ITA: An Energy-Efficient Attention and Softmax Accelerator for Quantized Transformers

Authors: Gamze İslamoğlu, Moritz Scherer, Gianna Paulin, Tim Fischer, Victor J. B. Jung, Angelo Garofalo, Luca Benini

Abstract: Transformer networks have emerged as the state-of-the-art approach for natural language processing tasks and are gaining popularity in other domains such as computer vision and audio processing. However, the efficient hardware acceleration of transformer models poses new challenges due to their high arithmetic intensities, large memory requirements, and complex dataflow dependencies. In this work,… ▽ More Transformer networks have emerged as the state-of-the-art approach for natural language processing tasks and are gaining popularity in other domains such as computer vision and audio processing. However, the efficient hardware acceleration of transformer models poses new challenges due to their high arithmetic intensities, large memory requirements, and complex dataflow dependencies. In this work, we propose ITA, a novel accelerator architecture for transformers and related models that targets efficient inference on embedded systems by exploiting 8-bit quantization and an innovative softmax implementation that operates exclusively on integer values. By computing on-the-fly in streaming mode, our softmax implementation minimizes data movement and energy consumption. ITA achieves competitive energy efficiency with respect to state-of-the-art transformer accelerators with 16.9 TOPS/W, while outperforming them in area efficiency with 5.93 TOPS/mm$^2$ in 22 nm fully-depleted silicon-on-insulator technology at 0.8 V. △ Less

Submitted 10 July, 2023; v1 submitted 7 July, 2023; originally announced July 2023.

Comments: Accepted for publication at the 2023 ACM/IEEE International Symposium on Low Power Electronics and Design (ISLPED)

Deconfined quantum criticality lost

Authors: Menghan Song, Jiarui Zhao, Meng Cheng, Cenke Xu, Michael M. Scherer, Lukas Janssen, Zi Yang Meng

Abstract: Over the past two decades, the enigma of the deconfined quantum critical point (DQCP) has attracted broad attention across the condensed matter, quantum field theory, and high-energy physics communities, as it is expected to offer a new paradigm in theory, experiment, and numerical simulations that goes beyond the Landau-Ginzburg-Wilson framework of symmetry breaking and phase transitions. However… ▽ More Over the past two decades, the enigma of the deconfined quantum critical point (DQCP) has attracted broad attention across the condensed matter, quantum field theory, and high-energy physics communities, as it is expected to offer a new paradigm in theory, experiment, and numerical simulations that goes beyond the Landau-Ginzburg-Wilson framework of symmetry breaking and phase transitions. However, the nature of DQCP has been controversial. For instance, in the square-lattice spin-1/2 $J$-$Q$ model, believed to realize the DQCP between Néel and valence bond solid states, conflicting results, such as first-order versus continuous transition, and critical exponents incompatible with conformal bootstrap bounds, have been reported. The enigma of DQCP is exemplified in its anomalous logarithmic subleading contribution in its entanglement entropy (EE), which was discussed in recent studies. In the current work, we demonstrate that similar anomalous logarithmic behavior persists in a class of models analogous to the DQCP. We systematically study the quantum EE of square-lattice SU($N$) DQCP spin models. Based on large-scale quantum Monte Carlo computation of the EE, we show that for a series of $N$ smaller than a critical value, the anomalous logarithmic behavior always exists in the EE, which implies that the previously determined DQCPs in these models do not belong to conformal fixed points. In contrast, when $N\ge N_c$ with a finite $N_c$ that we evaluate to lie between $7$ and $8$, the DQCPs are consistent with conformal fixed points that can be understood within the Abelian Higgs field theory with $N$ complex components. △ Less

Submitted 31 March, 2024; v1 submitted 5 July, 2023; originally announced July 2023.

Comments: The revised version focuses on the anomalous logarithmic correction to the EE arising from the smooth boundary, rather than corners. And the critical $N_c$ is determined based on the anomalous EE signal from smooth boundary

Spectra of Quotient Modules

Authors: Michael Didas, Jörg Eschmeier, Michael Hartz, Marcel Scherer

Abstract: We determine the Taylor spectra of quotient tuples of the $d$-shift on Drury-Arveson spaces with finite-dimensional coefficient spaces. We show the the Taylor spectrum can be described in terms of the approximate zero set of the annihilator ideal, and in terms of the pointwise behavior of the inner multiplier associated with the quotient tuple. We determine the Taylor spectra of quotient tuples of the $d$-shift on Drury-Arveson spaces with finite-dimensional coefficient spaces. We show the the Taylor spectrum can be described in terms of the approximate zero set of the annihilator ideal, and in terms of the pointwise behavior of the inner multiplier associated with the quotient tuple. △ Less

Submitted 5 June, 2023; originally announced June 2023.

Comments: 11 pages

MSC Class: Primary 47A13; Secondary 47A10; 47A45

Spin-valley magnetism on the triangular moiré lattice with SU(4) breaking interactions

Authors: Lasse Gresista, Dominik Kiese, Simon Trebst, Michael M. Scherer

Abstract: The discovery of correlated insulating states in moiré heterostructures has renewed the interest in strongly-coupled electron systems where spin and valley (or layer) degrees of freedom are intertwined. In the strong-coupling limit, such systems can be effectively described by SU(4) spin-valley models akin to Kugel-Khomskii models long studied in the context of spin-orbit coupled materials. Howeve… ▽ More The discovery of correlated insulating states in moiré heterostructures has renewed the interest in strongly-coupled electron systems where spin and valley (or layer) degrees of freedom are intertwined. In the strong-coupling limit, such systems can be effectively described by SU(4) spin-valley models akin to Kugel-Khomskii models long studied in the context of spin-orbit coupled materials. However, typical moiré heterostructures also exhibit interactions that break the SU(4) symmetry down to SU(2)${}_{\mathrm{spin}}\otimes$U(1)${}_{\mathrm{valley}}$. Here we investigate the impact of such symmetry-breaking couplings on the magnetic phase diagram for triangular superlattices considering a filling of two electrons (or holes) per moiré unit cell. We explore a broad regime of couplings -- including XXZ anisotropies, Dzyaloshinskii-Moriya exchange and on-site Hund's couplings -- using semi-classical Monte Carlo simulations. We find a multitude of classically ordered phases, including (anti-)ferromagnetic, incommensurate, and stripe order, manifesting in different sectors of the spin-valley model's parameter space. Zooming in on the regimes where quantum fluctuations are likely to have an effect, we employ pseudo-fermion functional renormalization group (pf-FRG) calculations to resolve quantum disordered ground states such as spin-valley liquids, which we indeed find for certain parameter regimes. As a concrete example, we discuss the case of trilayer graphene aligned with hexagonal boron nitride using material-specific parameters. △ Less

Submitted 7 July, 2023; v1 submitted 2 March, 2023; originally announced March 2023.

Comments: 20 pages, 16 figures

Journal ref: Phys. Rev. B 108, 045102 (2023)

Switchable Signaling Molecules for Media Modulation: Fundamentals, Applications, and Research Directions

Authors: Lukas Brand, Maike Scherer, Sebastian Lotter, Teena tom Dieck, Maximilian Schäfer, Andreas Burkovski, Heinrich Sticht, Kathrin Castiglione, Robert Schober

Abstract: Although visionary applications of molecular communication (MC), such as long-term continuous health monitoring by cooperative in-body nanomachines, have been proposed, MC is still in its infancy when it comes to practical implementation. In particular, long-term experiments and applications face issues such as depletion of signaling molecules (SMs) at the transmitter (TX) and inter-symbol interfe… ▽ More Although visionary applications of molecular communication (MC), such as long-term continuous health monitoring by cooperative in-body nanomachines, have been proposed, MC is still in its infancy when it comes to practical implementation. In particular, long-term experiments and applications face issues such as depletion of signaling molecules (SMs) at the transmitter (TX) and inter-symbol interference (ISI) at the receiver (RX). To overcome these practical challenges, a new class of SMs with switchable states seems to be promising for future MC applications. In this work, we provide an overview of existing switchable SMs, and classify them according to their properties. Furthermore, we highlight how switchable SMs can be utilized as information carriers for media modulation. In addition, we present theoretical and experimental results for an end-to-end MC system employing the green fluorescent protein variant "Dreiklang" (GFPD) as switchable SM. Our experimental results show, for the first time, successful information transmission in a closed-loop pipe system using media modulation. Finally, we discuss media modulation specific challenges and opportunities. △ Less

Submitted 15 August, 2023; v1 submitted 20 February, 2023; originally announced February 2023.

Comments: 7 pages, 6 figures. This work has been accepted for publication in IEEE Communications Magazine

TCN-CUTIE: A 1036 TOp/s/W, 2.72 uJ/Inference, 12.2 mW All-Digital Ternary Accelerator in 22 nm FDX Technology

Authors: Moritz Scherer, Alfio Di Mauro, Tim Fischer, Georg Rutishauser, Luca Benini

Abstract: Tiny Machine Learning (TinyML) applications impose uJ/Inference constraints, with a maximum power consumption of tens of mW. It is extremely challenging to meet these requirements at a reasonable accuracy level. This work addresses the challenge with a flexible, fully digital Ternary Neural Network (TNN) accelerator in a RISC-V-based System-on-Chip (SoC). Besides supporting Ternary Convolutional N… ▽ More Tiny Machine Learning (TinyML) applications impose uJ/Inference constraints, with a maximum power consumption of tens of mW. It is extremely challenging to meet these requirements at a reasonable accuracy level. This work addresses the challenge with a flexible, fully digital Ternary Neural Network (TNN) accelerator in a RISC-V-based System-on-Chip (SoC). Besides supporting Ternary Convolutional Neural Networks, we introduce extensions to the accelerator design that enable the processing of time-dilated Temporal Convolutional Neural Networks (TCNs). The design achieves 5.5 uJ/Inference, 12.2 mW, 8000 Inferences/sec at 0.5 V for a Dynamic Vision Sensor (DVS) based TCN, and an accuracy of 94.5 % and 2.72 uJ/Inference, 12.2 mW, 3200 Inferences/sec at 0.5 V for a non-trivial 9-layer, 96 channels-per-layer convolutional network with CIFAR-10 accuracy of 86 %. The peak energy efficiency is 1036 TOp/s/W, outperforming the state-of-the-art silicon-proven TinyML quantized accelerators by 1.67x while achieving competitive accuracy. △ Less

Submitted 1 December, 2022; originally announced December 2022.

Comments: Accepted at IEEE MICRO Journal

Kraken: A Direct Event/Frame-Based Multi-sensor Fusion SoC for Ultra-Efficient Visual Processing in Nano-UAVs

Authors: Alfio Di Mauro, Moritz Scherer, Davide Rossi, Luca Benini

Abstract: Small-size unmanned aerial vehicles (UAV) have the potential to dramatically increase safety and reduce cost in applications like critical infrastructure maintenance and post-disaster search and rescue. Many scenarios require UAVs to shrink toward nano and pico-size form factors. The key open challenge to achieve true autonomy on Nano-UAVs is to run complex visual tasks like object detection, trac… ▽ More Small-size unmanned aerial vehicles (UAV) have the potential to dramatically increase safety and reduce cost in applications like critical infrastructure maintenance and post-disaster search and rescue. Many scenarios require UAVs to shrink toward nano and pico-size form factors. The key open challenge to achieve true autonomy on Nano-UAVs is to run complex visual tasks like object detection, tracking, navigation and obstacle avoidance fully on board, at high speed and robustness, under tight payload and power constraints. With the Kraken SoC, fabricated in 22nm FDX technology, we demonstrate a multi-visual-sensor capability exploiting both event-based and BW/RGB imagers, combining their output for multi-functional visual tasks previously impossible on a single low-power chip for Nano-UAVs. Kraken is an ultra-low-power, heterogeneous SoC architecture integrating three acceleration engines and a vast set of peripherals to enable efficient interfacing with standard frame-based sensors and novel event-based DVS. Kraken enables highly sparse event-driven sub-uJ/inf SNN inference on a dedicated neuromorphic energy-proportional accelerator. Moreover, it can perform frame-based inference by combining a 1.8TOp\s\W 8-cores RISC-V processor cluster with mixed-precision DNN extensions with a 1036TOp\s\W} TNN accelerator. △ Less

Submitted 18 August, 2022; originally announced September 2022.

SO(4) multicriticality of two-dimensional Dirac fermions

Authors: Igor F. Herbut, Michael M. Scherer

Abstract: We study quantum multicritical behavior in a (2+1)-dimensional Gross-Neveu-Yukawa field theory with eight-component Dirac fermions coupled to two triplets of order parameters that act as Dirac masses, and transform as $(1,0) + (0,1)$ representation under the SO(4)$\simeq$SO(3)$\times$SO(3) symmetry group. This field theory is relevant to spin-1/2 fermions on honeycomb or $π$-flux lattices, for exa… ▽ More We study quantum multicritical behavior in a (2+1)-dimensional Gross-Neveu-Yukawa field theory with eight-component Dirac fermions coupled to two triplets of order parameters that act as Dirac masses, and transform as $(1,0) + (0,1)$ representation under the SO(4)$\simeq$SO(3)$\times$SO(3) symmetry group. This field theory is relevant to spin-1/2 fermions on honeycomb or $π$-flux lattices, for example, near the transition points between an $s$-wave superconductor and a charge-density wave, on one side, and Néel order, on the other. Two triplets of such order parameters always allow for a common pair of two other order parameters that would complete them to the maximal set of compatible (anticommuting) orders of five. We first derive a unitary transformation in the Nambu (particle-hole) space which maps any two such triplets, possibly containing some superconducting orders, onto purely insulating order parameters. This allows one to consider a universal SO(4) Gross-Neveu-Yukawa description of the multicriticality without any Nambu doubling. We then proceed to derive the renormalization-group flow of the coupling constants at one-loop order in $4-ε$ space-time dimensions, allowing also a more general set of order parameters transforming under SO($n_a$)$\times$SO($n_b$). While for $n_a=n_b > 2 $ in the bosonic sector and with fermions decoupled there is a stable fixed point of the flow, the Yukawa coupling to fermions quickly leads to its elimination by a generic fixed-point collision in the relevant range of fermion flavor numbers $N_f$. This suggests the replacement of the critical behavior by a runaway flow in the physical case $n_a=n_b=3$. The structure of the RG flow at $n_a\neq n_b$ is also discussed, and some non-perturbative arguments in favor of the stability of the decoupled critical point when $n_a=3$ and $n_b=1$ in $D=2+1$ are provided. △ Less

Submitted 8 June, 2022; originally announced June 2022.

Comments: 13 pages, 4 figures

Journal ref: Physical Review, vol. 106, 115136 (2022)

Functional renormalization of spinless triangular-lattice fermions: $N$-patch vs. truncated-unity scheme

Authors: Nico Gneist, Dominik Kiese, Ravn Henkel, Ronny Thomale, Laura Classen, Michael M. Scherer

Abstract: We study competing orders of spinless fermions in the triangular-lattice Hubbard model with nearest-neighbor interaction. We calculate the effective, momentum-resolved two-particle vertex in an unbiased way in terms of the functional renormalization group method and compare two different schemes for the momentum discretization, one based on dividing the Fermi surface into patches and one based on… ▽ More We study competing orders of spinless fermions in the triangular-lattice Hubbard model with nearest-neighbor interaction. We calculate the effective, momentum-resolved two-particle vertex in an unbiased way in terms of the functional renormalization group method and compare two different schemes for the momentum discretization, one based on dividing the Fermi surface into patches and one based on a channel decomposition. We study attractive and repulsive nearest-neighbor interaction and find a competition of pairing and charge instabilities. In the attractive case, a Pomeranchuk instability occurs at Van Hove filling and $f$-wave and $p$-wave pairing emerge when the filling is reduced. In the repulsive case, we obtain a charge density wave at Van Hove filling and extended $p$-wave pairing with reduced filling. The $p$-wave pairing solution is doubly degenerate and can realize chiral $p+ip$ superconductivity with different Chern numbers in the ground state. We discuss implications for strongly correlated spin-orbit coupled hexagonal electron systems such as moiré heterostructures. △ Less

Submitted 25 May, 2022; originally announced May 2022.

Competition of Density Waves and Superconductivity in Twisted Tungsten Diselenide

Authors: Lennart Klebl, Ammon Fischer, Laura Classen, Michael M. Scherer, Dante M. Kennes

Abstract: Evidence for correlated insulating and superconducting phases around regions of high density of states was reported in the strongly spin-orbit coupled van-der Waals material twisted tungsten diselenide (tWSe$_2$). We investigate their origin and interplay by using a functional renormalization group approach that allows to describe superconducting and spin/charge instabilities in an unbiased way. W… ▽ More Evidence for correlated insulating and superconducting phases around regions of high density of states was reported in the strongly spin-orbit coupled van-der Waals material twisted tungsten diselenide (tWSe$_2$). We investigate their origin and interplay by using a functional renormalization group approach that allows to describe superconducting and spin/charge instabilities in an unbiased way. We map out the phase diagram as function of filling and perpendicular electric field, and find that the moiré Hubbard model for tWSe$_2$ features mixed-parity superconducting order parameters with $s/f$-wave and topological $d/p$-wave symmetry next to (incommensurate) density wave states. Our work systematically characterizes competing interaction-driven phases in tWSe$_2$ beyond mean-field approximations and provides guidance for experimental measurements by outlining the fingerprint of correlated states in interacting susceptibilities. △ Less

Submitted 14 March, 2023; v1 submitted 1 April, 2022; originally announced April 2022.

Comments: 7 pages, 3 figures, supplemental material

Journal ref: Phys. Rev. Research 5, L012034, Published 10 March 2023

Bioformers: Embedding Transformers for Ultra-Low Power sEMG-based Gesture Recognition

Authors: Alessio Burrello, Francesco Bianco Morghet, Moritz Scherer, Simone Benatti, Luca Benini, Enrico Macii, Massimo Poncino, Daniele Jahier Pagliari

Abstract: Human-machine interaction is gaining traction in rehabilitation tasks, such as controlling prosthetic hands or robotic arms. Gesture recognition exploiting surface electromyographic (sEMG) signals is one of the most promising approaches, given that sEMG signal acquisition is non-invasive and is directly related to muscle contraction. However, the analysis of these signals still presents many chall… ▽ More Human-machine interaction is gaining traction in rehabilitation tasks, such as controlling prosthetic hands or robotic arms. Gesture recognition exploiting surface electromyographic (sEMG) signals is one of the most promising approaches, given that sEMG signal acquisition is non-invasive and is directly related to muscle contraction. However, the analysis of these signals still presents many challenges since similar gestures result in similar muscle contractions. Thus the resulting signal shapes are almost identical, leading to low classification accuracy. To tackle this challenge, complex neural networks are employed, which require large memory footprints, consume relatively high energy and limit the maximum battery life of devices used for classification. This work addresses this problem with the introduction of the Bioformers. This new family of ultra-small attention-based architectures approaches state-of-the-art performance while reducing the number of parameters and operations of 4.9X. Additionally, by introducing a new inter-subjects pre-training, we improve the accuracy of our best Bioformer by 3.39%, matching state-of-the-art accuracy without any additional inference cost. Deploying our best performing Bioformer on a Parallel, Ultra-Low Power (PULP) microcontroller unit (MCU), the GreenWaves GAP8, we achieve an inference latency and energy of 2.72 ms and 0.14 mJ, respectively, 8.0X lower than the previous state-of-the-art neural network, while occupying just 94.2 kB of memory. △ Less

Submitted 25 March, 2022; v1 submitted 24 March, 2022; originally announced March 2022.

Competing instabilities of the extended Hubbard model on the triangular lattice: Truncated-unity functional renormalization group and application to moiré materials

Authors: Nico Gneist, Laura Classen, Michael M. Scherer

Abstract: A simple yet paradigmatic model for the interplay of strong electronic correlations and geometric frustration is the triangular lattice Hubbard model. Recently it was proposed that moiré structures of transition metal dichalcogenides can be used to simulate extended versions that include non-local density-density interactions. We study competing instabilities of interacting electrons in such an ex… ▽ More A simple yet paradigmatic model for the interplay of strong electronic correlations and geometric frustration is the triangular lattice Hubbard model. Recently it was proposed that moiré structures of transition metal dichalcogenides can be used to simulate extended versions that include non-local density-density interactions. We study competing instabilities of interacting electrons in such an extended Hubbard model on the triangular lattice near a filling where the density of states has a Van Hove singularity. We employ a truncated-unity functional renormalization group approach to investigate two cases: a paradigmatic minimally extended Hubbard model and a specific model with parameters that are applicable to hetero-bilayers of transition metal dichalcogenides. We unravel rich phase diagrams, including tendencies to spin-density-wave order and unconventional pairing, which can give rise to topological superconductivity. We classify the symmetry of the superconducting instabilities according to their irreducible representations and show that higher lattice harmonics are dominant when the nearest-neighbor interaction is sizable indicating pair formation between second-nearest neighbors. The phenomenological consequences can be enhanced spin and thermal quantum Hall responses in a topological superconductor. △ Less

Submitted 2 March, 2022; originally announced March 2022.

Comments: 11+7 pages, 11+2 figures

Deeptime: a Python library for machine learning dynamical models from time series data

Authors: Moritz Hoffmann, Martin Scherer, Tim Hempel, Andreas Mardt, Brian de Silva, Brooke E. Husic, Stefan Klus, Hao Wu, Nathan Kutz, Steven L. Brunton, Frank Noé

Abstract: Generation and analysis of time-series data is relevant to many quantitative fields ranging from economics to fluid mechanics. In the physical sciences, structures such as metastable and coherent sets, slow relaxation processes, collective variables dominant transition pathways or manifolds and channels of probability flow can be of great importance for understanding and characterizing the kinetic… ▽ More Generation and analysis of time-series data is relevant to many quantitative fields ranging from economics to fluid mechanics. In the physical sciences, structures such as metastable and coherent sets, slow relaxation processes, collective variables dominant transition pathways or manifolds and channels of probability flow can be of great importance for understanding and characterizing the kinetic, thermodynamic and mechanistic properties of the system. Deeptime is a general purpose Python library offering various tools to estimate dynamical models based on time-series data including conventional linear learning methods, such as Markov state models (MSMs), Hidden Markov Models and Koopman models, as well as kernel and deep learning approaches such as VAMPnets and deep MSMs. The library is largely compatible with scikit-learn, having a range of Estimator classes for these different models, but in contrast to scikit-learn also provides deep Model classes, e.g. in the case of an MSM, which provide a multitude of analysis methods to compute interesting thermodynamic, kinetic and dynamical quantities, such as free energies, relaxation times and transition paths. The library is designed for ease of use but also easily maintainable and extensible code. In this paper we introduce the main features and structure of the deeptime software. △ Less

Submitted 11 December, 2021; v1 submitted 28 October, 2021; originally announced October 2021.

Journal ref: Machine Learning: Science and Technology, Volume 3, Number 1, 2021

On the role of turbulent large-scale streaks in generating sediment ridges

Authors: Markus Scherer, Markus Uhlmann, Aman G. Kidanemariam, Michael Krayer

Abstract: The role of turbulent large-scale streaks in forming subaqueous sediment ridges on an initially flat sediment bed is investigated with the aid of particle-resolved direct numerical simulations of open channel flow at bulk Reynolds numbers up to 9500. The regular arrangement of quasi-streamwise ridges and troughs at a characteristic spanwise spacing between 1 and 1.5 times the mean fluid height is… ▽ More The role of turbulent large-scale streaks in forming subaqueous sediment ridges on an initially flat sediment bed is investigated with the aid of particle-resolved direct numerical simulations of open channel flow at bulk Reynolds numbers up to 9500. The regular arrangement of quasi-streamwise ridges and troughs at a characteristic spanwise spacing between 1 and 1.5 times the mean fluid height is found to be a consequence of the preferential spanwise organization of turbulence in large-scale streamwise velocity streaks. Sediment ridges predominantly appear in regions of weaker erosion below large-scale low-speed streaks and vice versa for troughs. The interaction between the dynamics of the large-scale streaks in the bulk flow and the evolution of sediment ridges on the sediment bed is best described as a `top-down' process, as the arrangement of the sediment bedforms is seen to adapt to changes in the outer flow with a time delay of several bulk time units. The observed `top-down' interaction between the outer flow and the bed agrees fairly well with the conceptual model on causality in canonical channel flows proposed by Jimenez (J. Fluid Mech., vol. 842, 2018, P1, section 5.6). Mean secondary currents of Prandtl's second kind of comparable intensity and lateral spacing are found over developed sediment ridges and in single-phase smooth-wall channels alike in averages over O(10) bulk time units. This indicates that the secondary flow commonly observed together with sediment ridges is the statistical footprint of the regularly organized large-scale streaks. △ Less

Submitted 2 October, 2021; originally announced October 2021.

Comments: original submission, revised version has been accepted for publication in J. Fluid Mech

$\mathcal{N}=4$ chiral superconductivity in moiré transition metal dichalcogenides

Authors: Michael M. Scherer, Dante M. Kennes, Laura Classen

Abstract: Experimental demonstrations of tunable correlation effects in magic-angle twisted bilayer graphene have put two-dimensional moiré quantum materials at the forefront of condensed-matter research. Other twisted few-layer graphitic structures, boron-nitride, and homo- or hetero-stacks of transition metal dichalcogenides (TMDs) have further enriched the opportunities for analysis and utilization of co… ▽ More Experimental demonstrations of tunable correlation effects in magic-angle twisted bilayer graphene have put two-dimensional moiré quantum materials at the forefront of condensed-matter research. Other twisted few-layer graphitic structures, boron-nitride, and homo- or hetero-stacks of transition metal dichalcogenides (TMDs) have further enriched the opportunities for analysis and utilization of correlations in these systems. Recently, within the latter material class, strong spin-orbit coupling or excitonic physics were experimentally explored. The observation of a Mott insulating state and other fascinating collective phenomena such as generalized Wigner crystals, stripe phases and quantum anomalous Hall insulators confirmed the relevance of many-body interactions, and demonstrated the importance of their extended range. Since the interaction, its range, and the filling can be tuned experimentally by twist angle, substrate engineering and gating, we here explore Fermi surface instabilities and resulting phases of matter of hetero-bilayer TMDs. Using an unbiased renormalization group approach, we establish in particular that hetero-bilayer TMDs are unique platforms to realize topological superconductivity with winding number $|\mathcal{N}|=4$. We show that this state reflects in pronounced experimental signatures, such as distinct quantum Hall features. △ Less

Submitted 25 August, 2021; originally announced August 2021.

Comments: 8 pages, 3 figures

Entanglement dynamics of spins using a few complex trajectories

Authors: Matheus V. Scherer, Alexandre D. Ribeiro

Abstract: In this work, we consider two spins initially prepared in a product of coherent states and study their entanglement dynamics due to a general interacting Hamiltonian. We adopt an approach that allowed the derivation of a semiclassical formula for the linear entropy of the reduced density operator, assumed as an entanglement quantifier. The resulting expression depends on sets of four trajectories,… ▽ More In this work, we consider two spins initially prepared in a product of coherent states and study their entanglement dynamics due to a general interacting Hamiltonian. We adopt an approach that allowed the derivation of a semiclassical formula for the linear entropy of the reduced density operator, assumed as an entanglement quantifier. The resulting expression depends on sets of four trajectories, originated from the underlying classical description, and having mutually connected final phase-space points. Such classical elements, which are capable to reproduce the quantum entanglement even for long values of propagation time, arise when we assume a proper analytical continuation of the classical phase space onto a complex domain. We apply this theory to a particular physical system, showing that taking into account only a few sets of complex trajectories is enough to get an excellent agreement between the semiclassical linear entropy of the reduced density operator and its quantum counterpart. △ Less

Submitted 12 August, 2021; originally announced August 2021.

Comments: 12 pages, 5 figures

Essential spherical isometries

Authors: Marcel Scherer

Abstract: A result due to Williams, Stampfli and Fillmore shows that an essential isometry $T$ on a Hilbert space $\mathcal{H}$ is a compact perturbation of an isometry if and only if ind$(T)\le 0$. A recent result of S. Chavan yields an analogous characterization of essential spherical isometries $T=(T_1,\dots,T_n)\in\mathcal{B}(\mathcal{H})^n$ with dim($\bigcap_{i=1}^n\ker(T_i))\le$ dim… ▽ More A result due to Williams, Stampfli and Fillmore shows that an essential isometry $T$ on a Hilbert space $\mathcal{H}$ is a compact perturbation of an isometry if and only if ind$(T)\le 0$. A recent result of S. Chavan yields an analogous characterization of essential spherical isometries $T=(T_1,\dots,T_n)\in\mathcal{B}(\mathcal{H})^n$ with dim($\bigcap_{i=1}^n\ker(T_i))\le$ dim$(\bigcap_{i=1}^n\ker(T_i^*))$. In the present note we show that in dimension $n>1$ the result of Chavan holds without any condition on the dimensions of the joint kernels of $T$ and $T^*$. △ Less

Submitted 12 May, 2021; originally announced May 2021.

Comments: 3 pages

MSC Class: 47A13 (47B07 47B20)

Journal ref: Acta Sci. Math. (Szeged) 86 (2020), no. 3-4, 667-670

Universal principles of moiré band structures

Authors: Jan Attig, **hong Park, Michael M. Scherer, Simon Trebst, Alexander Altland, Achim Rosch

Abstract: Moiré materials provide a highly tunable environment for the realization of band structures with engineered physical properties. Specifically, moiré structures with Fermi surface flat bands - a synthetic environment for the realization of correlated phases - have moiré unit cells containing thousands of atoms and tantalizingly complex bands structures. In this paper we show that statistical princi… ▽ More Moiré materials provide a highly tunable environment for the realization of band structures with engineered physical properties. Specifically, moiré structures with Fermi surface flat bands - a synthetic environment for the realization of correlated phases - have moiré unit cells containing thousands of atoms and tantalizingly complex bands structures. In this paper we show that statistical principles go a long way in explaining universal physical properties of these systems. Our approach builds on three conceptual elements: the presence of quantum chaos caused by the effective irregularity of the atomic configurations on short length scales, Anderson localization in momentum space, and the presence of approximate crystalline symmetries. Which of these principles dominates depends on material parameters such as the extension of the Fermi surface or the strength of the moiré lattice potential. The phenomenological consequences of this competition are predictions for the characteristic group velocity of moiré bands, a primary indicator for their average flatness. In addition to these generic features, we identify structures outside the statistical context, notably almost flat bands close to the extrema of the unperturbed spectra, and the celebrated zero energy `magic angle' flat bands, where the latter require exceptionally fine tuned material parameters. △ Less

Submitted 20 April, 2021; originally announced April 2021.

Comments: 15 pages, 10 figures

Journal ref: 2D Materials 8, 044007 (2021)

Fractionalized quantum criticality in spin-orbital liquids from field theory beyond the leading order

Authors: Shouryya Ray, Bernhard Ihrig, Daniel Kruti, John A. Gracey, Michael M. Scherer, Lukas Janssen

Abstract: Two-dimensional spin-orbital magnets with strong exchange frustration have recently been predicted to facilitate the realization of a quantum critical point in the Gross-Neveu-SO(3) universality class. In contrast to previously known Gross-Neveu-type universality classes, this quantum critical point separates a Dirac semimetal and a long-range-ordered phase, in which the fermion spectrum is only p… ▽ More Two-dimensional spin-orbital magnets with strong exchange frustration have recently been predicted to facilitate the realization of a quantum critical point in the Gross-Neveu-SO(3) universality class. In contrast to previously known Gross-Neveu-type universality classes, this quantum critical point separates a Dirac semimetal and a long-range-ordered phase, in which the fermion spectrum is only partially gapped out. Here, we characterize the quantum critical behavior of the Gross-Neveu-SO(3) universality class by employing three complementary field-theoretical techniques beyond their leading orders. We compute the correlation-length exponent $ν$, the order-parameter anomalous dimension $η_φ$, and the fermion anomalous dimension $η_ψ$ using a three-loop $ε$ expansion around the upper critical space-time dimension of four, a second-order large-$N$ expansion (with the fermion anomalous dimension obtained even at the third order), as well as a functional renormalization group approach in the improved local potential approximation. For the physically relevant case of $N=3$ flavors of two-component Dirac fermions in 2+1 space-time dimensions, we obtain the estimates $1/ν= 1.03(15)$, $η_φ= 0.42(7)$, and $η_ψ= 0.180(10)$ from averaging over the results of the different techniques, with the displayed uncertainty representing the degree of consistency among the three methods. △ Less

Submitted 1 April, 2021; v1 submitted 25 January, 2021; originally announced January 2021.

Comments: 18 pages, 7 figures, 3 tables, 1 ancillary file "GNSO3-exponents.m". v2: additional explanations, minor corrections

Report number: LTH 1253

Journal ref: Phys. Rev. B 103, 155160 (2021)

The Good, the Bad and the Ugly: Pitfalls and Best Practices in Automated Sound Static Analysis of Ethereum Smart Contracts

Authors: Clara Schneidewind, Markus Scherer, Matteo Maffei

Abstract: Ethereum smart contracts are distributed programs running on top of the Ethereum blockchain. Since program flaws can cause significant monetary losses and can hardly be fixed due to the immutable nature of the blockchain, there is a strong need of automated analysis tools which provide formal security guarantees. Designing such analyzers, however, proved to be challenging and error-prone. We revie… ▽ More Ethereum smart contracts are distributed programs running on top of the Ethereum blockchain. Since program flaws can cause significant monetary losses and can hardly be fixed due to the immutable nature of the blockchain, there is a strong need of automated analysis tools which provide formal security guarantees. Designing such analyzers, however, proved to be challenging and error-prone. We review the existing approaches to automated, sound, static analysis of Ethereum smart contracts and highlight prevalent issues in the state of the art. Finally, we overview eThor, a recent static analysis tool that we developed following a principled design and implementation approach based on rigorous semantic foundations to overcome the problems of past works. △ Less

Submitted 14 January, 2021; originally announced January 2021.

Constraints on discrete global symmetries in quantum gravity

Authors: Passant Ali, Astrid Eichhorn, Martin Pauly, Michael M. Scherer

Abstract: The question whether global symmetries can be realized in quantum-gravity-matter-systems has far-reaching phenomenological consequences. Here, we collect evidence that within an asymptotically safe context, discrete global symmetries of the form $\mathbb{Z}_n$, $n>4$, cannot be realized in a near-perturbative regime. In contrast, an effective-field-theory approach to quantum gravity might feature… ▽ More The question whether global symmetries can be realized in quantum-gravity-matter-systems has far-reaching phenomenological consequences. Here, we collect evidence that within an asymptotically safe context, discrete global symmetries of the form $\mathbb{Z}_n$, $n>4$, cannot be realized in a near-perturbative regime. In contrast, an effective-field-theory approach to quantum gravity might feature such symmetries, providing a mechanism to generate mass hierarchies in the infrared without the need for additional fine-tuning. △ Less

Submitted 3 September, 2021; v1 submitted 14 December, 2020; originally announced December 2020.

Comments: 27 pages, 10 figures. v2: matches published version

Journal ref: JHEP 05 (2021) 036

CUTIE: Beyond PetaOp/s/W Ternary DNN Inference Acceleration with Better-than-Binary Energy Efficiency

Authors: Moritz Scherer, Georg Rutishauser, Lukas Cavigelli, Luca Benini

Abstract: We present a 3.1 POp/s/W fully digital hardware accelerator for ternary neural networks. CUTIE, the Completely Unrolled Ternary Inference Engine, focuses on minimizing non-computational energy and switching activity so that dynamic power spent on storing (locally or globally) intermediate results is minimized. This is achieved by 1) a data path architecture completely unrolled in the feature map a… ▽ More We present a 3.1 POp/s/W fully digital hardware accelerator for ternary neural networks. CUTIE, the Completely Unrolled Ternary Inference Engine, focuses on minimizing non-computational energy and switching activity so that dynamic power spent on storing (locally or globally) intermediate results is minimized. This is achieved by 1) a data path architecture completely unrolled in the feature map and filter dimensions to reduce switching activity by favoring silencing over iterative computation and maximizing data re-use, 2) targeting ternary neural networks which, in contrast to binary NNs, allow for sparse weights which reduce switching activity, and 3) introducing an optimized training method for higher sparsity of the filter weights, resulting in a further reduction of the switching activity. Compared with state-of-the-art accelerators, CUTIE achieves greater or equal accuracy while decreasing the overall core inference energy cost by a factor of 4.8x-21x. △ Less

Submitted 4 February, 2021; v1 submitted 3 November, 2020; originally announced November 2020.

Exchangeable min-id sequences: Characterization, exponent measures and non-decreasing id-processes

Authors: Florian Brück, Jan-Frederik Mai, Matthias Scherer

Abstract: We establish a one-to-one correspondence between (i) exchangeable sequences of random variables whose finite-dimensional distributions are minimum (or maximum) infinitely divisible and (ii) non-negative, non-decreasing, infinitely divisible stochastic processes. The exponent measure of an exchangeable minimum infinitely divisible sequence is shown to be the sum of a very simple ``drift measure'' a… ▽ More We establish a one-to-one correspondence between (i) exchangeable sequences of random variables whose finite-dimensional distributions are minimum (or maximum) infinitely divisible and (ii) non-negative, non-decreasing, infinitely divisible stochastic processes. The exponent measure of an exchangeable minimum infinitely divisible sequence is shown to be the sum of a very simple ``drift measure'' and a mixture of product probability measures, which uniquely corresponds to the Lévy measure of a non-negative and non-decreasing infinitely divisible process. The latter is shown to be supported on non-negative and non-decreasing functions. In probabilistic terms, the aforementioned infinitely divisible process is equal to the conditional cumulative hazard process associated with the exchangeable sequence of random variables with minimum (or maximum) infinitely divisible marginals. Our results provide an analytic umbrella which embeds the de Finetti subfamilies of many interesting classes of multivariate distributions, such as exogenous shock models, exponential and geometric laws with lack-of-memory property, min-stable multivariate exponential and extreme-value distributions, as well as reciprocal Archimedean copulas with completely monotone generator and Archimedean copulas with log-completely monotone generator. △ Less

Submitted 20 September, 2022; v1 submitted 7 October, 2020; originally announced October 2020.

Comments: 57 pages, 3 Figures

MSC Class: 60G99; 60E07; 62H05

TinyRadarNN: Combining Spatial and Temporal Convolutional Neural Networks for Embedded Gesture Recognition with Short Range Radars

Authors: Moritz Scherer, Michele Magno, Jonas Erb, Philipp Mayer, Manuel Eggimann, Luca Benini

Abstract: This work proposes a low-power high-accuracy embedded hand-gesture recognition algorithm targeting battery-operated wearable devices using low power short-range RADAR sensors. A 2D Convolutional Neural Network (CNN) using range frequency Doppler features is combined with a Temporal Convolutional Neural Network (TCN) for time sequence prediction. The final algorithm has a model size of only 46 thou… ▽ More This work proposes a low-power high-accuracy embedded hand-gesture recognition algorithm targeting battery-operated wearable devices using low power short-range RADAR sensors. A 2D Convolutional Neural Network (CNN) using range frequency Doppler features is combined with a Temporal Convolutional Neural Network (TCN) for time sequence prediction. The final algorithm has a model size of only 46 thousand parameters, yielding a memory footprint of only 92 KB. Two datasets containing 11 challenging hand gestures performed by 26 different people have been recorded containing a total of 20,210 gesture instances. On the 11 hand gesture dataset, accuracies of 86.6% (26 users) and 92.4% (single user) have been achieved, which are comparable to the state-of-the-art, which achieves 87% (10 users) and 94% (single user), while using a TCN-based network that is 7500x smaller than the state-of-the-art. Furthermore, the gesture recognition classifier has been implemented on a Parallel Ultra-Low Power Processor, demonstrating that real-time prediction is feasible with only 21 mW of power consumption for the full TCN sequence prediction network, while a system-level power consumption of less than 100 mW is achieved. We provide open-source access to all the code and data collected and used in this work on tinyradar.ethz.ch. △ Less

Submitted 16 March, 2021; v1 submitted 25 June, 2020; originally announced June 2020.

Competing orders at higher-order Van Hove points

Authors: Laura Classen, Andrey V. Chubukov, Carsten Honerkamp, Michael M. Scherer

Abstract: Van Hove points are special points in the energy dispersion, where the density of states exhibits analytic singularities. When a Van Hove point is close to the Fermi level, tendencies towards density wave orders, Pomeranchuk orders, and superconductivity can all be enhanced, often in more than one channel, leading to a competition between different orders and unconventional ground states. Here we… ▽ More Van Hove points are special points in the energy dispersion, where the density of states exhibits analytic singularities. When a Van Hove point is close to the Fermi level, tendencies towards density wave orders, Pomeranchuk orders, and superconductivity can all be enhanced, often in more than one channel, leading to a competition between different orders and unconventional ground states. Here we consider the effects from higher-order Van Hove points, around which the dispersion is flatter than near a conventional Van Hove point, and the density of states has a power-law divergence. We argue that such points are present in intercalated graphene and other materials. We use an effective low-energy model for electrons near higher-order Van Hove points and analyze the competition between different ordering tendencies using an unbiased renormalization group approach. For purely repulsive interactions, we find that two key competitors are ferromagnetism and chiral superconductivity. For a small attractive exchange interaction, we find a new type of spin Pomeranchuk order, in which the spin order parameter winds around the Fermi surface. The supermetal state, predicted for a single higher-order Van Hove point, is an unstable fixed point in our case. △ Less

Submitted 25 June, 2020; originally announced June 2020.

Comments: 11+10 pages, 9+4 figures

Journal ref: Phys. Rev. B 102, 125141 (2020)

On the scaling of the instability of a flat sediment bed with respect to ripple-like patterns

Authors: Markus Scherer, Aman G. Kidanemariam, Markus Uhlmann

Abstract: We investigate the formation of subaqueous transverse bedforms in turbulent open channel flow by means of direct numerical simulations with fully-resolved particles. The main goal of the present analysis is to address the question whether the initial pattern wavelength scales with the particle diameter or with the mean fluid height. A previous study (Kidanemariam and Uhlmann, J. Fluid Mech., vol.… ▽ More We investigate the formation of subaqueous transverse bedforms in turbulent open channel flow by means of direct numerical simulations with fully-resolved particles. The main goal of the present analysis is to address the question whether the initial pattern wavelength scales with the particle diameter or with the mean fluid height. A previous study (Kidanemariam and Uhlmann, J. Fluid Mech., vol. 818, 2017, pp. 716-743) has observed a lower bound for the most unstable pattern wavelength in the range 75-100 times the particle diameter, which was equivalent to 3-4 times the mean fluid height. In the current paper, we vary the streamwise box length in terms of the particle diameter and of the mean fluid height independently in order to distinguish between the two possible scaling relations. For the chosen parameter range, the obtained results clearly exhibit a scaling of the initial pattern wavelength with the particle diameter, with a lower bound around a streamwise extent of approximately 80 particle diameters. In longer domains, on the other hand, patterns are observed at initial wavelengths in the range 150-180 times the particle diameter, which is in good agreement with experimental measurements. Variations of the mean fluid height, on the other hand, seem to have no significant influence on the most unstable initial pattern wavelength. Furthermore, for the cases with the largest relative submergence, we observe spanwise and streamwise sediment waves of similar amplitude to evolve and superimpose, leading to three-dimensional sediment patterns. △ Less

Submitted 18 May, 2020; originally announced May 2020.

Comments: original submission, revised version has been accepted for publication in J. Fluid Mech

eThor: Practical and Provably Sound Static Analysis of Ethereum Smart Contracts

Authors: Clara Schneidewind, Ilya Grishchenko, Markus Scherer, Matteo Maffei

Abstract: Ethereum has emerged as the most popular smart contract development platform, with hundreds of thousands of contracts stored on the blockchain and covering a variety of application scenarios, such as auctions, trading platforms, and so on. Given their financial nature, security vulnerabilities may lead to catastrophic consequences and, even worse, they can be hardly fixed as data stored on the blo… ▽ More Ethereum has emerged as the most popular smart contract development platform, with hundreds of thousands of contracts stored on the blockchain and covering a variety of application scenarios, such as auctions, trading platforms, and so on. Given their financial nature, security vulnerabilities may lead to catastrophic consequences and, even worse, they can be hardly fixed as data stored on the blockchain, including the smart contract code itself, are immutable. An automated security analysis of these contracts is thus of utmost interest, but at the same time technically challenging for a variety of reasons, such as the specific transaction-oriented programming mechanisms, which feature a subtle semantics, and the fact that the blockchain data which the contract under analysis interacts with, including the code of callers and callees, are not statically known. In this work, we present eThor, the first sound and automated static analyzer for EVM bytecode, which is based on an abstraction of the EVM bytecode semantics based on Horn clauses. In particular, our static analysis supports reachability properties, which we show to be sufficient for capturing interesting security properties for smart contracts (e.g., single-entrancy) as well as contract-specific functional properties. Our analysis is proven sound against a complete semantics of EVM bytecode and an experimental large-scale evaluation on real-world contracts demonstrates that eThor is practical and outperforms the state-of-the-art static analyzers: specifically, eThor is the only one to provide soundness guarantees, terminates on 95% of a representative set of real-world contracts, and achieves an F-measure (which combines sensitivity and specificity) of 89%. △ Less

Submitted 13 May, 2020; originally announced May 2020.

Comments: Accepted for CCS 2020

Realization of Nearly Dispersionless Bands with Strong Orbital Anisotropy from Destructive Interference in Twisted Bilayer MoS2

Authors: Lede Xian, Martin Claassen, Dominik Kiese, Michael M. Scherer, Simon Trebst, Dante M. Kennes, Angel Rubio

Abstract: Recently, the twist angle between adjacent sheets of stacked van der Waals materials emerged as a new knob to engineer correlated states of matter in two-dimensional heterostructures in a controlled manner, giving rise to emergent phenomena such as superconductivity or correlated insulating states. Here,we use an ab initio based approach to characterize the electronic properties of twisted bilayer… ▽ More Recently, the twist angle between adjacent sheets of stacked van der Waals materials emerged as a new knob to engineer correlated states of matter in two-dimensional heterostructures in a controlled manner, giving rise to emergent phenomena such as superconductivity or correlated insulating states. Here,we use an ab initio based approach to characterize the electronic properties of twisted bilayer MoS2. We report that, in marked contrast to twisted bilayer graphene, slightly hole-doped MoS2 realizes a strongly asymmetric px-py Hubbard model on the honeycomb lattice, with two almost entirely dispersionless bands emerging due to destructive interference. We study the collective behavior of twisted bilayer MoS2 in the presence of interactions, and characterize an array of different magnetic and orbitally-ordered correlated phases,which may be susceptible to quantum fluctuations giving rise to exotic, purely quantum, states of matter. △ Less

Submitted 6 April, 2020; originally announced April 2020.

Journal ref: Nature Communications 12, 5644 (2021)

Confinement transition in the QED$_3$-Gross-Neveu-XY universality class

Authors: Lukas Janssen, Wei Wang, Michael M. Scherer, Zi Yang Meng, Xiao Yan Xu

Abstract: The coupling between fermionic matter and gauge fields plays a fundamental role in our understanding of nature, while at the same time posing a challenging problem for theoretical modeling. In this situation, controlled information can be gained by combining different complementary approaches. Here, we study a confinement transition in a system of $N_f$ flavors of interacting Dirac fermions charge… ▽ More The coupling between fermionic matter and gauge fields plays a fundamental role in our understanding of nature, while at the same time posing a challenging problem for theoretical modeling. In this situation, controlled information can be gained by combining different complementary approaches. Here, we study a confinement transition in a system of $N_f$ flavors of interacting Dirac fermions charged under a U(1) gauge field in 2+1 dimensions. Using Quantum Monte Carlo simulations, we investigate a lattice model that exhibits a continuous transition at zero temperature between a gapless deconfined phase, described by three-dimensional quantum electrodynamics, and a gapped confined phase, in which the system develops valence-bond-solid order. We argue that the quantum critical point is in the universality class of the QED$_3$-Gross-Neveu-XY model. We study this field theory within a $1/N_f$ expansion in fixed dimension as well as a renormalization group analysis in $4-ε$ space-time dimensions. The consistency between numerical and analytical results is revealed from large to intermediate flavor number. △ Less

Submitted 11 May, 2020; v1 submitted 3 March, 2020; originally announced March 2020.

Comments: 10 pages, 9 figures; v2: additional data and explanations, corrected erroneous interpretation of dimer scaling dimension

Journal ref: Phys. Rev. B 101, 235118 (2020)

Multivariate geometric distributions, (logarithmically) monotone sequences, and infinitely divisible laws (with erratum by Natalia Shenkman)

Authors: Jan-Frederik Mai, Matthias Scherer, Natalia Shenkman

Abstract: Two stochastic representations of multivariate geometric distributions are analyzed, both are obtained by lifting the lack-of-memory (LM) property of the univariate geometric law to the multivariate case. On the one hand, the narrow-sense multivariate geometric law can be considered a discrete equivalent of the well-studied Marshall-Olkin exponential law. On the other hand, the more general wide-s… ▽ More Two stochastic representations of multivariate geometric distributions are analyzed, both are obtained by lifting the lack-of-memory (LM) property of the univariate geometric law to the multivariate case. On the one hand, the narrow-sense multivariate geometric law can be considered a discrete equivalent of the well-studied Marshall-Olkin exponential law. On the other hand, the more general wide-sense geometric law is shown to be characterized by the LM property and can differ significantly from its continuous counterpart, e.g., by allowing for negative pairwise correlations. For both families of distributions, their $d$-dimensional exchangeable subclass is characterized analytically via $d$-log-monotone, resp.\ $d$-monotone, sequences of parameters. Using this reparameterization, the subfamilies of distributions with conditionally i.i.d.\ components in the sense of de Finetti's theorem are determined. For these, a third stochastic construction based on a non-decreasing random walk is presented. The narrow-sense family is embedded in this construction when the increments of the involved random walk are infinitely divisible. The exchangeable narrow-sense law is furthermore shown to exhibit the multivariate right tail increasing (MRTI) dependence. △ Less

Submitted 3 February, 2020; originally announced February 2020.

Comments: 39 pages

MSC Class: 62H05; 62H20

Exploring the added value of blockchain technology for the healthcare domain

Authors: Bas R. J. Bolmer, Monique Taverne, Marco Scherer

Abstract: In this report, the University Medical Center Groningen (UMCG) has written down lessons learned on how blockchain technology can have an impact on the healthcare domain. By looking at two use-cases, the hospital challenged several teams, participating in an open innovation program and blockchain hackathon, to find a solution that showed the added value of the technology for patient care and scient… ▽ More In this report, the University Medical Center Groningen (UMCG) has written down lessons learned on how blockchain technology can have an impact on the healthcare domain. By looking at two use-cases, the hospital challenged several teams, participating in an open innovation program and blockchain hackathon, to find a solution that showed the added value of the technology for patient care and scientific research. Besides this practical perspective, the report also considers literature discussing the current state of blockchain technology in regard to developments in the healthcare domain (touching on patient empowerment, data management, regulations, and interoperability between healthcare systems). △ Less

Submitted 15 November, 2019; originally announced November 2019.

Emergent symmetries and coexisting orders in Dirac fermion systems

Authors: Emilio Torres, Lukas Weber, Lukas Janssen, Stefan Wessel, Michael M. Scherer

Abstract: The quantum phase diagram and critical behavior of two-dimensional Dirac fermions coupled to two compatible order-parameter fields with $O(N_1)\oplus O(N_2)$ symmetry is investigated. Recent numerical studies of such systems have reported evidence for non-Landau-Ginzburg-Wilson transitions and emergent $O(N_1+N_2)$ symmetry between the two ordered states, which has been interpreted within a scenar… ▽ More The quantum phase diagram and critical behavior of two-dimensional Dirac fermions coupled to two compatible order-parameter fields with $O(N_1)\oplus O(N_2)$ symmetry is investigated. Recent numerical studies of such systems have reported evidence for non-Landau-Ginzburg-Wilson transitions and emergent $O(N_1+N_2)$ symmetry between the two ordered states, which has been interpreted within a scenario of deconfined quantum criticality in (2+1)-dimensional Dirac materials. Here, we provide two theoretical approaches to refine the phase diagrams of such systems. In the immediate vicinity of the multicritical point between the ordered phases and the semimetallic phase, we employ a non-perturbative field-theoretical analysis based on the functional renormalization group. For the particular case of $N_1=3$, $N_2=1$, we perform a large-scale quantum Monte Carlo analysis of the strong-coupling region, where both orders meet. Our findings support the robust emergence of enhanced symmetry at the multicritical point and suggest the transition between the two ordered phases to take place via a sequence of continuous transitions. In particular, we find that intermediate regimes of coexistence are present in the phase diagram for all values of $N_1$ and $N_2$. △ Less

Submitted 4 November, 2019; originally announced November 2019.

Comments: 4+9 pages, 3+4 figures, 1+0 tables

Journal ref: Phys. Rev. Research 2, 022005 (2020)

A size-dependent ductile fracture model: Constitutive equations, Numerical implementation and Validation

Authors: J. M. Scherer, J. Hure

Abstract: Size effects have been predicted at the micro- or nano-scale for porous ductile materials from Molecular Dynamics, Discrete Dislocation Dynamics and Continuum Mechanics numerical simulations, as a consequence of Geometrically Necessary Dislocations or due to the presence of a void matrix interface. As voids size decreases, higher stresses are needed to deform the material, for a given porosity. Ho… ▽ More Size effects have been predicted at the micro- or nano-scale for porous ductile materials from Molecular Dynamics, Discrete Dislocation Dynamics and Continuum Mechanics numerical simulations, as a consequence of Geometrically Necessary Dislocations or due to the presence of a void matrix interface. As voids size decreases, higher stresses are needed to deform the material, for a given porosity. However, the majority of the homogenized models for porous materials used in ductile fracture modeling are size-independent, even though micrometric or nanometric voids are commonly observed in structural materials. Based on yield criteria proposed in the literature for nanoporous materials, a size-dependent homogenized model for porous materials is proposed for axisymmetric loading conditions, including void growth and coalescence as well as void shape effects. Numerical implementation of the constitutive equations is detailed. The homogenized model is validated through comparisons to porous unit cells finite element simulations that consider interfacial stresses, consistently with the model used for the derivation of the yield criteria, aiming at modeling an additional hardening at the void matrix interface. Potential improvements of the model are finally discussed with respect to the theoretical derivation of refined yield criteria and evolution laws. △ Less

Submitted 26 October, 2019; originally announced October 2019.

Journal ref: Eur. J. Mech. A/Solids, 76, 135-145 (2019)

On the structure of exchangeable extreme-value copulas

Authors: Jan-Frederik Mai, Matthias Scherer

Abstract: We show that the set of $d$-variate symmetric stable tail dependence functions, uniquely associated with exchangeable $d$-dimensional extreme-value copulas, is a simplex and determine its extremal boundary. The subset of elements which arises as $d$-margins of the set of $(d+k)$-variate symmetric stable tail dependence functions is shown to be proper for arbitrary $k \geq 1$. Finally, we derive an… ▽ More We show that the set of $d$-variate symmetric stable tail dependence functions, uniquely associated with exchangeable $d$-dimensional extreme-value copulas, is a simplex and determine its extremal boundary. The subset of elements which arises as $d$-margins of the set of $(d+k)$-variate symmetric stable tail dependence functions is shown to be proper for arbitrary $k \geq 1$. Finally, we derive an intuitive and useful necessary condition for a bivariate extreme-value copula to arise as bi-margin of an exchangeable extreme-value copula of arbitrarily large dimension, and thus to be conditionally iid. △ Less

Submitted 20 September, 2019; originally announced September 2019.

Journal ref: Journal of Multivariate Analysis 180, 104670 [11 pages] (2020)

Emergence and stability of spin-valley entangled quantum liquids in moiré heterostructures

Authors: Dominik Kiese, Finn Lasse Buessen, Ciarán Hickey, Simon Trebst, Michael M. Scherer

Abstract: Twisting moiré heterostructures to the flatband regime allows for the formation of strongly correlated quantum states, since the dramatic reduction of the bandwidth can cause the residual electronic interactions to set the principal energy scale. An effective description for such correlated moiré heterostructures, derived in the strong-coupling limit at integer filling, generically leads to spin-v… ▽ More Twisting moiré heterostructures to the flatband regime allows for the formation of strongly correlated quantum states, since the dramatic reduction of the bandwidth can cause the residual electronic interactions to set the principal energy scale. An effective description for such correlated moiré heterostructures, derived in the strong-coupling limit at integer filling, generically leads to spin-valley Heisenberg models. Here we explore the emergence and stability of spin liquid behavior in an SU(2)$^{\mathrm{spin}}\otimes$SU(2)$^{\mathrm{valley}}$ Heisenberg model upon inclusion of Hund's-induced and longer-ranged exchange couplings, employing a pseudofermion functional renormalization group approach. We consider two lattice geometries, triangular and honeycomb (relevant to different moiré heterostructures), and find, for both cases, an extended parameter regime surrounding the SU(4) symmetric point where no long-range order occurs, indicating a stable realm of quantum spin liquid behavior. For large Hund's coupling, we identify the adjacent magnetic orders, with both antiferromagnetic and ferromagnetic ground states emerging in the separate spin and valley degrees of freedom. For both lattice geometries the inclusion of longer-ranged exchange couplings is found to have both stabilizing and destabilizing effects on the spin liquid regime depending on the sign of the additional couplings. △ Less

Submitted 22 July, 2019; originally announced July 2019.

Comments: 5 pages, 4 figures plus appendix (4 pages, 5 figures)

Journal ref: Phys. Rev. Research 2, 013370 (2020)

Abelian Higgs model at four loops, fixed-point collision and deconfined criticality

Authors: Bernhard Ihrig, Nikolai Zerf, Peter Marquard, Igor F. Herbut, Michael M. Scherer

Abstract: The abelian Higgs model is the textbook example for the superconducting transition and the Anderson-Higgs mechanism, and has become pivotal in the description of deconfined quantum criticality. We study the abelian Higgs model with $n$ complex scalar fields at unprecedented four-loop order in the $4-ε$ expansion and find that the annihilation of the critical and bicritical points occurs at a criti… ▽ More The abelian Higgs model is the textbook example for the superconducting transition and the Anderson-Higgs mechanism, and has become pivotal in the description of deconfined quantum criticality. We study the abelian Higgs model with $n$ complex scalar fields at unprecedented four-loop order in the $4-ε$ expansion and find that the annihilation of the critical and bicritical points occurs at a critical number of $n_c \approx 182.95\left(1 - 1.752ε+ 0.798 ε^2 + 0.362ε^3\right) + \mathcal{O}\left(ε^4\right)\nonumber$. Consequently, below $n_c$, the transition turns from second to first order. Resummation of the series to extract the result in three-dimensions provides strong evidence for a critical $n_c(d=3)$ which is significantly below the leading-order value, but the estimates for $n_c$ are widely spread. Conjecturing the topology of the renormalization group flow between two and four dimensions, we obtain a smooth interpolation function for $n_c(d)$ and find $n_c(3)\approx 12.2\pm 3.9$ as our best estimate in three dimensions. Finally, we discuss Miransky scaling occurring below $n_c$ and comment on implications for weakly first-order behavior of deconfined quantum transitions. We predict an emergent hierarchy of length scales between deconfined quantum transitions corresponding to different $n$. △ Less

Submitted 18 July, 2019; originally announced July 2019.

Comments: 15 pages, 6 figures, 3 tables

Report number: HU-EP-19/19, DESY 19-124, SAGEX-19-16

Journal ref: Phys. Rev. B 100, 134507 (2019)

Cluster functional renormalization group and absence of a bilinear spin liquid in the $J_1$-$J_2$-Heisenberg model

Authors: Dietrich Roscher, Nico Gneist, Michael M. Scherer, Simon Trebst, Sebastian Diehl

Abstract: The pseudofermion functional renormalization group (pf-FRG) has been put forward as a semi-analytical scheme that, for a given microscopic spin model, allows to discriminate whether the low-temperature states exhibit magnetic ordering or a tendency towards the formation of quantum spin liquids. However, the precise nature of the putative spin liquid ground state has remained hard to infer from the… ▽ More The pseudofermion functional renormalization group (pf-FRG) has been put forward as a semi-analytical scheme that, for a given microscopic spin model, allows to discriminate whether the low-temperature states exhibit magnetic ordering or a tendency towards the formation of quantum spin liquids. However, the precise nature of the putative spin liquid ground state has remained hard to infer from the original (single-site) pf-FRG scheme. Here we introduce a cluster pf-FRG approach, which allows for a more stringent connection between a microscopic spin model and its low-temperature spin liquid ground states. In particular, it allows to calculate spatially structured fermion bilinear expectation values on spatial clusters, which are formed by splitting the original lattice into several sublattices, thereby allowing for the positive identification of a family of bilinear spin liquid states. As an application of this cluster pf-FRG approach, we consider the $J_1$-$J_2$ SU($N$)-Heisenberg model on a square lattice, which is a paradigmatic example for a frustrated quantum magnet exhibiting quantum spin liquid behavior for intermediate coupling strengths. In the well-established large-$N$ limit of this model, we show that our approach correctly captures the emergence of the $π$-flux spin liquid state at low temperatures. For small $N$, where the precise nature of the ground state remains controversial, we focus on the widely studied case of $N=2$, for which we determine the low-temperature phase diagram near the strongly-frustrated regime after implementing the fermion number constraint by the flowing Popov-Fedotov method. Our results suggest that the $J_1$-$J_2$-Heisenberg model does not support the formation of a fermion bilinear spin liquid state. △ Less

Submitted 3 May, 2019; originally announced May 2019.

Comments: 13 pages, 9 figures

Journal ref: Phys. Rev. B 100, 125130 (2019)

Competing phases of interacting electrons on triangular lattices in moiré heterostructures

Authors: Laura Classen, Carsten Honerkamp, Michael M. Scherer

Abstract: We study the quantum many-body instabilities of interacting electrons with SU(2)$\times$SU(2) symmetry in spin and orbital degrees of freedom on the triangular lattice near van-Hove filling. Our work is motivated by effective models for the flat bands in hexagonal moiré heterostructures like twisted bilayer boron nitride and trilayer graphene-boron nitride systems. We consider an extended Hubbard… ▽ More We study the quantum many-body instabilities of interacting electrons with SU(2)$\times$SU(2) symmetry in spin and orbital degrees of freedom on the triangular lattice near van-Hove filling. Our work is motivated by effective models for the flat bands in hexagonal moiré heterostructures like twisted bilayer boron nitride and trilayer graphene-boron nitride systems. We consider an extended Hubbard model including onsite Hubbard and Hund's couplings, as well as nearest-neighbor exchange interactions and analyze the different ordering tendencies with the help of an unbiased functional renormalization group approach. We find three classes of instabilities controlled by the filling and bare interactions. For a nested Fermi surface at van-Hove filling, Hund-like couplings induce a weak instability towards spin or orbital density wave phases. An SU(4) exchange interaction moves the system towards a Chern insulator, which is robust with respect to perturbations from Hund-like interactions or deviations from perfect nesting. Further, in an extended range of fillings and interactions, we find topological $d\pm id$ and (spin-singlet)-(orbital-singlet) $f$-wave superconductivity. △ Less

Submitted 14 February, 2019; originally announced February 2019.

Comments: 5+4 pages, 4+1 figures

Journal ref: Phys. Rev. B 99, 195120 (2019)

Variational Selection of Features for Molecular Kinetics

Authors: Martin K. Scherer, Brooke E. Husic, Moritz Hoffmann, Fabian Paul, Hao Wu, Frank Noé

Abstract: The modeling of atomistic biomolecular simulations using kinetic models such as Markov state models (MSMs) has had many notable algorithmic advances in recent years. The variational principle has opened the door for a nearly fully automated toolkit for selecting models that predict the long-time kinetics from molecular dynamics simulations. However, one yet-unoptimized step of the pipeline involve… ▽ More The modeling of atomistic biomolecular simulations using kinetic models such as Markov state models (MSMs) has had many notable algorithmic advances in recent years. The variational principle has opened the door for a nearly fully automated toolkit for selecting models that predict the long-time kinetics from molecular dynamics simulations. However, one yet-unoptimized step of the pipeline involves choosing the features, or collective variables, from which the model should be constructed. In order to build intuitive models, these collective variables are often sought to be interpretable and familiar features, such as torsional angles or contact distances in a protein structure. However, previous approaches for evaluating the chosen features rely on constructing a full MSM, which in turn requires additional hyperparameters to be chosen, and hence leads to a computationally expensive framework. Here, we present a method to optimize the feature choice directly, without requiring the construction of the final kinetic model. We demonstrate our rigorous preprocessing algorithm on a canonical set of twelve fast-folding protein simulations, and show that our procedure leads to more efficient model selection. △ Less

Submitted 25 April, 2019; v1 submitted 28 November, 2018; originally announced November 2018.

Comments: 13 pages, 8 figures

Journal ref: J. Chem. Phys. 2019, 150, 194108

Subordinators which are infinitely divisible w.r.t. time: Construction, properties, and simulation of max-stable sequences and infinitely divisible laws

Authors: Jan-Frederik Mai, Matthias Scherer

Abstract: The concept of a Lévy subordinator is generalized to a family of non-decreasing stochastic processes, which are parameterized in terms of two Bernstein functions. Whereas the independent increments property is only maintained in the Lévy subordinator special case, the considered family is always strongly infinitely divisible with respect to time, meaning that a path can be represented in distribut… ▽ More The concept of a Lévy subordinator is generalized to a family of non-decreasing stochastic processes, which are parameterized in terms of two Bernstein functions. Whereas the independent increments property is only maintained in the Lévy subordinator special case, the considered family is always strongly infinitely divisible with respect to time, meaning that a path can be represented in distribution as a finite sum with arbitrarily many summands of independent and identically distributed paths of another process. Besides distributional properties of the process, we present two applications to the design of accurate and efficient simulation algorithms. First, each member of the considered family corresponds uniquely to an exchangeable max-stable sequence of random variables, and we demonstrate how the associated extreme-value copula can be simulated exactly and efficiently from its Pickands dependence measure. Second, we show how one obtains different series and integral representations for infinitely divisible probability laws by varying the parameterizing pair of Bernstein functions, without changing the law of one-dimensional margins of the process. As a particular example, we present an exact simulation algorithm for compound Poisson distributions from the Bondesson class, for which the generalized inverse of the distribution function of the associated Stieltjes measure can be evaluated accurately. △ Less

Submitted 12 October, 2018; originally announced October 2018.

Journal ref: ALEA, Lat. Am. J. Probab. Math. Stat. 16, 977-1005 (2019)

Mass hierarchy in collective modes of pair-density-wave superconductors

Authors: Shao-Kai Jian, Michael M. Scherer, Hong Yao

Abstract: We study collective modes near the quantum critical point of a pair-density-wave (PDW) superconductor in 2+1 dimensions. The fate of gaps of various collective modes is investigated by functional renormalization. For incommensurate PDW superconductors, we show that the gapless Leggett mode, protected by the emergent $U(1)$ symmetry, can induce an exponentially small Higgs mass compared to the supe… ▽ More We study collective modes near the quantum critical point of a pair-density-wave (PDW) superconductor in 2+1 dimensions. The fate of gaps of various collective modes is investigated by functional renormalization. For incommensurate PDW superconductors, we show that the gapless Leggett mode, protected by the emergent $U(1)$ symmetry, can induce an exponentially small Higgs mass compared to the superconducting gap. Further, for commensurate PDW superconductors, we find an emergent mass hierarchy in the collective modes, i.e. the masses of Leggett boson, Higgs boson, and the superconducting gap can differ by several magnitudes in the infrared. This may shed light to a mechanism underlying the hierarchy problem in the Standard Model of particle physics. △ Less

Submitted 10 January, 2020; v1 submitted 2 October, 2018; originally announced October 2018.

Comments: 8 pages, 5 figures

Journal ref: Phys. Rev. Research 2, 013034 (2020)