As custom hardware accelerators become increasingly central to machine learning workloads, efficient data transfer is critical for maximizing accelerator performance on linear algebra kernels. AXI4MLIR, an extension of the Multi-Level Intermediate Representation (MLIR) compiler framework for automated generation of host-accelerator driver code, incurs significant runtime overhead due to non-zero-copy CPU-accelerator data movement. During transfers from the host to the accelerator, data is copied from heap-allocated memory buffers into contiguous Direct Memory Access (DMA)-mapped buffers. This work identifies this copy as a redundant staging operation and eliminates it through zero-copy data movement. The optimization extends accel, an MLIR dialect introduced by AXI4MLIR, and implements lowering support that allocates buffers directly within DMA-mapped memory, thereby omitting the staging copy. We evaluate the proposed scheme using a configurable matrix-matrix multiplication accelerator and show that the zero-copy optimization reduces main memory data movement by up to 2x, increasing overall accelerator utilization.
We present and show how to implement a non-trivial all-to-all communication algorithm for arbitrary $d$-dimensional tori effectively in MPI. Given a factorization of the number of processes $p$ into $d$ factors that can be mapped onto a $d$-dimensional torus, we first utilize a Cartesian communicator to split a given $p$-process MPI communicator into, for each MPI process, $d$ smaller communicators spanning each of the dimensions of the torus to which the process belongs, and cache these communicators in order to avoid expensive splitting at each all-to-all operation. The all-to-all operation itself is decomposed into a sequence of $d$ MPI_Alltoall operations on the dimension-wise communicators. The non-trivial data rearrangement before and after each MPI_Alltoall call is implicit only and effected by MPI derived datatypes. This makes the implementation of the algorithm formally \emph{zero-copy}, meaning that no explicit process-local reordering of data blocks ever has to be performed. In order to achieve this, the algorithm employs a double-buffering scheme with modest temporary buffer requirements. By choosing the factorization of $p$ and selecting appropriate implementations for
Robot applications, comprising independent components that mutually publish/subscribe messages, are built on inter-process communication (IPC) middleware such as Robot Operating System 2 (ROS 2). In large-scale ROS 2 systems like autonomous driving platforms, true zero-copy communication -- eliminating serialization and deserialization -- is crucial for efficiency and real-time performance. However, existing true zero-copy middleware solutions lack widespread adoption as they fail to meet three essential requirements: 1) Support for all ROS 2 message types including unsized ones; 2) Minimal modifications to existing application code; 3) Selective implementation of zero-copy communication between specific nodes while maintaining conventional communication mechanisms for other inter-node communications including inter-host node communications. This first requirement is critical, as production-grade ROS 2 projects like Autoware rely heavily on unsized message types throughout their codebase to handle diverse use cases (e.g., various sensors), and depend on the broader ROS 2 ecosystem, where unsized message types are pervasive in libraries. The remaining requirements facilitate seamles
Per-ticker forecasting models dominate financial time-series work yet remain blind to cross-company propagation: a foundry disruption in Taiwan does not register in a single-asset model until Apple's own price has already moved. To address this limitation, we introduce a heterogeneous Rust-Python streaming architecture that maps cross-company attention as a continuous-time graph driven directly from text. We show that on the ingestion side, a zero-copy Rust edge parses news records in $\sim$100 ns and scans the target equity universe in $\sim$1.2 $μ$s. On the inference end, a multivariate Neural Hawkes Process featuring per-node continuous-time LSTM states and a bilinear latent projection propagates directed excitation, while an adaptive pruning rule bounds the computational cost of dynamic neighborhood updates. Combining these stages, we demonstrate an end-to-end processing latency of $\sim$13 ms per incoming news record on a single commodity CPU. Evaluated on a one-month temporal holdout of the FNSPID corpus (638 articles across 47 tickers), the system delivers a $1.70\times$ precision lift over random at the 90th-percentile next-day return threshold, and $3.36\times$ over a same
Serverless computing promises enhanced resource efficiency and lower user costs, yet is burdened by a heavyweight, CPU-bound data plane. Prior efforts exploiting shared memory reduce overhead locally but fall short when scaling across nodes. Furthermore, serverless environments can have unpredictable and large-scale multi-tenancy, leading to contention for shared network resources. We present Palladium, a DPU-centric serverless data plane that reduces the CPU burden and enables efficient, zero-copy communication in multi-tenant serverless clouds. Despite the limited general-purpose processing capability of the DPU cores, Palladium strategically exploits the DPU's potential by (1) offloading data transmission to high-performance NIC cores via RDMA, combined with intra-node shared memory to eliminate data copies across nodes, and (2) enabling cross-processor (CPU-DPU) shared memory to eliminate redundant data movement, which overwhelms wimpy DPU cores. At the core of Palladium is the DPU-enabled network engine (DNE) -- a lightweight reverse proxy that isolates RDMA resources from tenant functions, orchestrates inter-node RDMA flows, and enforces fairness under contention. To further
We propose in this paper to revisit the design of existing encrypted transport protocols to improve their efficiency. We call the methodology ``Reverso'' from reversing the order of field elements within a protocol specification. We detail how such a benign-looking change within the specifications may unlock contiguous zero-copy for encrypted protocols during data transport. To demonstrate our findings, we release \texttt{quiceh}, a QUIC implementation of QUIC VReverso, an extension of the QUIC V1 standard (RFC9000). Our methodology applied to the QUIC protocol reports $\approx 30\%$ of CPU efficiency improvement for processing packets at no added cost on the sender side and without relaxing any security guarantee from QUIC V1. We also implement a fork of Cloudflare's HTTP/3 module and client/server demonstrator using \texttt{quiceh} and show our optimizations to directly transfer to HTTP/3 as well, resulting in our new HTTP/3 to be $\approx 38\%$ more efficient than the baseline implementation using QUIC V1. We argue that Reverso applies to any modern encrypted protocol and its implementations and that similar efficiency improvement can also be unlocked for them, independently of
High performance networks (e.g. Infiniband) rely on zero-copy operations for performance. Zero-copy operations, as the name implies, avoid copying buffers for sending and receiving data. Instead, hardware devices directly read and write to application specified areas of memory. Since these networks can send and receive at nearly the same speed as the memory bus inside machines, zero-copy operations are necessary to achieve peak performance for many applications. Unfortunately, programming with zero-copy APIs is a *giant pain*. Users must carefully avoid using buffers that may be accessed by a device. Typically this either results in spaghetti code (where every access to a buffer is checked before usage), or blocking operations (which pretty much defeat the whole point of zero-copy). We show that by abusing memory protection hardware, we can offer the best of both worlds: a simple zero-copy mechanism which allows for non-blocking send and receives while protecting against incorrect accesses.
Bauplan is a FaaS-based lakehouse specifically built for data pipelines: its execution engine uses Apache Arrow for data passing between the nodes in the DAG. While Arrow is known as the "zero copy format", in practice, limited Linux kernel support for shared memory makes it difficult to avoid copying entirely. In this work, we introduce several new techniques to eliminate nearly all copying from pipelines: in particular, we implement a new kernel module that performs de-anonymization, thus eliminating a copy to intermediate data. We conclude by sharing our preliminary evaluation on different workloads types, as well as discussing our plan for future improvements.
As TLS 1.3 encryption limits traditional Deep Packet Inspection (DPI), the security community has pivoted to Euclidean Transformer-based classifiers (e.g., ET-BERT) for encrypted traffic analysis. However, these models remain vulnerable to byte-level adversarial morphing -- recent pre-padding attacks reduced ET-BERT accuracy to 25.68%, while VLESS Reality bypasses certificate-based detection entirely. We introduce AEGIS: an Adversarial Entropy-Guided Immune System powered by a Thermodynamic Variance-Guided Hyperbolic Liquid State Space Model (TVD-HL-SSM). Rather than competing in the Euclidean payload-reading domain, AEGIS discards payload bytes in favor of 6-dimensional continuous-time flow physics projected into a non-Euclidean Poincare manifold. Liquid Time-Constants measure microsecond IAT decay, and a Thermodynamic Variance Detector computes sequence-wide Shannon Entropy to expose automated C2 tunnel anomalies. A pure C++ eBPF Harvester with zero-copy IPC bypasses the Python GIL, enabling a linear-time O(N) Mamba-3 core to process 64,000-packet swarms at line-rate. Evaluated on a 400GB, 4-tier adversarial corpus spanning backbone traffic, IoT botnets, zero-days, and proprietar
Entanglement can hide in two fundamentally different ways. First, multi-copy correlations can carry information that no single-copy measurement on an unknown state is able to access. Second, bound entangled states possess a positive partial transpose, which makes them invisible to the Peres-Horodecki criterion and all moment inequalities that depend on it. Here we show that the moment difference between the partial transpose and purity decomposes exactly as a chirality-chirality correlator, where the relevant operator is the scalar spin chirality -- the same quantity that governs chiral spin liquids and the topological Hall effect. This decomposition identifies the specific physical structure that multi-copy entanglement detection probes. Using the same controlled-SWAP circuits, we develop a multi-channel spectral classifier for bound entanglement. The classifier combines realignment spectral features with chirality corrections and achieves 99.9% recall at zero false positives across all three known 3x3 bound entangled families, compared with ~40% for the CCNR criterion alone. We also introduce a marginal-noise construction that produces CCNR-invisible bound entangled states, which
A Tychonoff space $X$ has the two-disjoint-copies property (2DCP) if there exists a sequence $(K_n)_{n\inω}$ of non-empty compact subsets of $X$ such that each $K_n$ contains two disjoint subsets homeomorphic to $K_{n+1}$. Banakh, Kąkol and Śliwa showed that 2DCP yields an infinite-dimensional metrizable quotient of $C_p(X)$, while it is still a long-standing open question whether $C_p(X)$ has such a quotient for any infinite compact space $X$. The above concept as well as the last problem are closely related to Efimov's problem that has remained open for 40 years. We will discuss a number of conditions that imply 2DCP. For example, every locally homogeneous compact space, every space containing a copy of $βω$ or $2^ω$ has 2DCP although compact $h$-homogeneous spaces with 2DCP without such copies exist in ZFC. We prove that no scattered compact space has 2DCP and there exist in ZFC compact perfect spaces without 2DCP. This implies that for compact metric spaces $X$ the 2DCP is equivalent to uncountability of $X$. There exist explicit uncountable separable compact spaces failing 2DCP, for example the Isbell-Mrówka compacta. We give positive classes among zero-dimensional compact spa
Copy trading has become the dominant entry strategy in meme coin markets. However, due to the market's extremely illiquid and volatile nature, the strategy exposes an exploitable attack surface: adversaries deploy manipulative bots to front-run trades, conceal positions, and fabricate sentiment, systematically extracting value from naïve copiers at scale. Despite its prevalence, bot-driven manipulation remains largely unexplored, and no robust defensive framework exists. We propose a manipulation-resistant copy-trading system based on a multi-agent architecture powered by a multi-modal large language model (LLM) and chain-of-thought (CoT) reasoning. Our approach outperforms zero-shot and most statistic-driven baselines in prediction accuracy as well as all baselines in economic performance, achieving an average copier return of 3% per meme coin investment under realistic market frictions. Overall, our results demonstrate the effectiveness of agent-based defenses and predictability of trader profitability in adversarial meme coin markets, providing a practical foundation for robust copy trading.
For integers $1\le \ell<k$, the $\ell$-degree Turán density $π_\ell(F)$ measures the minimum $\ell$-degree threshold that forces a copy of a fixed $k$-uniform hypergraph $F$, generalizing both the classical Turán density $π_1$ and the codegree Turán density $π_{k-1}$. Motivated by Erdős' characterization of $k$-graphs with zero Turán density, we study the structural implications of vanishing $\ell$-degree Turán density. Our main result concerns the case $\ell=2$. We prove that, for every $k\ge3$, if a $k$-graph $F$ satisfies $π_2(F)=0$, then $F$ admits a $2$-vanishing order, that is, a global vertex ordering under which all edges align canonically with respect to their pairs. This extends to all uniformities a structural phenomenon previously known for $3$-graphs, and gives a higher-degree analogue of the classical fact that $π_1(F)=0$ forces $F$ to be $k$-partite. In particular, the absence of a $2$-vanishing order is a structural obstruction to vanishing $2$-degree Turán density. We also establish a suspension principle connecting consecutive degree parameters. Given a $(k-1)$-graph $F$, let $\mathcal{S}_F$ be the $k$-graph obtained by adding an apex vertex $v$ and replacing e
We explore the multimetric theory of gravitation, also known as multigravity. We derive additional new exact solutions for the theory in proportional Kerr-Schild and double Kerr-Schild forms. We extend several solutions from the theory of General Relativity, characterized by a constant Ricci scalar in single and double Kerr-Schild forms, to derive solutions in the multi-gravity context. We also examine and extend the classical double copy relations that can be constructed out from these solutions in multigravity exploring the dynamics of the single copy and zero copy fields.
Chain-of-thought (CoT) prompting is necessary for arithmetic in small language models, yet shuffling its steps preserves most performance. What does CoT contribute if not logical sequencing? In three 1-3B instruction-tuned LMs on GSM8K, we isolate the answer-readout stage via prefix completion and identify a positional shortcut: the model copies whichever number occupies the trailing position before the answer delimiter, regardless of intermediate reasoning. Gold-answer presence accounts for 54-92 pp of accuracy (89-92% of each model's teacher-forcing ceiling); even on incorrect items, the final answer matches the last CoT number 95-96% of the time. The copy channel takes precedence over retained-context completion: replacing the trailing number with a wrong value collapses accuracy to near-zero despite correct intermediates, yet removing it recovers 5-32 pp above that floor--even single-step arithmetic the model can otherwise perform is suppressed when a copyable number is present. Qwen and Llama copy novel distractors 87-95% of the time; Gemma gates selectively. Head-level ablation implicates architecture-specific head sets; the effect replicates on GSM-Symbolic. On non-arithmeti
Fine-grained, per-micro-batch load balancing is essential for efficient Mixture-of-Experts (MoE) training, yet every prior dynamic scheduling scheme pays for it with extra communication that is hard to hide. Especially on modern bulk-transfer backends such as DeepEP. We make a simple but consequential observation: on the NVIDIA Hopper architecture the NVLink Copy Engine can move data between intra-node GPUs without consuming any SM cycles, effectively providing a nearly free communication channel that runs in parallel with compute kernels. FEPLB turns this idle hardware into a new parallel dimension for MoE load rebalancing. Its Two-Phase Dispatch first routes tokens across nodes via the standard EP backend, then redistributes dynamic-expert tokens and weights within the NVLink domain through the Copy Engine at nearly zero cost, while a lightweight CPU scheduler runs concurrently with static expert computation. Because FEPLB uses only Copy Engine and CPU that are orthogonal to those consumed by EP and PP, it coexists with existing parallel strategies without reconfiguration. On GLM-5's MoE layers (128 experts, no auxiliary loss, up to 16 H100 GPUs), FEPLB reduces the token straggle
We consider four-dimensional Euclidean Yang-Mills theories quantized in the maximal Abelian and linear covariant gauges at finite temperature. Non-perturbatively, the Faddeev-Popov procedure must be improved to take into account the existence of the so-called Gribov copies. Tapping on previous results about the elimination of infinitesimal Gribov copies in maximal Abelian and linear covariant gauges at zero temperature, we explore the interplay between finite temperature effects and the removal of gauge copies. We focus in a hybrid approach where the thermal masses are derived through perturbative propagators as a stepping stone for a self-consistent treatment. The resulting action collects the effects of the elimination of infinitesimal Gribov copies as well as the thermal masses. We verify the existence of three different phases for the gluonic degrees of freedom; one of complete confinement at low temperatures, an intermediate one of partial confinement, and one of complete deconfinement at high temperatures.
We study subspaces of Orlicz spaces $L_M$ spanned by independent copies $f_k$, $k=1,2,\dots$, of a function $f\in L_M$, $\int_0^1 f(t)\,dt=0$. Any such a subspace $H$ is isomorphic to some Orlicz sequence space $\ell_ψ$. In terms of dilations of the function $f$, a description of strongly embedded subspaces of this type is obtained, and conditions, guaranteeing that the unit ball of such a subspace consists of functions with equicontinuous norms in $L_M$, are found. In particular, we prove that there is a wide class of Orlicz spaces $L_M$ (containing $L^p$-spaces, $1\le p< 2$), for which each of the above properties of $H$ holds if and only if the Matuszewska-Orlicz indices of the functions $M$ and $ψ$ satisfy the inequality: $α_ψ^0>β_M^\infty$.
We study the possibility of deterministic protocols for extracting work from quantum systems. Focusing on the two-point measurement work extraction scenario, we prove that, with enough copies of the system, such zero-fluctuation protocols always exist if the Hamiltonian has a rational spectrum. Leveraging this result, we show that for any Hamiltonian, it is possible to construct an unitary driving protocol on sufficiently many copies of the system with work fluctuations strictly bounded within an arbitrary interval $\pm δ$, albeit requiring exponentially many copies in $1/δ$.