arXiv subjects

Shay Vargaftik

Publications and source records attributed to Shay Vargaftik.

At least 19 recordsLinked to original sources

arXiv2026

Entropy-Constrained Adaptive Stochastic Quantization

Adaptive stochastic quantization (ASQ) is a recently introduced quantization approach that optimizes the Mean Squared Error (MSE) for a given input while preserving unbiasedness. It is designed to alleviate the communication and memory bottlenecks of modern data and machine learning workloads, including model, gradient, and KV-cache compression and nearest-neighbor search. Further, practical systems can then compress quantized data with a lossless entropy encoder. However, existing unbiased methods, including ASQ, choose their quantization values without considering this later encoding stage, leaving accuracy on the table. We formulate the Entropy Constrained Adaptive Stochastic Quantization (ECASQ) problem, which jointly selects adaptive quantization values to minimize MSE under an entropy budget and an unbiasedness constraint. We give an optimal dynamic program with $O(sd^2)$ time and $O(d^2)$ space for a length-d vector and at most s quantization values, and a GPU-friendly approximate dynamic program with $O(sd^2)$ time and $O(d)$ space. The approximation guarantees that the solution has an MSE no larger than the optimal solution that uses one fewer bit of entropy per entry. We also provide an iterative refinement procedure for the approximation solution that, in our experiments, yields near-optimal results while retaining a substantial speed advantage over our solver for the optimal solution.

cs.LG↗

arXiv2026

Quantizing With Randomized Hadamard Transforms: Efficient Heuristic Now Proven

Uniform random rotations (URRs) are a common preprocessing step in modern quantization approaches used for gradient compression, inference acceleration, KV-cache compression, model weight quantization, and approximate nearest-neighbor search in vector databases. In practice, URRs are often replaced by randomized Hadamard transforms (RHTs), which preserve orthogonality while admitting fast implementations. The remaining issue is the performance for worst-case inputs. With a URR, each coordinate is individually distributed as a shifted beta distribution, which converges to a Gaussian distribution in high dimensions. Generally, one RHT is not suitable in the worst case, as individual coordinates can be far from these distributions. We show that after composing two RHTs on any $d$-sized input vector, the marginal distribution of every fixed coordinate of the normalized rotated vector is within $O(d^{-1/2})$ of a standard Gaussian both in Kolmogorov distance and in $1$-Wasserstein distance. We then plug these bounds into the analyses of modern compression schemes, namely DRIVE and QUIC-FL, and show that two RHTs achieve performance that asymptotically matches URRs. However, we show that two RHTs may not be sufficient for Vector Quantization (VQ), which often requires weak correlation across fixed-size blocks of coordinates (as opposed to only marginal distribution convergence for single coordinates). We prove that a composition of three RHTs leads to decaying coordinate covariance. This ensures that any fixed, bounded, multi-dimensional VQ codebook optimized for URRs has the same expected error when using three RHTs, up to an additive term that vanishes with the dimension. Finally, because practical inputs are rarely adversarial, we propose a linear-time ${O}(d)$ check on the input's moments to dynamically adapt the number of RHTs used at runtime to improve performance.

cs.LG↗

arXiv2026

A Note on TurboQuant and the Earlier DRIVE/EDEN Line of Work

This note clarifies the relationship between the recent TurboQuant work and the earlier DRIVE (NeurIPS 2021) and EDEN (ICML 2022) schemes. DRIVE is a 1-bit quantizer that EDEN extended to any $b>0$ bits per coordinate; we refer to them collectively as EDEN. First, TurboQuant$_{\text{mse}}$ is a special case of EDEN obtained by fixing EDEN's scalar scale parameter to $S=1$. EDEN supports both biased and unbiased quantization, each optimized by a different $S$ (chosen via methods described in the EDEN works). The fixed choice $S=1$ used by TurboQuant is generally suboptimal, although the optimal $S$ for biased EDEN converges to $1$ as the dimension grows; accordingly TurboQuant$_{\text{mse}}$ approaches EDEN's behavior for large $d$. Second, TurboQuant$_{\text{prod}}$ combines a biased $(b-1)$-bit EDEN step with an unbiased 1-bit QJL quantization of the residual. It is suboptimal in three ways: (1) its $(b-1)$-bit step uses the suboptimal $S=1$; (2) its 1-bit unbiased residual quantization has worse MSE than (unbiased) 1-bit EDEN; (3) chaining a biased $(b-1)$-bit step with a 1-bit unbiased residual step is inferior to unbiasedly quantizing the input directly with $b$-bit EDEN. Third, some of the analysis in the TurboQuant work mirrors that of the EDEN works: both exploit the connection between random rotations and the shifted Beta distribution, use the Lloyd-Max algorithm, and note that Randomized Hadamard Transforms can replace uniform random rotations. Experiments support these claims: biased EDEN (with optimized $S$) is more accurate than TurboQuant$_{\text{mse}}$, and unbiased EDEN is markedly more accurate than TurboQuant$_{\text{prod}}$, often by more than a bit (e.g., 2-bit EDEN beats 3-bit TurboQuant$_{\text{prod}}$). We also repeat all accuracy experiments from the TurboQuant paper, showing that EDEN outperforms it in every setup we have tried.

cs.LG↗

arXiv2026

DynamiQ: Accelerating Gradient Synchronization using Compressed Multi-hop All-reduce

Multi-hop all-reduce is the de facto backbone of large model training. As the training scale increases, the network often becomes a bottleneck, motivating the reduction of the volume of transmitted data. Accordingly, recent systems have demonstrated significant acceleration of the training process using gradient quantization. However, these systems are not optimized for multi-hop aggregation, where entries are partially summed multiple times along their aggregation topology. We present DynamiQ, a quantization framework that bridges the gap between quantization best practices and multi-hop aggregation. DynamiQ introduces novel techniques to better represent partial sums, codesigned with a decompress accumulate recompress fused kernel to facilitate fast execution. We extend PyTorch DDP to support DynamiQ over NCCL P2P, and across different LLMs, tasks, and scales, we demonstrate consistent improvement of up to 34.2% over the best among state-of-the-art methods such as Omni-Reduce, THC, and emerging standards such as MXFP4, MXFP6, and MXFP8. Further, DynamiQ is the only evaluated method that consistently reaches near-baseline accuracy (e.g., 99.9\% of the BF16 baseline) and does so while significantly accelerating the training.

cs.LG↗

arXiv2025

OptiNIC: A Resilient and Tail-Optimal RDMA NIC for Distributed ML Workloads

As distributed machine learning (ML) workloads scale to thousands of GPUs connected by high-speed interconnects, tail latency in collective communication has become a major bottleneck. Existing RDMA transports, such as RoCE, IRN, SRNIC, and Falcon, enforce strict reliability and in-order delivery, relying on retransmissions and packet sequencing to ensure correctness. While these approaches work well for general-purpose workloads, they introduce complexity and latency that scale poorly in ML, where even rare packet delays can stall entire model pipelines. We present OptiNIC, a domain-specific RDMA transport that revisits traditional reliability guarantees based on ML's tolerance for partial or missing data. OptiNIC eliminates retransmissions and in-order delivery from the NIC, enabling a best-effort, out-of-order transport model for RDMA. Unlike traditional RDMA, which signals completion only after complete data delivery, OptiNIC introduces adaptive timeouts to trigger forward progress when data may be lost or delayed. OptiNIC retains standard congestion control mechanisms (e.g., DCQCN, EQDS, or Swift) while shifting loss recovery to the ML pipeline itself (e.g., via the Hadamard Transform and Erasure Coding). Our evaluation shows that OptiNIC improves time-to-accuracy (TTA) by 2x and increases throughput by 1.6x for training and inference, respectively, across two public clouds (i.e., Hyperstack and CloudLab). OptiNIC also lowers 99th-percentile latency by 3.5x, cuts BRAM usage by 2.7x, and nearly doubles NIC resilience to faults-delivering a resilient, tail-optimized RDMA transport purpose-built for distributed ML workloads.

cs.DC↗

arXiv2025

Reimagining RDMA Through the Lens of ML

As distributed machine learning (ML) workloads scale to thousands of GPUs connected by ultra-high-speed inter-connects, tail latency in collective communication has emerged as a primary bottleneck. Prior RDMA designs, like RoCE, IRN, and SRNIC, enforce strict reliability and in-order delivery, relying on retransmissions and packet sequencing to ensure correctness. While effective for general-purpose workloads, these mechanisms introduce complexity and latency that scale poorly, where even rare packet losses or delays can consistently degrade system performance. We introduce Celeris, a domain-specific RDMA transport that revisits traditional reliability guarantees based on ML's tolerance for lost or partial data. Celeris removes retransmissions and in-order delivery from the RDMA NIC, enabling best-effort transport that exploits the robustness of ML workloads. It retains congestion control (e.g., DCQCN) and manages communication with software-level mechanisms such as adaptive timeouts and data prioritization, while shifting loss recovery to the ML pipeline (e.g., using the Hadamard Transform). Early results show that Celeris reduces 99th-percentile latency by up to 2.3x, cuts BRAM usage by 67%, and nearly doubles NIC resilience to faults -- delivering a resilient, scalable transport tailored for ML at cluster scale.

cs.DC↗

arXiv2025

Optimal and Near-Optimal Adaptive Vector Quantization

Quantization is a fundamental optimization for many machine-learning use cases, including compressing gradients, model weights and activations, and datasets. The most accurate form of quantization is \emph{adaptive}, where the error is minimized with respect to a given input, rather than optimizing for the worst case. However, optimal adaptive quantization methods are considered infeasible in terms of both their runtime and memory requirements. We revisit the Adaptive Vector Quantization (AVQ) problem and present algorithms that find optimal solutions with asymptotically improved time and space complexity. We also present an even faster near-optimal algorithm for large inputs. Our experiments show our algorithms may open the door to using AVQ more extensively in a variety of machine learning applications.

cs.LG↗

arXiv2025

OptiReduce: Resilient and Tail-Optimal AllReduce for Distributed Deep Learning in the Cloud

We present OptiReduce, a new collective-communication system for the cloud with bounded, predictable completion times for deep-learning jobs in the presence of varying computation (stragglers) and communication (congestion and gradient drops) variabilities. OptiReduce exploits the inherent resiliency and the stochastic nature of distributed deep-learning (DDL) training and fine-tuning to work with approximated (or lost) gradients -- providing an efficient balance between (tail) performance and the resulting accuracy of the trained models. Exploiting this domain-specific characteristic of DDL, OptiReduce introduces (1) mechanisms (e.g., unreliable bounded transport with adaptive timeout) to improve the DDL jobs' tail execution time, and (2) strategies (e.g., Transpose AllReduce and Hadamard Transform) to mitigate the impact of gradient drops on model accuracy. Our evaluation shows that OptiReduce achieves 70% and 30% faster time-to-accuracy (TTA), on average, when operating in shared, cloud environments (e.g., CloudLab) compared to Gloo and NCCL, respectively.

cs.DC↗

arXiv2025

Counter Pools: Counter Representation for Efficient Stream Processing

Due to the large data volume and number of distinct elements, space is often the bottleneck of many stream processing systems. The data structures used by these systems often consist of counters whose optimization yields significant memory savings. The challenge lies in balancing the size of the counters: too small, and they overflow; too large, and memory capacity limits their number. In this work, we suggest an efficient encoding scheme that sizes each counter according to its needs. Our approach uses fixed-sized pools of memory (e.g., a single memory word or 64 bits), where each pool manages a small number of counters. We pay special attention to performance and demonstrate considerable improvements for various streaming algorithms and workload characteristics.

cs.NI↗

arXiv2025

HACK: Homomorphic Acceleration via Compression of the Key-Value Cache for Disaggregated LLM Inference

Disaggregated Large Language Model (LLM) inference has gained popularity as it separates the computation-intensive prefill stage from the memory-intensive decode stage, avoiding the prefill-decode interference and improving resource utilization. However, transmitting Key-Value (KV) data between the two stages can be a bottleneck, especially for long prompts. Additionally, the computation time overhead for prefill and decode is key for optimizing Job Completion Time (JCT), and KV data size can become prohibitive for long prompts and sequences. Existing KV quantization methods can alleviate the transmission bottleneck and reduce memory requirements, but they introduce significant dequantization overhead, exacerbating the computation time. We propose Homomorphic Acceleration via Compression of the KV cache (HACK) for disaggregated LLM inference. HACK eliminates the heavy KV dequantization step, and directly performs computations on quantized KV data to approximate and reduce the cost of the expensive matrix-multiplication step. Extensive trace-driven experiments show that HACK reduces JCT by up to 70.9% compared to disaggregated LLM inference baseline and by up to 52.3% compared to state-of-the-art KV quantization methods.

cs.DC↗

arXiv2024

Beyond Throughput and Compression Ratios: Towards High End-to-end Utility of Gradient Compression

Gradient aggregation has long been identified as a major bottleneck in today's large-scale distributed machine learning training systems. One promising solution to mitigate such bottlenecks is gradient compression, directly reducing communicated gradient data volume. However, in practice, many gradient compression schemes do not achieve acceleration of the training process while also preserving accuracy. In this work, we identify common issues in previous gradient compression systems and evaluation methodologies. These include excessive computational overheads; incompatibility with all-reduce; and insufficient evaluation methods, such as not using an end-to-end metric or using a 32-bit baseline instead of the stronger 16-bit baseline. We revisit common compression approaches (sparsification, quantization, and low-rank decomposition) and demonstrate how considering the above issues can lead to minor but strategic design changes, resulting in notably better performance. Our goal is to raise awareness of the need for design and evaluation standards that naturally translate to the end-to-end utility of gradient compression.

cs.LG↗

arXiv2024

Lucy: Think and Reason to Solve Text-to-SQL

Large Language Models (LLMs) have made significant progress in assisting users to query databases in natural language. While LLM-based techniques provide state-of-the-art results on many standard benchmarks, their performance significantly drops when applied to large enterprise databases. The reason is that these databases have a large number of tables with complex relationships that are challenging for LLMs to reason about. We analyze challenges that LLMs face in these settings and propose a new solution that combines the power of LLMs in understanding questions with automated reasoning techniques to handle complex database constraints. Based on these ideas, we have developed a new framework that outperforms state-of-the-art techniques in zero-shot text-to-SQL on complex benchmarks

cs.AI↗

arXiv2024

ScionFL: Efficient and Robust Secure Quantized Aggregation

Secure aggregation is commonly used in federated learning (FL) to alleviate privacy concerns related to the central aggregator seeing all parameter updates in the clear. Unfortunately, most existing secure aggregation schemes ignore two critical orthogonal research directions that aim to (i) significantly reduce client-server communication and (ii) mitigate the impact of malicious clients. However, both of these additional properties are essential to facilitate cross-device FL with thousands or even millions of (mobile) participants. In this paper, we unite both research directions by introducing ScionFL, the first secure aggregation framework for FL that operates efficiently on quantized inputs and simultaneously provides robustness against malicious clients. Our framework leverages (novel) multi-party computation (MPC) techniques and supports multiple linear (1-bit) quantization schemes, including ones that utilize the randomized Hadamard transform and Kashin's representation. Our theoretical results are supported by extensive evaluations. We show that with no overhead for clients and moderate overhead for the server compared to transferring and processing quantized updates in plaintext, we obtain comparable accuracy for standard FL benchmarks. Moreover, we demonstrate the robustness of our framework against state-of-the-art poisoning attacks.

cs.CR↗

arXiv2024

THC: Accelerating Distributed Deep Learning Using Tensor Homomorphic Compression

Deep neural networks (DNNs) are the de facto standard for essential use cases, such as image classification, computer vision, and natural language processing. As DNNs and datasets get larger, they require distributed training on increasingly larger clusters. A main bottleneck is the resulting communication overhead where workers exchange model updates (i.e., gradients) on a per-round basis. To address this bottleneck and accelerate training, a widely-deployed approach is compression. However, previous deployments often apply bi-directional compression schemes by simply using a uni-directional gradient compression scheme in each direction. This results in significant computational overheads at the parameter server and increased compression error, leading to longer training and lower accuracy. We introduce Tensor Homomorphic Compression (THC), a novel bi-directional compression framework that enables the direct aggregation of compressed values and thus eliminating the aforementioned computational overheads. Moreover, THC is compatible with in-network aggregation (INA), which allows for further acceleration. Our evaluation shows that training representative vision and language models with THC reaches target accuracy by 1.40x to 1.47x faster using INA and 1.28x to 1.33x faster using a software PS compared with state-of-the-art systems.

cs.LG↗

arXiv2023

DoCoFL: Downlink Compression for Cross-Device Federated Learning

Many compression techniques have been proposed to reduce the communication overhead of Federated Learning training procedures. However, these are typically designed for compressing model updates, which are expected to decay throughout training. As a result, such methods are inapplicable to downlink (i.e., from the parameter server to clients) compression in the cross-device setting, where heterogeneous clients $\textit{may appear only once}$ during training and thus must download the model parameters. Accordingly, we propose $\textsf{DoCoFL}$ -- a new framework for downlink compression in the cross-device setting. Importantly, $\textsf{DoCoFL}$ can be seamlessly combined with many uplink compression schemes, rendering it suitable for bi-directional compression. Through extensive evaluation, we show that $\textsf{DoCoFL}$ offers significant bi-directional bandwidth reduction while achieving competitive accuracy to that of a baseline without any compression.

cs.LG↗

arXiv2023

QUIC-FL: Quick Unbiased Compression for Federated Learning

Distributed Mean Estimation (DME), in which $n$ clients communicate vectors to a parameter server that estimates their average, is a fundamental building block in communication-efficient federated learning. In this paper, we improve on previous DME techniques that achieve the optimal $O(1/n)$ Normalized Mean Squared Error (NMSE) guarantee by asymptotically improving the complexity for either encoding or decoding (or both). To achieve this, we formalize the problem in a novel way that allows us to use off-the-shelf mathematical solvers to design the quantization.

cs.LG↗

arXiv2022

Distributed Dispatching in the Parallel Server Model

With the rapid increase in the size and volume of cloud services and data centers, architectures with multiple job dispatchers are quickly becoming the norm. Load balancing is a key element of such systems. Nevertheless, current solutions to load balancing in such systems admit a paradoxical behavior in which more accurate information regarding server queue lengths degrades performance due to herding and detrimental incast effects. Indeed, both in theory and in practice, there is a common doubt regarding the value of information in the context of multi-dispatcher load balancing. As a result, both researchers and system designers resort to more straightforward solutions, such as the power-of-two-choices to avoid worst-case scenarios, potentially sacrificing overall resource utilization and system performance. A principal focus of our investigation concerns the value of information about queue lengths in the multi-dispatcher setting. We argue that, at its core, load balancing with multiple dispatchers is a distributed computing task. In that light, we propose a new job dispatching approach, called Tidal Water Filling, which addresses the distributed nature of the system. Specifically, by incorporating the existence of other dispatchers into the decision-making process, our protocols outperform previous solutions in many scenarios. In particular, when the dispatchers have complete and accurate information regarding the server queues, our policies significantly outperform all existing solutions.

cs.DC↗

arXiv2022

IIsy: Practical In-Network Classification

The rat race between user-generated data and data-processing systems is currently won by data. The increased use of machine learning leads to further increase in processing requirements, while data volume keeps growing. To win the race, machine learning needs to be applied to the data as it goes through the network. In-network classification of data can reduce the load on servers, reduce response time and increase scalability. In this paper, we introduce IIsy, implementing machine learning classification models in a hybrid fashion using off-the-shelf network devices. IIsy targets three main challenges of in-network classification: (i) mapping classification models to network devices (ii) extracting the required features and (iii) addressing resource and functionality constraints. IIsy supports a range of traditional and ensemble machine learning models, scaling independently of the number of stages in a switch pipeline. Moreover, we demonstrate the use of IIsy for hybrid classification, where a small model is implemented on a switch and a large model at the backend, achieving near optimal classification results, while significantly reducing latency and load on the servers.

cs.NI↗