26,098 papers · page 81 of 1,305
Concurrent queues and stacks are important components of many software systems. They are well-known contended data structures due to intense contention on hotspots, resulting in limited performance and poor scalability. To achieve higher performance, several attempts applied the …
Large Language Models (LLMs) have demonstrated remarkable performance in various natural language processing tasks. However, the training of these models is computationally intensive and susceptible to faults, particularly in the attention mechanism, which is a critical component…
Training large language models (LLMs) has become increasingly expensive due to the rapid expansion in model size. Pipeline parallelism is a widely used distributed training technique. However, as LLMs with larger context become prevalent and memory optimization techniques advance…
Large language models have to be trained in parallel due to their large number of parameters and significant memory footprint. Among various parallelism techniques, pipeline parallelism is widely adopted in inter-node scenarios with minimal communication overhead. However, state-…
Deterministic parallelism is a key building block for distributed and fault-tolerant systems that offers substantial performance benefits while guaranteeing determinism. By studying existing deterministically parallel systems (DPS), we identify certain design pitfalls, such as ba…
There are several strategies to parallelize graph neural network (GNN) training over multiple GPUs. We observe that there is no consistent winner (i.e., with the shortest running time), and the optimal strategy depends on the graph dataset, GNN model, training algorithm, and hard…
The n-body problem involves calculating the effect of bodies on each other. n-body simulations are ubiquitous in the fields of physics and astronomy and notoriously computationally expensive. The naïve algorithm for n-body simulations has the prohibiting O(n2) time complexity. Re…
In-memory key-value (KV) caching bridges the performance gap between high-performance networks and disk devices. However, prior in-memory KV caching systems either consider large objects or introduce additional memory overhead. In this paper, we conduct a systematic analysis over…
Breadth First Search (BFS) plays a key role in computational science, networking, and artificial intelligence applications. Although the BFS approach has been extensively studied, particularly in its direction-optimized form, existing implementations still present three main issu…
Molecular dynamics simulation emerges as an important area that HPC+AI helps to investigate the physical properties, with machine-learning interatomic potentials (MLIPs) being used. General-purpose machine-learning (ML) tools have been leveraged in MLIPs, but they are not perfect…
Many concurrent algorithms require processes to perform fetch-and-add operations on a single memory location, which can be a hot spot of contention. We present a novel algorithm called Aggregating Funnels that reduces this contention by spreading the fetch-and-add operations acro…
Transactional Data Structure Programming Systems (TD-SPSs) let programmers compose method invocations on concurrent data structures into coarse-grained, isolated transactions. They detect conflicts among concurrent transactions and recover when operations do not commute. We pres…
Sparse Matrix-matrix Multiplication (SpMM) and Sampled Dense-dense Matrix Multiplication (SDDMM) are important sparse operators in scientific computing and deep learning. Tensor Core Units (TCUs) enhance modern accelerators with superior computing power, which is promising to boo…
Safe memory reclamation techniques that utilize per read reservations, such as hazard pointers and hazard eras, often cause significant overhead in traversals of linked concurrent data structures. This is primarily due to the need to announce a reservation, and fence to make it g…
Second-order optimization methods have been developed to enhance convergence and generalization in deep neural network (DNN) training compared to first-order methods like Stochastic Gradient Descent (SGD). However, these methods face challenges in distributed settings due to high…
Online GNN inference has been widely explored by applications such as online recommendation and financial fraud detection systems, where even minor delays can result in significant financial impact. Real-time dynamic graph sampling enables online GNN inference to reflect the late…
In today's data-driven era, the explosive growth of global data volume has led to an increasing consumption of computing and storage resources. Effective management of virtual machines (VMs) memory usage is critical for cloud vendors to optimize system performance and resource ut…
Tridiagonalization, which is a key step in symmetric eigenvalue decomposition (EVD), aims to convert a symmetric matrix to a tridiagonal form. In Nvidia's cuSOLVER library, the FP64 precision tridiagonalization process only reach 2.1 TFLOPs out of 67 TFLOPs on H100 GPU, and it co…
The Louvain algorithm is one of the most popular algorithms for community detection. Observing that existing implementations suffer from inaccurate pruning and inefficient intermediate state management, we introduce GALA, GPU-Accelerated Louvain Algorithm, which incorporates two …
Scaling blockchain performance through parallel smart contract execution has gained significant attention, as traditional methods remain constrained by the performance of a single virtual machine (VM), even in multi-chain or Layer-2 systems. Parallel VMs offer a compelling soluti…