Related papers: M3R: Increased performance for in-memory Hadoop jo…
With the rapid advancement of Big Data platforms such as Hadoop, Spark, and Dataflow, many tools are being developed that are intended to provide end users with an interactive environment for large-scale data analysis (e.g., IQmulus).…
Most of the prior work on multi-agent reinforcement learning (MARL) achieves optimal collaboration by directly controlling the agents to maximize a common reward. In this paper, we aim to address this from a different angle. In particular,…
Space Cyber-Physical Systems (S-CPS) such as spacecraft and satellites strongly rely on the reliability of onboard computers to guarantee the success of their missions. Relying solely on radiation-hardened technologies is extremely…
Hierarchical matrices provide a powerful representation for significantly reducing the computational complexity associated with dense kernel matrices. For general kernel functions, interpolation-based methods are widely used for the…
Transformer-based large language models (LLM) have been widely used in language processing applications. However, due to the memory constraints of the devices, most of them restrict the context window. Even though recurrent models in…
Training LLMs as interactive agents for multi-turn decision-making remains challenging, particularly in long-horizon tasks with sparse and delayed rewards, where agents must execute extended sequences of actions before receiving meaningful…
Database search and clustering are fundamental components of many data analytics problems, such as mass spectrometry-driven proteomics. Traditional full clustering and search algorithms suffer from high resource usage and long latencies. We…
Growing interest in Artificial Intelligence (AI) has resulted in a surge in demand for faster methods of Machine Learning (ML) model training and inference. This demand for speed has prompted the use of high performance computing (HPC)…
HPC systems keep growing in size to meet the ever-increasing demand for performance and computational resources. Apart from increased performance, large scale systems face two challenges that hinder further growth: energy efficiency and…
An immutable multi-map is a many-to-many thread-friendly map data structure with expected fast insert and lookup operations. This data structure is used for applications processing graphs or many-to-many relations as applied in static…
With the rapid evolution of GPU architectures, the heterogeneity of model training infrastructures is steadily increasing. In such environments, effectively utilizing all available heterogeneous accelerators becomes critical for distributed…
Despite recent advances in feed-forward 3D Gaussian Splatting, generalizable 3D reconstruction remains challenging, particularly in multi-view correspondence modeling. Existing approaches face a fundamental trade-off: explicit methods…
We show that utilizing attribution maps for training neural networks can improve regularization of models and thus increase performance. Regularization is key in deep learning, especially when training complex models on relatively small…
Analyzing and working with big data could be very diffi cult using classical means like relational database management systems or desktop software packages for statistics and visualization. Instead, big data requires large clusters with…
In a large-scale computing cluster, the job completions can be substantially delayed due to two sources of variability, namely, variability in the job size and that in the machine service capacity. To tackle this issue, existing works have…
Matrix-vector multiplication is a fundamental building block in neural networks, vector databases, and large language models, particularly during inference. As a result, efficient matrix-vector multiplication engines directly translate into…
Although LLMs drive automation, it is critical to ensure immense consideration for high-stakes enterprise workflows such as those involving legal matters, risk management, and privacy compliance. For Meta, and other organizations like ours,…
In this work, we consider the integration of MPI one-sided communication and non-blocking I/O in HPC-centric MapReduce frameworks. Using a decoupled strategy, we aim to overlap the Map and Reduce phases of the algorithm by allowing…
MapReduce framework is the de facto standard in Hadoop. Considering the data locality in data centers, the load balancing problem of map tasks is a special case of affinity scheduling problem. There is a huge body of work on affinity…
Many big data algorithms executed on MapReduce-like systems have a shuffle phase that often dominates the overall job execution time. Recent work has demonstrated schemes where the communication load in the shuffle phase can be traded off…