Related papers: Eliminating Multi-GPU Performance Taxes: A Systems…

The rapid development of large language models (LLM) has greatly enhanced everyday applications. While many FPGA-based accelerators, with flexibility for fine-grained data control, exhibit superior speed and energy efficiency compared to…

Heterogeneous computing environments combining CPU and GPU resources provide a great boost to large-scale scientific computing applications. Code generation utilities that partition the work into CPU and GPU tasks while considering data…

There is an urgent and pressing need to optimize usage of Graphical Processing Units (GPUs), which have arguably become one of the most expensive and sought after IT resources. To help with this goal, several of the current generation of…

Efficiently scaling deep neural networks across GPU clusters requires navigating complex trade-offs between computational throughput, memory utilization, and synchronization overhead. This paper presents a unified empirical evaluation of…

High-level synthesis (HLS) enhances digital hardware design productivity through a high abstraction level. Even if the HLS abstraction prevents fine-grained manual register-transfer level (RTL) optimizations, it also enables automatable…

Large language models (LLMs) are computationally intensive. The computation workload and the memory footprint grow quadratically with the dimension (layer width). Most of LLMs' parameters come from the linear layers of the transformer…

As inference workloads for large language models (LLMs) scale to meet growing user demand, pipeline parallelism (PP) has become a widely adopted strategy for multi-GPU deployment, particularly in cross-node setups, to improve key-value (KV)…

B-spline modeling is fundamental to CAD systems, and its evaluation and manipulation algorithms currently in use were developed decades ago, specifically for CPU architectures. While remaining effective for many applications, these…

With the rapid adoption of Large Language Models (LLMs), LLM-adapters have become increasingly common, providing lightweight specialization of large-scale models. Serving hundreds or thousands of these adapters on a single GPU allows…

Deploying large language model (LLM) inference at scale requires jointly selecting base models, provisioning heterogeneous GPUs, configuring parallelism, and distributing workloads under tight latency, accuracy, and budget constraints.…

This paper presents ServerlessLLM, a distributed system designed to support low-latency serverless inference for Large Language Models (LLMs). By harnessing the substantial near-GPU storage and memory capacities of inference servers,…

Data Parallelism (DP), Tensor Parallelism (TP), and Pipeline Parallelism (PP) are the three strategies widely adopted to enable fast and efficient Large Language Model (LLM) training. However, these approaches rely on data-intensive…

Machine learning (ML) models are increasingly trained in clusters with non-dedicated workers possessing heterogeneous resources. In such scenarios, model training efficiency can be negatively affected by stragglers -- workers that run much…

We discuss an approach for solving sparse or dense banded linear systems ${\bf A} {\bf x} = {\bf b}$ on a Graphics Processing Unit (GPU) card. The matrix ${\bf A} \in {\mathbb{R}}^{N \times N}$ is possibly nonsymmetric and moderately large;…

Deploying a large language model (LLM) inference service remains costly because centralized serving depends on specialized GPU clusters and high-bandwidth interconnects in datacenters. An appealing alternative is to leverage collaborative…

Large-scale models rely heavily on 3D parallelism for distributed training, which utilizes tensor parallelism (TP) as the intra-operator parallelism to partition model states across GPUs. However, TP introduces significant communication…

Large deep learning models have demonstrated strong ability to solve many tasks across a wide range of applications. Those large models typically require training and inference to be distributed. Tensor parallelism is a common technique…

Large language models have demonstrated extraordinary performance in many AI tasks but are expensive to use, even after training, due to their requirement of high-end GPUs. Recently, a distributed system called PETALS was developed to lower…

State of art DL models are growing in size and complexity, with many modern models also increasing in heterogeneity of behavior. GPUs are still the dominant platform for DL applications, relying on a bulk-synchronous execution model which…

Binarized Neural Networks (BNNs) significantly reduce the computation and memory demands with binarized weights and activations compared to full-precision NNs. Executing a layer in a BNN on different devices of a heterogeneous…