Related papers: Stripe: Tensor Compilation via the Nested Polyhedr…

Modern recommendation models have increased to trillions of parameters. As cluster scales expand to O(1k), distributed training bottlenecks shift from computation and memory to data movement, especially lookup and communication latency…

Mixed-integer programming (MIP) is a well-established framework for computer-aided molecular design (CAMD). By precisely encoding the molecular space and score functions, e.g., a graph neural network, the molecular design problem is…

During the past decade, Deep Learning (DL) algorithms, programming systems and hardware have converged with the High Performance Computing (HPC) counterparts. Nevertheless, the programming methodology of DL and HPC systems is stagnant,…

An optimizing compiler consists of a front end parsing a textual programming language into an intermediate representation (IR), a middle end performing optimizations on the IR, and a back end lowering the IR to a target representation (TR)…

Mixed Integer Linear Programs (MILPs) are highly flexible and powerful tools for modeling and solving complex real-world combinatorial optimization problems. Recently, machine learning (ML)-guided approaches have demonstrated significant…

Neural network (NN) accelerators with multi-chip-module (MCM) architectures enable integration of massive computation capability; however, they face challenges of computing resource underutilization and off-chip communication overheads.…

Dense linear algebra kernels are critical for wireless applications, and the oncoming proliferation of 5G only amplifies their importance. Many such matrix algorithms are inductive, and exhibit ample amounts of fine-grain ordered…

Fast machine code generation is especially important for fast start-up just-in-time compilation, where the compilation time is part of the end-to-end latency. However, widely used compiler frameworks like LLVM do not prioritize fast…

Matrix-accelerated stencil computation is a hot research topic, yet its application to three-dimensional (3D) high-order stencils and HPC remains underexplored. With the emergence of matrix units on multicore CPUs, we analyze matrix-based…

As deep learning becomes more expensive, both in terms of time and compute, inefficiencies in machine learning (ML) training prevent practical usage of state-of-the-art models for most users. The newest model architectures are simply too…

Large Language Models (LLMs) with Mixture-of-Expert (MoE) architectures achieve superior model performance with reduced computation costs, but at the cost of high memory capacity and bandwidth requirements. Near-Memory Processing (NMP)…

Modern polyhedral compilers excel at aggressively optimizing codes with static control parts, but the state-of-practice to find high-performance polyhedral transformations especially for different hardware targets still largely involves…

Linear Sequence Modeling (LSM) like linear attention, state space models and linear RNNs, and Mixture-of-Experts (MoE) have recently emerged as significant architectural improvements. In this paper, we introduce Linear-MoE, a…

Mixed-integer linear programming (MILP) is widely employed for modeling combinatorial optimization problems. In practice, similar MILP instances with only coefficient variations are routinely solved, and machine learning (ML) algorithms are…

Distributed inference of large language models (LLMs) using tensor parallelism can introduce communication overheads of $20$% even over GPUs connected via NVLink, a high-speed GPU interconnect. Several techniques have been proposed to…

Deep Neural Networks (DNNs) have revolutionized many aspects of our lives. The use of DNNs is becoming ubiquitous including in softwares for image recognition, speech recognition, speech synthesis, language translation, to name a few. he…

Sparse triangular solve (SpTRSV) is an extensively studied computational kernel. An important obstacle in parallel SpTRSV implementations is that in some parts of a sparse matrix the computation is serial. By transforming the dependency…

In this paper, we evaluate the performance of various parallel optimization methods for Kernel Support Vector Machines on multicore CPUs and GPUs. In particular, we provide the first comparison of algorithms with explicit and implicit…

The rapid growth of deep learning models has increased the demand for efficient distributed training strategies. Fully sharded approaches like ZeRO-3 and FSDP partition model parameters across GPUs and apply optimizations such as…

Over the past few decades, neuroscience experiments have become increasingly complex and naturalistic. Experimental design has in turn become more challenging, as experiments must conform to an ever-increasing diversity of design…