Approximate multipliers can reduce hardware area and energy consumption in Deep Neural Network (DNN) inference; however, they introduce computational errors. Assessing the accuracy of numerous approxi...

Deploying Vision Transformer (ViT) models on edge platforms remains challenging due to their high computational demands and the architectural heterogeneity of modern hybrid ViT models, which incorpora...

Quantization of Deep Neural Network (DNN) activations is a commonly used technique to reduce compute and memory demands during DNN inference, which can be particularly beneficial on...

As custom hardware accelerators become increasingly central to machine learning workloads, efficient data transfer is critical for maximizing accelerator performance on linear algebra kernels. AXI4MLI...

The FPL 2026 conference app.

Power-of-two (PoT) quantization significantly reduces the size of Deep Neural Networks (DNNs) and allows the use of bit-shifts as core operations instead of multiplications for DNN inference. Previous...

Fully homomorphic encryption (FHE) has recently attracted significant attention as both a cryptographic primitive and a systems challenge. Given the latest advances in accelerated computing, FHE prese...

AccML workshop homepage

7th C4ML workshop, at CGO 2026 Sunday, February 1, 2026 In person: Sydney, Australia

AccML workshop homepage

AccML workshop homepage

Large Language Models (LLMs) have become increasingly prominent for daily tasks, from improving sound-totext translation to generating additional frames for the latest video games. With the help of LL...

AccML workshop homepage

23rd International Workshop on Algorithms, Models and Tools for Parallel Computing on Heterogeneous Platforms

23rd International Workshop on Algorithms, Models and Tools for Parallel Computing on Heterogeneous Platforms

Pruning is a widely used method for compressing Deep Neural Networks (DNNs), where less relevant parameters are removed from a DNN model to reduce its size. However, removing parameters reduces model ...

23rd International Workshop on Algorithms, Models and Tools for Parallel Computing on Heterogeneous Platforms

The deployment of deep neural networks (DNNs) in privacy-sensitive environments is constrained by computational overheads in fully homomorphic encryption (FHE). This paper explores unstructured sparsi...

The 5th Workshop on Machine Learning and Systems (EuroMLSys)