no code implementations • 27 May 2024 • Zhenyu Bai, Pranav Dangi, Huize Li, Tulika Mitra
In response, we propose a dataflow-aware FPGA-based accelerator design, SWAT, that efficiently leverages the sparsity to achieve scalable performance for long input.
1 code implementation • 24 Nov 2023 • Shivam Aggarwal, Kuluhan Binici, Tulika Mitra
Machine learning pipelines for classification tasks often train a universal model to achieve accuracy across a broad range of classes.
no code implementations • 21 Nov 2023 • Shivam Aggarwal, Alessandro Pappalardo, Hans Jakob Damsgaard, Giuseppe Franco, Thomas B. Preußer, Michaela Blott, Tulika Mitra
However, the exploration of floating-point formats smaller than 8 bits and their comparison with integer quantization remains relatively limited.
no code implementations • 20 Sep 2023 • Dan Wu, Zhaoying Li, Tulika Mitra
Our experiments with three GNN models on four large graphs demonstrate that InkStream accelerates by 2. 5-427$\times$ on a CPU cluster and 2. 4-343$\times$ on two different GPU clusters while producing identical outputs as GNN model inference on the latest graph snapshot.
1 code implementation • 12 Sep 2023 • Dhananjaya Wijerathne, Zhaoying Li, Tulika Mitra
Coarse-Grained Reconfigurable Arrays (CGRAs) hold great promise as power-efficient edge accelerator, offering versatility beyond AI applications.
1 code implementation • 9 Jan 2022 • Kuluhan Binici, Shivam Aggarwal, Nam Trung Pham, Karianto Leman, Tulika Mitra
In particular, we design a Variational Autoencoder (VAE) with a training objective that is customized to learn the synthetic data representations optimally.
no code implementations • 11 Aug 2021 • Kuluhan Binici, Nam Trung Pham, Tulika Mitra, Karianto Leman
Moreover, the sample generation strategies in some of these methods could result in a mismatch between the synthetic and real data distributions.
1 code implementation • 2 Sep 2019 • Guanhua Wang, Sudipta Chattopadhyay, Arnab Kumar Biswas, Tulika Mitra, Abhik Roychoudhury
Spectre attacks disclosed in early 2018 expose data leakage scenarios via cache side channels.
Cryptography and Security
no code implementations • 24 Aug 2019 • Siqi Wang, Anuj Pathania, Tulika Mitra
Mobile devices are empowered with heterogeneous multi-processor Systems-on-Chips (SoCs) to process ML workloads such as Convolutional Neural Network (CNN) inference.
no code implementations • 14 Mar 2019 • Siqi Wang, Gayathri Ananthanarayanan, Yifan Zeng, Neeraj Goel, Anuj Pathania, Tulika Mitra
IoT Edge intelligence requires Convolutional Neural Network (CNN) inference to take place in the edge devices itself.
2 code implementations • 16 Jul 2018 • Guanhua Wang, Sudipta Chattopadhyay, Ivan Gotovchits, Tulika Mitra, Abhik Roychoudhury
In this paper, we propose oo7, a static analysis approach that can mitigate Spectre attacks by detecting potentially vulnerable code snippets in program binaries and protecting them against the attack by patching them.
Cryptography and Security
no code implementations • 28 Mar 2018 • Guanwen Zhong, Akshat Dubey, Tan Cheng, Tulika Mitra
Convolutional Neural Networks (CNN) have been widely deployed in diverse application domains.