[ASPLOS'26] HILOS: A Cost-Effective Near-Storage Processing Solution for Offline Inference of Long-Context LLMs

Latr: Lazy Translation Coherence - ASPLOS'18

NVDLA (An Opensource DL Accelerator Framework) implementation on FPGA.

NPUWattch: ML-based Power, Area, and Timing Modeling for Neural Accelerators

FPGA Virtualization and NMU Source

Open-source AI acceleration on FPGA: from ONNX to RTL

Official repository of the paper: Detection of LLM-Paraphrased Code and Identification of the Responsible LLM Using Coding Style Features (To appear in Engineering Applications of Artificial Intell…

A modular program analysis tool framework for accelerators (NVIDIA, AMD, and DL workloads).

Inline i18n for JavaScript/TypeScript. Write translations where you use them.

This is the public repo for DREAM accepted to ICS 2025.

Fully homomorphic encryption compiler for CKKS

Synthesisable SIMT-style RISC-V GPGPU

LoCaLUT_AE

Original Proof-of-Concepts for React2Shell CVE-2025-55182

Preemption-Enhanced Benchmark Suite for FPGAs

Simple Python Module to access PCIe Endpoint BARs

Vortex Graphics

Artifacts for pact25 paper

OPS is an API with associated libraries and preprocessors to generate parallel executables for applications on mulit-block structured meshes.

The framework for the paper "ScaleMoE: A Fast and Scalable Distributed Training Framework for Large-Scale Mixture-of-Experts Models" in PACT 2025.

This repository contains examples of DRAMA reverse-engineering and side-channel attacks