Computing Systems Group
Research group focussing on Embedded Machine Learning, GPU Computing, HPC and energy efficiency. These are our tools.
Heidelberg, Germany
Pinned Repositories
camuy
Fast evaluation of CNNs on configurable systolic arrays based on abstract metrics
CLAIRE-ROP
Rapid Partitioning-based Deformable Image Registration on Multi-GPU Accelerator
cuda-flux
CUDA Flux is a profiler for GPU applications which reports the basic block executions frequencies of compute kernels
cuda-memtrace
LLVM Plugin to Instrument Global Memory Accesses in CUDA Kernels
gpu-mangrove
machine learning model for execution time and power prediction of CUDA kernels
gpugraph500
A GPU-based Graph500 implementation providing compressed data movements.
grapholator
Simulator for memory access patterns of FPGA-based graph processing accelerators
mekong-cuda
Automatically partitioning compiler for CUDA (WIP) based on the LLVM infrastructure.
MScTI_APC
Accompanying material for course "Advanced Parallel Computing", Institute of Computer Engineering, Ruprecht-Karls University of Heidelberg, Germany
sonar
Trace analysis utility for OTF traces
Computing Systems Group's Repositories
UniHD-CEG/cuda-flux
CUDA Flux is a profiler for GPU applications which reports the basic block executions frequencies of compute kernels
UniHD-CEG/cuda-memtrace
LLVM Plugin to Instrument Global Memory Accesses in CUDA Kernels
UniHD-CEG/mekong-cuda
Automatically partitioning compiler for CUDA (WIP) based on the LLVM infrastructure.
UniHD-CEG/gpugraph500
A GPU-based Graph500 implementation providing compressed data movements.
UniHD-CEG/gpu-mangrove
machine learning model for execution time and power prediction of CUDA kernels
UniHD-CEG/grapholator
Simulator for memory access patterns of FPGA-based graph processing accelerators
UniHD-CEG/MScTI_APC
Accompanying material for course "Advanced Parallel Computing", Institute of Computer Engineering, Ruprecht-Karls University of Heidelberg, Germany
UniHD-CEG/camuy
Fast evaluation of CNNs on configurable systolic arrays based on abstract metrics
UniHD-CEG/CLAIRE-ROP
Rapid Partitioning-based Deformable Image Registration on Multi-GPU Accelerator
UniHD-CEG/arm-peak
Measure computational peak performance on embedded ARM processors.
UniHD-CEG/DeepHYDRA
UniHD-CEG/galen
Galen: Hardware-specific Automatic Compression of Neural Networks
UniHD-CEG/gtod-resolution
Simple tool to measure the average and worst case resolution of the gettimeofday call.
UniHD-CEG/sonar
Trace analysis utility for OTF traces
UniHD-CEG/walking-noise
UniHD-CEG/CUDAsap
UniHD-CEG/ECML2018
UniHD-CEG/gpu-mangrove-tutorial
Tutorial files for ''Instrumentation and Modeling of Performance and Power Consumption for Massively Parallel Processors'' at HiPEAC 2021 Conference -- https://www.hipeac.net/2021/spring-virtual/#/program/sessions/7856/
UniHD-CEG/interval_ranges
A no-dependency python library (with okayish performance) for sets of interval ranges
UniHD-CEG/llm-brain-damage-experiment
UniHD-CEG/torchprofilingutils
UniHD-CEG/ZCU216-PYNQ
Build repo for PYNQ on the ZCU216 RFSOC