OpenCores

FlexGripPlus General Purpose Graphics Processing Unit (GPGPU) core

Project maintainers

Details

Name: flexgripplus
Created: Apr 11, 2020
Updated: Sep 15, 2020
SVN Updated: Sep 15, 2020
SVN: Browse
Latest version: download (might take a bit to start...)
Statistics: View
Bugs: 1 reported / 1 solved
Star2you like it: star it!

Other project properties

Category:Video controller
Language:VHDL
Development status:Mature
Additional info:FPGA proven
WishBone compliant: No
WishBone version: n/a
License: LGPL

Description

FlexGripPLus is an open-source General Purpose Graphics Processing Units (soft-GPGPU) model based on the NVIDIA-G80 microarchitecture and described in VHDL. The original version of the GPGPU model was developed by the University of Massachusetts and was designed to be fully compatible with the CUDA programming environment using the SM 1.0 compatibility.

The FlexGripPlus model is an extension of the original GPGPU with multiple improvements and based on the description of a Streaming Multiprocessor (SM) module and has the support of 28 assembly (SASS) instructions.

The internal description of the SM is mainly composed of five stages of pipeline (Fetch, Decode, Read, Execute, and Write-Back) following the descriptions of the G80 microarchitecture Tesla. The total number of SP in the execute stage is selectable before synthesis among 8, 16, and 32 parallel cores.

Please feel free to use and modify.

Additional descriptions and suppport can be found in:

J. E. R. Condia, B. Du, M. Sonza Reorda and L. Sterpone, "FlexGripPlus: An improved GPGPU model to support reliability analysis," Microelectronics Reliability 109, 113660, 2020

Other documents:

B. Du, J. E. R. Condia and M. S. Reorda, "An extended model to support detailed GPGPU reliability analysis," 2019 14th International Conference on Design & Technology of Integrated Systems In Nanoscale Era (DTIS), 2019

B. Du, J. E. R. Condia, M. S. Reorda and L. Sterpone, "About the functional test of the GPGPU scheduler," 2018 IEEE 24th International Symposium on On-Line Testing And Robust System Design (IOLTS), 2018

B. Du, J. E. Rodriguez Condia, M. Sonza Reorda and L. Sterpone, "On the evaluation of SEU effects in GPGPUs," 2019 IEEE Latin American Test Symposium (LATS), 2019

K. Andryc, M. Merchant and R. Tessier, "FlexGrip: A soft GPGPU for FPGAs," 2013 International Conference on Field-Programmable Technology (FPT)

Andryc, Kevin, Tedy Thomas, and Russell Tessier. "Soft GPGPUs for embedded FPGAs: An architectural evaluation.", 2016.

The GPGPU model can be simulated and sinthesized in multiple simulators and platforms.

The latest version can be simulated in multiple environments including ModelSim, QuestaSim, Xcelium Parallel Simulator, ZamiaCAD and ModelSim-Altera Starter Edition.

Full download of the project, user manuals and programmer manuals can be consulted and downloaded from:

Programmer's manual is available in: Manual

User's Manual is available in: Coming Soon

User's Starting Guide is available in: Coming Soon

https://github.com/Jerc007/Open-GPGPU-FlexGrip-

https://opencores.org/websvn/listing/flexgripplus/flexgripplus

The FlexGripPlus model was developed in the CAD group of Politecnico di Torino, Turin, Italy and Supported with funding by the European Comission through the Horizon 2020 RESCUE-ETN project under grant 722325. For more information: http://rescue-etn.eu/

Rescue

Authors{Contact}

Josie Esteban Rodriguez Condia {josie.rodriguez@polito.it}

Boyang Du {boyang.du@polito.it}

Gianluca Roascio {gianluca.roascio@polito.it}

Edouard Scie {edouard.scie@grenoble-inp.org}

Juan David Guerrero Balaguera {juandavid.guerrero@uptc.edu.co}