{"id":13642678,"url":"https://github.com/pulp-platform/mempool","last_synced_at":"2026-02-07T23:35:52.398Z","repository":{"id":37866700,"uuid":"223218149","full_name":"pulp-platform/mempool","owner":"pulp-platform","description":"A scalable 256/1024-RISC-V-core system with low-latency access into shared L1 memory.","archived":false,"fork":false,"pushed_at":"2026-02-01T15:06:53.000Z","size":78450,"stargazers_count":312,"open_issues_count":7,"forks_count":58,"subscribers_count":7,"default_branch":"main","last_synced_at":"2026-02-02T00:28:56.933Z","etag":null,"topics":["asic","manycore","risc-v"],"latest_commit_sha":null,"homepage":"","language":"C","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"apache-2.0","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/pulp-platform.png","metadata":{"files":{"readme":"README.md","changelog":"CHANGELOG.md","contributing":"CONTRIBUTING.md","funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":".github/CODEOWNERS","security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2019-11-21T16:34:37.000Z","updated_at":"2026-01-20T21:32:40.000Z","dependencies_parsed_at":"2023-10-11T12:43:28.702Z","dependency_job_id":"6c0d31f0-346c-441e-bed9-bb5417f23202","html_url":"https://github.com/pulp-platform/mempool","commit_stats":null,"previous_names":[],"tags_count":6,"template":false,"template_full_name":null,"purl":"pkg:github/pulp-platform/mempool","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/pulp-platform%2Fmempool","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/pulp-platform%2Fmempool/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/pulp-platform%2Fmempool/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/pulp-platform%2Fmempool/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/pulp-platform","download_url":"https://codeload.github.com/pulp-platform/mempool/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/pulp-platform%2Fmempool/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":29212753,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-02-07T23:14:30.912Z","status":"ssl_error","status_checked_at":"2026-02-07T23:14:17.253Z","response_time":63,"last_error":"SSL_connect returned=1 errno=0 peeraddr=140.82.121.5:443 state=error: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["asic","manycore","risc-v"],"created_at":"2024-08-02T01:01:34.824Z","updated_at":"2026-02-07T23:35:52.390Z","avatar_url":"https://github.com/pulp-platform.png","language":"C","funding_links":[],"categories":["C"],"sub_categories":[],"readme":"[![ci](https://github.com/pulp-platform/mempool/actions/workflows/ci.yml/badge.svg)](https://github.com/pulp-platform/mempool/actions/workflows/ci.yml)\n[![lint](https://github.com/pulp-platform/mempool/actions/workflows/lint.yml/badge.svg)](https://github.com/pulp-platform/mempool/actions/workflows/lint.yml)\n[![License](https://img.shields.io/badge/License-Apache%202.0-blue.svg)](https://opensource.org/licenses/Apache-2.0)\n\n# MemPool\n\nMemPool is a many-core system targeting image processing and wireless applications. It implements 256 RISC-V cores that can access a large, shared L1 memory in at most five cycles. TeraPool and MinPool, respectively a 1024 RISC-V cores scaled-up and a 16 RISC-V cores scaled-down parametrizations of MemPool are also supported.\n\nThis repository contains the software and hardware of MinPool, MemPool, and TeraPool, as well as infrastructure for compilation and simulation.\n\n## Structure\n\nThe repository is structured as follows:\n\n- `config` contains the global configurations that are used by software as well as hardware.\n- `hardware` is where the RTL code and simulation scripts are.\n- `scripts` contains useful scripts such as linting or formatting scripts.\n- `software` provides example applications and MemPool's runtime.\n- `toolchain` holds third-party packages\n    * `halide` is the compiler infrastructure for the _Halide_ language.\n    * `llvm-project` provides the LLVM compiler infrastructure.\n    * `riscv-gnu-toolchain` contains the RISC-V GCC compiler.\n    * `riscv-isa-sim` is an extended version of [Spike](https://github.com/riscv/riscv-isa-sim) and is used as the golden model and to parse simulation traces.\n    * `riscv-opcodes` is an extended version of [riscv-opcodes](https://github.com/riscv/riscv-opcodes) that contains our custom image processing extension.\n    * `verilator` provides the open-source RTL simulator Verilator.\n\n## Get Started\n\nMake sure you clone this repository recursively to get all the necessary submodules:\n\n```bash\ngit submodule update --init --recursive\n```\n\nIf the repository path of any submodule changes, run the following command to change your submodule's pointer to the remote repository:\n\n```bash\ngit submodule sync --recursive\n```\n\n## Build dependencies\n### Compiler\n\nMemPool requires at least the RISC-V GCC toolchain to compile applications. It also supports LLVM, which depends on GCC. To implement image processing kernels, MemPool also supports Halide, a domain-specific language built on top of C++. Its compilation process is based on LLVM.\n\nTo build these toolchains, run the following commands in the project's root directory.\n\n```bash\n# Build both compilers (GCC and LLVM)\nmake toolchain\n# Build only GCC\nmake tc-riscv-gcc\n# Build only LLVM\nmake tc-llvm\n# Build Halide\nmake halide\n```\n\n### RTL Simulation\n\nWe use [Bender](https://github.com/pulp-platform/bender) to checkout hardware dependencies and to generate our simulation scripts. Make sure you have Bender installed, or install it in the MemPool repository with:\n\n```bash\n# Install Bender\nmake bender\n```\n\nTo checkout the hardware dependencies using Bender run:\n\n```bash\n# Update hardware dependencies\nmake update-deps\n```\n\nThe RTL simulation, or more specifically, the tracing in the simulation, relies on the SPIKE simulator. To build it, run the following command in the project's directory:\n\n```bash\n# Build Spike\nmake riscv-isa-sim\n```\n\nMemPool supports ModelSim and the open-source Verilator for RTL simulation. Use the following command to build and install Verilator:\n```bash\n# Build Verilator\nmake verilator\n```\nYou will need an LLVM installation.\n\n## Software\n\n### Build Applications\n\nThe `software/apps` folder contains example applications that work on MemPool. MemPool also contains some Halide example applications in the `software/halide` folder and OpenMP applications in the `software/omp` folder. Run the following command to build an application. E.g., `hello_world`:\n\n```bash\n# Bare-metal applications\ncd software/apps/baremetal\nmake hello_world\n# Halide applications\ncd software/halide\nmake matmul\n# OpenMP applications\ncd software/omp\nmake omp_parallel\n```\n\nYou can also choose the compiler to build the application with using the `COMPILER` option. The possible options are `gcc` and `llvm`, the former being the default.\n\n```bash\n# Compile with LLVM instead of GCC\nmake COMPILER=llvm hello_world\n```\n\nTo run applications designed for the **Xpulpimg** extension, be sure to select the `gcc` compiler option.\nIf all the Xpulpimg instructions implemented in Snitch at compilation time are supported by the Xpulpimg subset of the GCC compiler, you can build your application with the option `XPULPIMG` set to `1`:\n\n```bash\n# Compile with GCC supporting Xpulpimg instruction set\nmake COMPILER=gcc XPULPIMG=1 hello_world\n```\n\nOtherwise, if new Xpulpimg instructions have been implemented in Snitch, but the Xpulpimg extension in the compiler does not support them yet, you must be sure to use Xpulpimg instructions only in an `asm volatile` construct within your C/C++ application, and set `XPULPIMG=0`. This will work as long as Xpulpimg is a subset of Xpulpv2.\n\nIf `XPULPIMG` is not forced while launching `make`, it will be defaulted to the `xpulpimg` value configured in `config/config.mk`. Note that such parameter in the configuration file also defines whether the Xpulpimg extension is enabled or not in the RTL of the Snitch core, and whether such Xpulpimg functionalities have to be tested or not by the `riscv-tests` unit tests.\n\n### Unit tests\n\nThe system is provided with an automatic unit tests suit for verification purposes; the tests are located in `riscv-tests/isa`, and can be launched from the top-level directory with:\n```bash\nmake riscv-tests\n```\nThe unit tests will be compiled, simulated in Spike, and run in RTL simulation of MemPool.\nThe compilation and simulation (for both Spike simulator and MemPool RTL) of the unit tests also depends on the `xpulpimg` parameter in `config/config.mk`: the test cases dedicated to the Xpulpimg instructions will be compiled and simulated only if `xpulpimg=1`.\nTo add more tests, you must add your own ones to the `riscv-isa` infrastructure; more information can be found in `software/riscv-tests/README.md`.\n\nThe unit tests are included in the software package of `software` and can be compiled for MemPool by launching in the `software` directory:\n```bash\nmake COMPILER=gcc riscv-tests\n```\nNote that the unit tests need to be compiled with `gcc`. The same logic of normal applications concerning the `XPULPIMG` parameter applies for tests.\n\n### Writing Applications\n\nMemPool follows [LLVM's coding style guidelines](https://llvm.org/docs/CodingStandards.html) when it comes to C and C++ code. We use `clang-format` to format all C code. Use `make format` in the project's root directory before committing software changes to make them conform with our style guide through *clang-format*.\n\n## RTL Simulation\n\nTo simulate the MemPool system with ModelSim, go to the `hardware` folder, which contains all the SystemVerilog files. Use the following command to run your simulation:\n\n```bash\n# Go to the hardware folder\ncd hardware\n# Only compile the hardware without running the simulation.\nmake compile\n# Run the simulation with the *hello_world* binary loaded\napp=baremetal/hello_world make sim\n# For Halide applications use the `halide-` prefix: E.g., to run `matmul`:\napp=halide-matmul make sim\n# Run the simulation with the *some_binary* binary. This allows specifying the full path to the binary\npreload=/some_path/some_binary make sim\n# Run the simulation without starting the gui\napp=hello_world make simc\n# Generate the human-readable traces after simulation is completed\nmake trace\n# Generate a visualization of the traces\napp=hello_world make tracevis\n# Automatically run the benchmark (headless), extract the traces, and log the results\napp=hello_world make benchmark\n```\n\nYou can set up the configuration of the system in the file `config/config.mk`, controlling the total number of cores, the number of cores per tile and whether the Xpulpimg extension is enabled or not in the Snitch core; the `xpulpimg` parameter also control the default core architecture considered when compiling applications for MemPool.\n\nTo simulate the MemPool system with Verilator use the same format, but with the target\n```bash\nmake verilate\n```\nIf, during the Verilator model compilation, you run out of space on your disk, use\n```bash\nexport OBJCACHE=''\n```\nto disable the use of `ccache`. Keep in mind that this will make the following compilations slower since compiled object files will no longer be cached.\n\nIf the tracer is enabled, its output traces are found under `hardware/build`, for both ModelSim and Verilator simulations.\n\nTracing can be controlled per core with a custom `trace` CSR register. The CSR is of type WARL and can only be set to zero or one. For debugging, tracing can be enabled persistently with the `snitch_trace` environment variable.\n\nTo get a visualization of the traces, check out the `scripts/tracevis.py` script. It creates a JSON file that can be viewed with [Trace-Viewer](https://github.com/catapult-project/catapult/tree/master/tracing) or in Google Chrome by navigating to `about:tracing`.\n\nWe also provide Synopsys Spyglass linting scripts in the `hardware/spyglass`. Run `make lint` in the `hardware` folder, with a specific MemPool configuration, to run the tests associated with the `lint_rtl` target.\n\n## Hardware Configurations\n\nMemPool's core-count is parametrizable. In `./config`, the `config.mk` file includes other configuration files, which represent specific parametrizations of MemPool. We currently support three parametrizations:\n- `terapool`: 1024 cores, organized into 128 tiles with eight cores each\n- `mempool`: 256 cores, organized into 64 tiles with four cores each (default)\n- `minpool`: 16 cores, organized into 4 tiles with four cores each\n\nThe software and the hardware for different MemPool configurations can be compiled specifying the desired configuration on the command line. For example, to compile and simulate the `hello_world` app for TeraPool:\n```bash\ncd hardware\n# Compile the hello_world for TeraPool configuration\nconfig=terapool make COMPILER=gcc hello_world\n# Run the simulation with the *hello_world* binary loaded\nconfig=terapool app=baremetal/hello_world make sim\n```\n\n## DRAMsys Co-Simulation\n\nThe MemPool system supports both on-chip SRAM or off-chip DRAM co-simulation for higher hierarchy memory transfering. For off-chip DRAM co-simulation, it incorporates the `dram_rtl_sim` tool as a submodule, build at `hardware/deps/dram_rtl_sim`. Leveraging DRAMSys5.0, it facilitates an effective co-simulation environment between RTL models and DRAMSys5.0 for the simulation of DRAM + CTRL models, with contemporary off-chip DRAM technologies (e.g., LPDDR, DDR, HBM).\n\nThe DRAMsys tool aids are open-sourced and can be found here:\n[https://github.com/pulp-platform/dram_rtl_sim](https://github.com/pulp-platform/dram_rtl_sim)\n\n### Building DRAMsys Co-Simulation\n\nTo prepare for DRAMsys co-simulation, adjust the system configuration by setting `l2_sim_type` to `dram` in `config/config.mk`. Then, execute the following command in the project's root directory to establish the DRAMsys tool aids environment:\n\n```bash\nmake setup-dram\n```\n\nThis makefile target automates several tasks:\n1. Cleans up the existing DRAMSys5.0 repository, if previously built.\n2. Rebuilds the DRAMSys5.0 repository and applies necessary patches within `hardware/deps/dramsys_rtl_sim/dramsys_lib/`.\n3. Applies HBM2 DRAM configuration patches tailored for the MemPool system simulation.\n4. Compiles the DRAMSys dynamic linkable library located at `hardware/deps/dramsys_rtl_sim/dramsys_lib/DRAMSys`.\n\n**Important:** This environment requires `cmake` version 3.28.1 or higher and GCC version 11.2.0 or above.\n\n### DRAM Chip Configuration\n\nDRAMsys supports a range of contemporary off-chip DRAM technologies, including LPDDR, DDR, and HBM. Configuration files, formatted as `.json`, are accessible in the following directory: `hardware/deps/dramsys_rtl_sim/dramsys_lib/DRAMSys/configs`. Additionally, we provide a recommended HBM2 configuration for the MemPool system located within `hardware/deps/dramsys_rtl_sim/dramsys_lib/DRAMSys`. This configuration is automatically applied as the default setting when establishing the DRAMsys tool aids environment. You are encouraged to review and modify these configurations as necessary to meet your specific simulation requirements.\n\n### Testing MemPool-DRAMSys Co-Simulation\n\nFor data transfer testing between the MemPool system and higher hierarchy memory through DMA transfer, use the prepared example kernel located in `software/tests/baremetal/memcpy`. For more detailed methods on building applications and setting up RTL simulation, please refer to the sections aboves.\n\n**Note:** Currently, the simulation crafting tool for off-chip DRAM co-simulation is not open-sourced. We utilize the `Questasim` simulator exclusively.\n\n## Publications\nIf you use MemPool or TeraPool in your work or research, you can cite us:\n\n**MemPool: A Scalable Manycore Architecture with a Low-Latency Shared L1 Memory**\n\n```\n@article{Riedel2023MemPool,\n  title = {{MemPool}: A Scalable Manycore Architecture with a Low-Latency Shared {L1} Memory},\n  author = {Riedel, Samuel and Cavalcante, Matheus and Andri, Renzo and Benini, Luca},\n  journal = {IEEE Transactions on Computers},\n  year = {2023},\n  volume = {72},\n  number = {12},\n  pages = {3561--3575},\n  publisher = {IEEE Computer Society},\n  doi = {10.1109/TC.2023.3307796}\n}\n```\nThis paper was published on [IEEE Xplore](https://ieeexplore.ieee.org/document/10227739) and is also available on [arXiv:2303.17742 [cs.AR]](https://arxiv.org/abs/2303.17742) and the [ETH Research Collection](https://doi.org/10.3929/ethz-b-000643341).\n\n**TeraPool: A Physical Design Aware, 1024 RISC-V Cores Shared-L1-Memory Scaled-Up Cluster Design With High Bandwidth Main Memory Link**\n\n```\n@article{Zhang2025TeraPool,\n  title={TeraPool: A Physical Design Aware, 1024 RISC-V Cores Shared-L1-Memory Scaled-Up Cluster Design With High Bandwidth Main Memory Link},\n  author={Zhang, Yichao and Bertuletti, Marco and Zhang, Chi and Riedel, Samuel and Shen, Diyou and Wang, Bowen and Vanelli-Coralli, Alessandro and Benini, Luca},\n  journal={IEEE Transactions on Computers},\n  year={2025},\n  volume={74},\n  number={11},\n  pages={3667-3681},\n  doi={10.1109/TC.2025.3603692}\n}\n```\nThis paper was published on [IEEE Xplore](https://ieeexplore.ieee.org/document/11145329).\n\nThe following publications give more details about MemPool, TeraPool, its extensions, and use cases:\n\n### 2021\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003ci\u003eMemPool: A Shared-L1 Memory Many-Core Cluster with a Low-Latency Interconnect\u003c/i\u003e\u003c/summary\u003e\n\u003cp\u003e\n\n```\n@inproceedings{Cavalcante2021MemPool,\n  title = {{MemPool}: A Shared-{L1} Memory Many-Core Cluster with a Low-Latency Interconnect},\n  author = {Cavalcante, Matheus and Riedel, Samuel and Pullini, Antonio and Benini, Luca},\n  booktitle = {2021 Design, Automation, and Test in Europe Conference and Exhibition},\n  address = {Grenoble, France},\n  year = {2021},\n  month = mar,\n  pages = {701--706},\n  publisher = {IEEE},\n  doi = {10.23919/DATE51398.2021.9474087}\n}\n```\nThis paper was published on [IEEE Xplore](https://ieeexplore.ieee.org/document/9474087) and is also available on [arXiv:2012.02973 [cs.AR]](https://arxiv.org/abs/2012.02973).\n\n\u003c/p\u003e\n\u003c/details\u003e\n\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003ci\u003e3D SoC integration, beyond 2.5D chiplets\u003c/i\u003e\u003c/summary\u003e\n\u003cp\u003e\n\n```\n@inproceedings{Beyne2021,\n  title = {{3D} {SoC} integration, beyond {2.5D} chiplets},\n  author = {Beyne, Eric and Milojevic, Dragomir and {Van Der Plas}, Geert and Beyer, Gerald},\n  booktitle = {Technical Digest - International Electron Devices Meeting, IEDM},\n  year = {2021},\n  pages = {79--82},\n  publisher = {IEEE},\n  doi = {10.1109/IEDM19574.2021.9720614}\n}\n```\nThis paper was published on [IEEE Xplore](https://ieeexplore.ieee.org/document/9720614).\n\n\u003c/p\u003e\n\u003c/details\u003e\n\n\n### 2022\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003ci\u003eMemPool-3D: Boosting Performance and Efficiency of Shared-L1 Memory Many-Core Clusters with 3D Integration\u003c/i\u003e\u003c/summary\u003e\n\u003cp\u003e\n\n```\n@inproceedings{Cavalcante2022MemPool3D,\n  title = {{MemPool-3D}: Boosting Performance and Efficiency of Shared-{L1} Memory Many-Core Clusters with {3D} Integration},\n  author = {Cavalcante, Matheus and Agnesina, Anthony and Riedel, Samuel and Brunion, Moritz and Garcia-Ortiz, Alberto and Milojevic, Dragomir and Catthoor, Francky and Lim, Sung Kyu and Benini, Luca},\n  booktitle = {2022 Design, Automation, and Test in Europe Conference and Exhibition},\n  address = {Online},\n  year = {2022},\n  month = mar,\n  pages = {394--399},\n  publisher = {IEEE},\n  doi = {10.23919/DATE54114.2022.9774726}\n}\n```\nThis paper was published on [IEEE Xplore](https://ieeexplore.ieee.org/document/9774726) and is also available on [arXiv:2112.01168 [cs.AR]](https://arxiv.org/abs/2112.01168).\n\n\u003c/p\u003e\n\u003c/details\u003e\n\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003ci\u003eHier-3D: A Hierarchical Physical Design Methodology for Face-to-Face-Bonded 3D ICs\u003c/i\u003e\u003c/summary\u003e\n\u003cp\u003e\n\n```\n@inproceedings{Agnesina2022,\n  title = {{Hier-3D}: A Hierarchical Physical Design Methodology for Face-to-Face-Bonded {3D} ICs},\n  author = {Agnesina, Anthony and Brunion, Moritz and Garcia-Ortiz, Alberto and Catthoor, Francky and Milojevic, Dragomir and Komalan, Manu and Cavalcante, Matheus and Riedel, Samuel and Benini, Luca and Lim, Sung Kyu},\n  booktitle = {Proceedings of the ACM/IEEE International Symposium on Low Power Electronics and Design},\n  address = {New York, NY, USA},\n  year = {2022},\n  month = aug,\n  publisher = {Association for Computing Machinery},\n  doi = {10.1145/3531437.3539702}\n}\n```\nThis paper was published on [ACM DL](https://dl.acm.org/doi/10.1145/3531437.3539702).\n\n\n\u003c/p\u003e\n\u003c/details\u003e\n\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003ci\u003eSpatz: A Compact Vector Processing Unit for High-Performance and Energy-Efficient Shared-L1 Clusters\u003c/i\u003e\u003c/summary\u003e\n\u003cp\u003e\n\n```\n@inproceedings{Cavalcante2022Spatz,\n  title = {Spatz: A Compact Vector Processing Unit for High-Performance and Energy-Efficient Shared-{L1} Clusters},\n  author = {Cavalcante, Matheus and W{\\\"{u}}thrich, Domenic and Perotti, Matteo and Riedel, Samuel and Benini, Luca},\n  booktitle = {2022 IEEE/ACM International Conference On Computer Aided Design (ICCAD)},\n  address = {San Diego, California, USA},\n  year = {2022},\n  month = oct,\n  pages = {159--167},\n  publisher = {Association for Computing Machinery},\n  doi = {10.1145/3508352.3549367}\n}\n```\nThis paper was published on [IEEE Xplore](https://ieeexplore.ieee.org/document/10069431) and is also available on [arXiv:2207.07970 [cs.AR]](https://arxiv.org/abs/2207.07970).\n\n\u003c/p\u003e\n\u003c/details\u003e\n\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003ci\u003eThermal Performance Analysis of Mempool RISC-V Multicore SoC\u003c/i\u003e\u003c/summary\u003e\n\u003cp\u003e\n\n```\n@article{Venkateswarlu2022,\n  title = {Thermal Performance Analysis of Mempool RISC-V Multicore {SoC}},\n  author = {Venkateswarlu, Sankatali and Mishra, Subrat and Oprins, Herman and Vermeersch, Bjorn and Brunion, Moritz and Han, Jun Han and Stan, Mircea R. and Weckx, Pieter and Catthoor, Francky},\n  journal = {IEEE Transactions on Very Large Scale Integration (VLSI) Systems},\n  year = {2022},\n  volume = {30},\n  number = {11},\n  pages = {1668--1676},\n  publisher = {IEEE},\n  doi = {10.1109/TVLSI.2022.3207553}\n}\n```\nThis paper was published on [IEEE Xplore](https://ieeexplore.ieee.org/document/9905665).\n\n\u003c/p\u003e\n\u003c/details\u003e\n\n\n### 2023\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003ci\u003eTowards Chip-Package-System Co-optimization of Thermally-limited System-On-Chips (SOCs)\u003c/i\u003e\u003c/summary\u003e\n\u003cp\u003e\n\n```\n@inproceedings{Mishra2023,\n  title = {Towards Chip-Package-System Co-optimization of Thermally-limited System-On-Chips (SOCs)},\n  author = {Mishra, S. and Sankatali, V. and Vermeersch, B. and Brunion, M. and Lofrano, M. and Abdi, D. and Oprins, H. and Biswas, D. and Zografos, O. and Hiblot, G. and {Van Der Plas}, G. and Weckx, P. and Hellings, G. and Myers, J. and Catthoor, F. and Ryckaert, J.},\n  booktitle = {IEEE International Reliability Physics Symposium Proceedings},\n  address = {Monterey, CA, USA},\n  year = {2023},\n  month = mar,\n  publisher = {IEEE},\n  doi = {10.1109/IRPS48203.2023.10117979}\n}\n```\nThis paper was published on [IEEE Xplore](https://ieeexplore.ieee.org/document/10117979).\n\n\u003c/p\u003e\n\u003c/details\u003e\n\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003ci\u003eEfficient Parallelization of 5G-PUSCH on a Scalable RISC-V Many-Core Processor\u003c/i\u003e\u003c/summary\u003e\n\u003cp\u003e\n\n```\n@inproceedings{Bertuletti2023PUSCH,\n  title = {Efficient Parallelization of {5G-PUSCH} on a Scalable {RISC-V} Many-Core Processor},\n  author = {Bertuletti, Marco and Zhang, Yichao and Vanelli-Coralli, Alessandro and Benini, Luca},\n  booktitle = {2023 Design, Automation, and Test in Europe Conference and Exhibition},\n  address = {Antwerp, Belgium},\n  year = {2023},\n  month = apr,\n  pages = {396--401},\n  publisher = {IEEE},\n  doi = {10.23919/DATE56975.2023.10137247}\n}\n```\nThis paper was published on [IEEE Xplore](https://ieeexplore.ieee.org/document/10137247) and is also available on [arXiv:2210.09196 [cs.DC]](https://arxiv.org/abs/2210.09196).\n\n\u003c/p\u003e\n\u003c/details\u003e\n\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003ci\u003eMemPool Meets Systolic: Flexible Systolic Computation in a Large Shared-Memory Processor Cluster\u003c/i\u003e\u003c/summary\u003e\n\u003cp\u003e\n\n```\n@inproceedings{Riedel2023MmS,\n  title = {{MemPool} Meets Systolic: Flexible Systolic Computation in a Large Shared-Memory Processor Cluster},\n  author = {Riedel, Samuel and Khov, Gua Hao and Mazzola, Sergio and Cavalcante, Matheus and Andri, Renzo and Benini, Luca},\n  booktitle = {2023 Design, Automation, and Test in Europe Conference and Exhibition},\n  address = {Antwerp, Belgium},\n  year = {2023},\n  month = apr,\n  pages = {503--504},\n  publisher = {IEEE},\n  doi = {10.23919/DATE56975.2023.10136909}\n}\n```\nThis paper was published on [IEEE Xplore](https://ieeexplore.ieee.org/document/10136909).\n\n\u003c/p\u003e\n\u003c/details\u003e\n\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003ci\u003eFast Shared-Memory Barrier Synchronization for a 1024-Cores RISC-V Many-Core Cluster\u003c/i\u003e\u003c/summary\u003e\n\u003cp\u003e\n\n```\n@inproceedings{Bertuletti2023Barrier,\n  title = {Fast Shared-Memory Barrier Synchronization for a 1024-Cores {RISC-V} Many-Core Cluster},\n  author = {Bertuletti, Marco and Riedel, Samuel and Zhang, Yichao and Vanelli-Coralli, Alessandro and Benini, Luca},\n  booktitle = {Embedded Computer Systems: Architectures, Modeling, and Simulation},\n  editor = {Silvano, Cristina and Pilato, Christian and Reichenbach, Marc},\n  address = {Samos},\n  year = {2023},\n  month = jul,\n  pages = {241--254},\n  publisher = {Springer Nature Switzerland},\n  doi = {10.1007/978-3-031-46077-7_16}\n}\n```\nThis paper was published on [Springer Link](https://link.springer.com/chapter/10.1007/978-3-031-46077-7_16) and is also available on [arXiv:2307.10248 [cs.DC]](https://arxiv.org/abs/2307.10248) and the [ETH Research Collection](https://doi.org/10.3929/ethz-b-000648454).\n\n\u003c/p\u003e\n\u003c/details\u003e\n\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003ci\u003eMemPool: A Scalable Manycore Architecture with a Low-Latency Shared L1 Memory\u003c/i\u003e\u003c/summary\u003e\n\u003cp\u003e\n\n```\n@article{Riedel2023MemPool,\n  title = {{MemPool}: A Scalable Manycore Architecture with a Low-Latency Shared {L1} Memory},\n  author = {Riedel, Samuel and Cavalcante, Matheus and Andri, Renzo and Benini, Luca},\n  journal = {IEEE Transactions on Computers},\n  year = {2023},\n  volume = {72},\n  number = {12},\n  pages = {3561--3575},\n  publisher = {IEEE Computer Society},\n  doi = {10.1109/TC.2023.3307796}\n}\n```\nThis paper was published on [IEEE Xplore](https://ieeexplore.ieee.org/document/10227739) and is also available on [arXiv:2303.17742 [cs.AR]](https://arxiv.org/abs/2303.17742) and the [ETH Research Collection](https://doi.org/10.3929/ethz-b-000643341).\n\n\u003c/p\u003e\n\u003c/details\u003e\n\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003ci\u003eImpact of 3-D Integration on Thermal Performance of RISC-V MemPool Multicore SOC\u003c/i\u003e\u003c/summary\u003e\n\u003cp\u003e\n\n```\n@article{Venkateswarlu2023,\n  title = {Impact of 3-D Integration on Thermal Performance of {RISC-V} {MemPool} Multicore {SOC}},\n  author = {Venkateswarlu, Sankatali and Mishra, Subrat and Oprins, Herman and Vermeersch, Bjorn and Brunion, Moritz and Han, Jun Han and Stan, Mircea R. and Biswas, Dwaipayan and Weckx, Pieter and Catthoor, Francky},\n  journal = {IEEE Transactions on Very Large Scale Integration (VLSI) Systems},\n  year = {2023},\n  volume = {31},\n  number = {12},\n  pages = {1896-1904},\n  publisher = {IEEE},\n  doi = {10.1109/TVLSI.2023.3314135}\n}\n```\nThis paper was published on [IEEE Xplore](https://ieeexplore.ieee.org/document/10261872).\n\n\u003c/p\u003e\n\u003c/details\u003e\n\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003ci\u003eMinPool: A 16-core NUMA-L1 Memory RISC-V Processor Cluster for Always-on Image Processing in 65nm CMOS\u003c/i\u003e\u003c/summary\u003e\n\u003cp\u003e\n\n```\n@inproceedings{Riedel2023MinPool,\n  author={Riedel, Samuel and Cavalcante, Matheus and Frouzakis, Manos and Wüthrich, Domenic and Mustafa, Enis and Billa, Arlind and Benini, Luca},\n  title={{MinPool}: A 16-core {NUMA-L1} Memory {RISC-V} Processor Cluster for Always-on Image Processing in 65nm {CMOS}},\n  booktitle={2023 30th IEEE International Conference on Electronics, Circuits and Systems (ICECS)},\n  address = {Istanbul, Turkiye},\n  year={2023},\n  month=dec,\n  pages={1--4},\n  publisher={IEEE},\n  doi={10.1109/ICECS58634.2023.10382925}\n}\n```\nThis paper was published on [IEEE Xplore](https://ieeexplore.ieee.org/document/10382925) and is also available on the [ETH Research Collection](https://doi.org/10.3929/ethz-b-000653598).\n\n\u003c/p\u003e\n\u003c/details\u003e\n\n### 2024\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003ci\u003eLRSCwait: Enabling Scalable and Efficient Synchronization in Manycore Systems through Polling-Free and Retry-Free Operation\u003c/i\u003e\u003c/summary\u003e\n\u003cp\u003e\n\n```\n@inproceedings{Riedel2024LRSCwait,\n  author={Riedel, Samuel and Gantenbein, Marc and Ottaviano, Alessandro and Hoefler, Torsten and Benini, Luca},\n  title={{LRSCwait}: Enabling Scalable and Efficient Synchronization in Manycore Systems Through Polling-Free and Retry-Free Operation},\n  booktitle={2024 Design, Automation \\\u0026 Test in Europe Conference \\\u0026 Exhibition (DATE)},\n  year={2024},\n  month=mar,\n  pages={1-6}\n}\n```\nThis paper was published on [IEEE Xplore](https://ieeexplore.ieee.org/document/10546857) and is also available on [arXiv:2401.09359 [cs.AR]](https://arxiv.org/abs/2401.09359).\n\n\u003c/p\u003e\n\u003c/details\u003e\n\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003ci\u003eEnabling Efficient Hybrid Systolic Computation in Shared L1-Memory Manycore Clusters\u003c/i\u003e\u003c/summary\u003e\n\u003cp\u003e\n\n```\n@article{Mazzola2024Systolic,\n      title={Enabling Efficient Hybrid Systolic Computation in Shared {L1}-Memory Manycore Clusters},\n      author={Sergio Mazzola and Samuel Riedel and Luca Benini},\n      journal={arXiv:2402.12986 [cs.AR]},\n      year={2024},\n      month=feb\n}\n```\nThis paper is available on [arXiv:2402.12986 [cs.AR]](https://arxiv.org/abs/2402.12986).\n\n\u003c/p\u003e\n\u003c/details\u003e\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003ci\u003eMX: Enhancing RISC-V's Vector ISA for Ultra-Low Overhead, Energy-Efficient Matrix Multiplication\u003c/i\u003e\u003c/summary\u003e\n\u003cp\u003e\n\n```\n@inproceedings{perotti2024mx,\n      author={Perotti, Matteo and Zhang, Yichao and Cavalcante, Matheus and Mustafa, Enis and Benini, Luca},\n      booktitle={2024 Design, Automation \u0026 Test in Europe Conference \u0026 Exhibition (DATE)},\n      title={MX: Enhancing RISC-V's Vector ISA for Ultra-Low Overhead, Energy-Efficient Matrix Multiplication},\n      year={2024},\n      pages={1-6},\n      isbn = {979-8-3503-4860-6}\n}\n\n```\nThis paper is available on [IEEE Xplore](https://ieeexplore.ieee.org/document/10546720).\n\n\u003c/p\u003e\n\u003c/details\u003e\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003ci\u003eTeraPool-SDR: An 1.89TOPS 1024 RV-Cores 4MiB Shared-L1 Cluster for Next-Generation Open-Source Software-Defined Radios\u003c/i\u003e\u003c/summary\u003e\n\u003cp\u003e\n\n```\n@inproceedings{Yichao2024Terapool,\n      author={Zhang, Yichao and Bertuletti, Marco and Riedel, Samuel and Cavalcante, Matheus and Vanelli-Coralli, Alessandro and Benini, Luca},\n      title={TeraPool-SDR: An 1.89TOPS 1024 RV-Cores 4MiB Shared-L1 Cluster for Next-Generation Open-Source Software-Defined Radios},\n      year={2024},\n      isbn={9798400706059},\n      publisher={Association for Computing Machinery},\n      address={New York, NY, USA},\n      url={https://doi.org/10.1145/3649476.3658735},\n      doi={10.1145/3649476.3658735},\n      booktitle={Proceedings of the Great Lakes Symposium on VLSI 2024},\n      pages={86–91},\n      numpages={6},\n      series={GLSVLSI '24}\n}\n```\nThis paper is available on [ACM DIGITAL LIBRARY](https://dl.acm.org/doi/10.1145/3649476.3658735).\n\n### 2025\n\n\u003c/p\u003e\n\u003c/details\u003e\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003ci\u003eA 66-Gb/s/5.5-W RISC-V Many-Core Cluster for 5G+ Software-Defined Radio Uplinks\u003c/i\u003e\u003c/summary\u003e\n\u003cp\u003e\n\n```\n@article{Bertuletti2025TeraPool,\n  title={A 66-Gb/s/5.5-W RISC-V Many-Core Cluster for 5G+ Software-Defined Radio Uplinks},\n  author={Bertuletti, Marco and Zhang, Yichao and Vanelli-Coralli, Alessandro and Benini, Luca},\n  journal={IEEE Transactions on Very Large Scale Integration (VLSI) Systems},\n  year={2025},\n  volume={33},\n  number={8},\n  pages={2225-2238},\n  doi={10.1109/TVLSI.2025.3576855}\n}\n```\nThis paper is available on [IEEE Xplore](https://ieeexplore.ieee.org/abstract/document/11038837).\n\n\u003c/p\u003e\n\u003c/details\u003e\n\n## Chips\n\nThe MemPool architecture has been taped out in the following chips:\n\n- 2021 [**MinPool**](http://asic.ethz.ch/2021/Minpool.html): A 16-core prototype of MemPool.\n- 2024 [**Heartstream**](http://asic.ethz.ch/2024/Heartstream.html): A 64-core version of MemPool with systolic, FPU and S-DIVSQRT(Shared Division Square-Root) unit support.\n\n## License\nMemPool is released under permissive open source licenses. Most of MemPool's source code is released under the Apache License 2.0 (`Apache-2.0`) see [`LICENSE`](LICENSE). The code in `hardware` is released under Solderpad v0.51 (`SHL-0.51`) see [`hardware/LICENSE`](hardware/LICENSE).\n\nNote, MemPool includes several third-party packages with their own licenses:\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003ci\u003eNote, MemPool includes several third-party packages with their own licenses:\u003c/i\u003e\u003c/summary\u003e\n\u003cp\u003e\n\n### Software\n\n- `software/runtime/printf.{c,h}` is licensed under the MIT license.\n- `software/runtime/omp/libgomp.h` is licensed under the GPL license.\n- `software/riscv-tests` is an extended version of RISC-V's [riscv-tests](https://github.com/riscv/riscv-tests/) repository licensed under a BSD license. See [`software/riscv-tests/LICENSE`](software/riscv-tests/LICENSE) for details.\n\n### Hardware\n\nThe `hardware` folder is licensed under Solderpad v0.51 see [`hardware/LICENSE`](hardware/LICENSE). We use the following exceptions:\n\n- `hardware/tb/dpi/elfloader.cpp` is licensed under a BSD license.\n- `hardware/tb/verilator/*` is licensed under Apache License 2.0 see [`LICENSE`](LICENSE)\n- `hardware/tb/verilator/lowrisc_*` contain modified versions of lowRISC's helper libraries. They are licensed under Apache License 2.0.\n\n### Scripts\n\n- `scripts/run_clang_format.py` is licensed under the MIT license.\n\n### Toolchains\n\nThe following compilers can be used to build applications for MemPool:\n\n- `toolchain/halide` is licensed under the MIT license. See [Halide's license](https://github.com/halide/Halide/blob/master/LICENSE.txt) for details.\n- `toolchain/llvm-project`is licensed under the Apache License v2.0 with LLVM Exceptions. See [LLVM's DeveloperPolicy](https://llvm.org/docs/DeveloperPolicy.html#new-llvm-project-license-framework) for more details.\n- `toolchain/riscv-gnu-toolchain`'s licensing information is available [here](https://github.com/pulp-platform/pulp-riscv-gnu-toolchain/blob/master/LICENSE)\n\nWe use the following RISC-V tools to parse simulation traces and keep opcodes consistent throughout the project.\n\n- `toolchain/riscv-isa-sim` is licensed under a BSD license. See [riscv-isa-sim's license](https://github.com/riscv/riscv-isa-sim/blob/master/LICENSE) for details.\n- `toolchain/riscv-opcodes` contains an extended version of [riscv-opcodes](https://github.com/riscv/riscv-opcodes) licensed under the BSD license. See [`toolchain/riscv-opcodes/LICENSE`](toolchain/riscv-opcodes/LICENSE) for details.\n\nThe open-source simulator [Verilator](https://www.veripool.org/verilator) can be used for RTL simulation.\n\n- `toolchain/verilator` is licensed under GPL. See [Verilator's license](https://github.com/verilator/verilator/blob/master/LICENSE) for more details.\n\n### DRAMsys5.0\n\n- The `dram_rtl_sim` submodule, located at `hardware/deps/dram_rtl_sim`, is licensed under the Solderpad Hardware License 0.51. You can review the license [here](https://github.com/pulp-platform/dram_rtl_sim/blob/main/LICENSE).\n- [DRAMSys5.0](https://github.com/tukl-msd/DRAMSys) is utilized for DRAM simulations. For details on its usage and licensing, please refer to the DRAMSys5.0 [license information](https://github.com/tukl-msd/DRAMSys).\n\n\u003c/p\u003e\n\u003c/details\u003e\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fpulp-platform%2Fmempool","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fpulp-platform%2Fmempool","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fpulp-platform%2Fmempool/lists"}