Bug#894628: ITP: cupy/2.5.0 -- NumPy-like API accelerated with CUDA
Hello, As per the discussion on the Debian Science mailing list, CuPy also has an AMD ROCm backend [1]. I believe all required components for building CuPy for AMD hardware are currently available on Debian Sid. The ROCm version of CuPy could therefore be available in main rather than contrib (as would be required for the NVIDIA backend). Sincerely, Cory Bloor [1]: https://lists.debian.org/debian-science/2024/06/msg00033.html
Bug#1079952: RFP: vitis-ai -- AMD Xilinx tools and libraries for AI inference
Package: wnpp Severity: wishlist X-Debbugs-Cc: debian...@lists.debian.org * Package name: vitis-ai Version : 3.5 * URL : https://github.com/Xilinx/Vitis-AI/ * License : Apache-2.0 Programming Lang: C++, Python Description : AMD Xilinx tools and libraries for AI inference Vitis AI is a collection of tools and libraries for accelerating AI inference on AMD Xilinx hardware. It provides an optimizer to prune over-parameterized models, quantizers to reduce model precision, a runtime to submit jobs to accelerator hardware, and a "model zoo" that provides metadata describing compatible AI models along with recommended optimization and quantization parameters for acceleration. This package provides tooling used to run AI models on AMD Xilinx hardware, including Versal and Zynq adaptive systems-on-a-chip, Alveo adaptive accelerators for data centres, and Ryzen AI accelerators for personal computers. This package would be maintained by the Debian AI Team.
Bug#1064257: ITP: rocm-tensile -- ROCm tool for generating and benchmarking assembly kernels
Package: wnpp Severity: wishlist Owner: Cordell Bloor X-Debbugs-Cc: debian-de...@lists.debian.org, c...@slerp.xyz, debian...@lists.debian.org * Package name: rocm-tensile Version : 6.0.2 * URL : https://github.com/ROCm/Tensile * License : Expat Programming Lang: Python, HIP Description : ROCm tool for generating and benchmarking assembly kernels Tensile is a set of tools and libraries primarily for selecting parameters of GPU kernels implementing the general matrix multiply (GEMM) operation. There are three components that comprise Tensile: . 1. A command-line tool for generating kernels, benchmarking them, and saving the parameters used for generating the best kernels (a.k.a. "solutions") in YAML files. 2. A build system component that reads YAML solution files, generates kernel source files, and invokes the compiler to turn them into code object files. The kernels are indexed by their parameters in either YAML or MessagePack format within a TensileLibrary file. 3. A runtime library for loading and executing the best available solution for a given set of GEMM input parameters (a.k.a. "a problem"). The rocm-tensile library sources are currently packaged as part of rocblas in a multi-upstream tarball package, but they should be split out so that the command-line tool can be packaged. Tensile kernels are a vital part of the performance of the rocblas library. It is often necessary to add tuned kernels for particular problem sizes to achieve optimal performance in a new application or on a new hardware architecture. This is therefore an important development tool for BLAS performance on AMD GPUs. A fork of the Tensile library is also used by hipblaslt. Splitting Tensile out from the rocblas package may be helpful in preventing the duplication of embedded copies. The Tensile library can also be used by MIOpen. This package is part of AMD's ROCm stack and will be maintained under the Debian AI team umbrella.
Bug#1064071: RFP: hipblaslt -- portable interface for extended general matrix-matrix operations
Package: wnpp Severity: wishlist X-Debbugs-Cc: c...@slerp.xyz, debian...@lists.debian.org * Package name: hipblaslt Version : 6.0.2 * URL : https://github.com/ROCm/hipBLASLt * License : Expat Programming Lang: C, C++, HIP Description : portable interface for extended general matrix-matrix operations hipBLASLt is a library for general matrix-matrix operations that extends the traditional BLAS interface to support more options controlling data layout, data types, and algorithm selection. The hipBLASLt library is implemented in the HIP programming language and is supplemented by tuned assembly kernels. The hipBLASLt library helps users port code using cuBLASLt to the AMD ROCm platform. It is used by PyTorch for accelerating operations on MI200 and MI300 GPU hardware. The package is part of AMD's ROCm stack and would be maintained under the Debian AI team umbrella.
Bug#1064070: RFP: cxxheaderparser -- python library for parsing C++ headers
Package: wnpp Severity: wishlist X-Debbugs-Cc: c...@slerp.xyz, debian...@lists.debian.org * Package name: cxxheaderparser Version : 1.3.1 Upstream Contact: RobotPy Development Team * URL : https://github.com/robotpy/cxxheaderparser * License : BSD-3-Clause Programming Lang: Python Description : Python library for parsing C++ headers The cxxheaderparser library is used to parse syntactically valid C++ code and operate on the results. It provides both a visitor-style interface to process the results as they are being parsed or the option of a single data structure containing all parsed information. This library is a successor to CppHeaderParser, which is a build dependency of the AMD ROCm GPU profiling libraries used by PyTorch. Specifically, CppHeaderParser is required for roctracer and parts of rocm-hipamd. As CppHeaderParser has been deprecated by its authors, those libraries will need to be migrated to cxxheaderparser.
Bug#1063741: ITP: hipify -- CUDA to HIP source-to-source translation tools
Package: wnpp Severity: wishlist Owner: Cordell Bloor X-Debbugs-Cc: debian-de...@lists.debian.org, debian...@lists.debian.org, c...@slerp.xyz * Package name: hipify Version : 6.0.2 * URL : https://github.com/ROCm/HIPIFY * License : Expat Programming Lang: C++, Perl Description : CUDA to HIP source-to-source translation tools hipify is a set of tools to convert CUDA sources into HIP sources. It provides hipify-clang, which uses a clang-based parser and can therefore translate complex C++ constructs, but requires complete input sources, including access to any CUDA headers used. For cases where this is not possible, it provides hipify-perl, which uses a simple perl-based parser that can translate syntactically invalid inputs and source file fragments, but may not recognize complex C++ constructs. Hipify is often used by developers porting existing CUDA-based libraries to the ROCm platform. Additionally, the hipify-perl tool is sometimes used at build-time by projects that were written in CUDA to convert to HIP at the last minute when targeting AMD GPUs. For example, the rccl library introduced a build dependency on hipify-perl in ROCm 5.5. A hipify-perl package is needed to update the Debian rccl package to the most recent upstream release. This package is part of AMD's ROCm stack and will be maintained under the Debian AI team umbrella.
Bug#1040616: ITP: rocm-validation-suite -- AMD GPU system validation tools
Package: wnpp Severity: wishlist Owner: Cordell Bloor X-Debbugs-Cc: debian-de...@lists.debian.org, c...@slerp.xyz, debian...@lists.debian.org * Package name: rocm-validation-suite Version : 5.6.0 * URL : https://github.com/ROCm-Developer-Tools/ROCmValidationSuite * License : Expat Programming Lang: C++ Description : AMD GPU system validation tools The ROCm Validation Suite (RVS) is a collection of utilities for verifying the correct functioning of the AMD GPUs installed on a system. It provides system administrators with tests, benchmarks and other tools for troubleshooting common problems found in high-performance computing environments. . RVS provides utilities for querying GPU properties, monitoring GPU information, monitoring the PCI Express link speeds and power, querying relevent PCI Express bus properties for a GPU, verifying the GPU SBIOS mapping, benchmarking peer-to-peer links between GPUs, benchmarking the PCI Express bus, stress-testing installed GPUs, stress-testing the system PSU, verifying GPU memory to detect hardware errors, and benchmarking device global memory. This package provides a variety of tools for checking the correct functioning of AMD GPU hardware, which would be valuable for ensuring that malfunctioning hardware and misconfigured systems are identified. It is useful for ruling out hardware problems when unexpected software behaviours are encountered. This package is part of AMD's ROCm stack and will be maintained under the Debian AI team umbrella.
Bug#1040516: ITP: rocdbgapi -- AMD GPU debugger support library
Package: wnpp Severity: wishlist Owner: Cordell Bloor X-Debbugs-Cc: debian-de...@lists.debian.org, c...@slerp.xyz, debian...@lists.debian.org * Package name: rocdbgapi Version : 5.6.0 * URL : https://github.com/ROCm-Developer-Tools/ROCdbgapi * License : Expat Programming Lang: C, C++ Description : AMD GPU debugger support library The ROCdbpgapi library provides debugging routines to inspect and control programs running on AMD GPUs. This library is used by debuggers to receive notifications of events, query the GPU state, and step through execution. However, operations such as symbolic mappings, code object decoding and stack unwinding are outside the scope of this library. The rocdbgapi is a dependency of gdb and, together with the kernel debug interface, allows users to debug GPU code in a similar manner to CPU code. This is therefore an important tool for developing and maintaining AMD GPU software. This package is part of AMD's ROCm stack and will be maintained under the Debian AI team umbrella.
Bug#1037459: ITP: miopen -- ROCm library for machine learning primitives
Package: wnpp Severity: wishlist Owner: Cordell Bloor X-Debbugs-Cc: debian-de...@lists.debian.org, c...@slerp.xyz, debian...@lists.debian.org * Package name: miopen Version : 5.5.0 Upstream Author : Advanced Micro Devices, Inc. * URL : https://github.com/ROCmSoftwarePlatform/MIOpen * License : Expat Programming Lang: C++, HIP Description : ROCm library for machine learning primitives MIOpen is a library for machine learning on AMD GPUs. It provides high-performance primitives for a number of operations, including convolutions, batch normalization, pooling layers, and softmax layers for Recurrent Neural Networks. HIP and OpenCL are both supported as programming models. MIOpen is a key library for machine learning on AMD GPU hardware and is a dependency of both PyTorch and Tensorflow. It is an open-source alternative to nvidia-cudnn. This package is part of AMD's ROCm stack and will be maintained under the Debian AI team umbrella.
Bug#1031761: ITP: hipfort -- Fortran bindings for ROCm and HIP libraries
Package: wnpp Severity: wishlist Owner: Cordell Bloor X-Debbugs-Cc: debian-de...@lists.debian.org, c...@slerp.xyz, debian...@lists.debian.org * Package name: hipfort Version : 5.4.3 Upstream Author : Advanced Micro Devices, Inc. * URL : https://github.com/ROCmSoftwarePlatform/hipfort * License : Expat Programming Lang: Fortran Description : Fortran bindings for ROCm and HIP libraries hipfort is a library that provides Fortran bindings the ROCm and HIP libraries, as well as tooling to help write portable code targeting both AMD and NVIDIA GPUs. . hipfort provides Fortran 2003 and Fortran 2008 bindings for the HIP runtime, rocBLAS, hipBLAS, rocSPARSE, hipSPARSE, rocFFT, hipFFT, rocRAND, hipRAND, rocSOLVER, and hipSOLVER. Additionally, it provides the hipfc compiler wrapper and a Makefile to support a set of standard flags and environment variables for building portable Fortran programs that use GPU compute. Fortran remains a popular language for high-performance computing and bindings for GPU math libraries are useful to enable Fortran users to more fully utilize modern computing hardware. This package is part of the AMD ROCm stack and would be maintained by the Debian ROCm Team.
Bug#1023168: ITP: rocalution -- ROCm library for iterative sparse solvers
Package: wnpp Severity: wishlist Owner: Cordell Bloor X-Debbugs-Cc: debian-de...@lists.debian.org, c...@slerp.xyz, debian...@lists.debian.org * Package name: rocalution Version : 5.3.0 Upstream Author : Advanced Micro Devices, Inc. * URL : https://github.com/ROCmSoftwarePlatform/rocALUTION * License : Expat (MIT/X) Programming Lang: C++ Description : ROCm library for iterative sparse solvers rocALUTION is a library that provides iterative sparse preconditioners and solvers. The rocALUTION project began as a port of PARALUTION to the AMD ROCm platform. As such, it supports an OpenMP backend for multi-core CPUs, a HIP backend for discrete AMD GPUs, and an MPI backend for multi-node clusters and multi-GPU setups. . rocALUTION provides a C++ API containing implementations of fixed-point iteration schemes such as Jacobi iteration and Gauss-Seidel; Krylov subspace methods such as the conjugate gradient method and the biconjugate gradient stabilized method; a mixed-precision defect correction scheme; a Chebyshev iteration scheme; as well as geometric and algebraic multigrid solvers. There are also a wide variety of sparse preconditioners, including several based on matrix splitting schemes, factorization schemes, and approximate inverses. I'm not sure which libraries use rocALUTION. It is one of the few libraries in the ROCm stack that's not a dependency of PyTorch or Tensorflow. However, it is an incredibly useful library nonetheless. These sparse iterative solvers are commonly used in Eulerian fluid simulations for the pressure solve. The pressure solve is typically the slowest part of an incompressible fluid simulation and rocALUTION provides highly optimized and parallelized implementations, enabling much larger simulations than could be run on a single machine. Long ago, I implemented some of these algorithms myself and it took me weeks. Even then, my implementation didn't nearly approach the quality of rocALUTION. (There's probably lots of other uses for rocALUTION, but I did my M.Sc. on fluid sims so that's what I know.) This package is part of AMD's ROCm stack and will be maintained under the Debian AI team umbrella.
Bug#1023098: ITP: hipsolver -- portable interface for GPU LAPACK routines
Package: wnpp Severity: wishlist Owner: Cordell Bloor X-Debbugs-Cc: debian-de...@lists.debian.org, c...@slerp.xyz, debian...@lists.debian.org * Package name: hipsolver Version : 5.3.0 Upstream Author : Advanced Micro Devices, Inc. * URL : https://github.com/ROCmSoftwarePlatform/hipSOLVER * License : Expat (MIT/X) Programming Lang: C, C++ Description : portable interface for GPU LAPACK routines hipSOLVER is a wrapper library that provides a common interface to rocSOLVER and cuSOLVER. The hipSOLVER library is designed to help applications using cuSOLVER to port their code to the ROCm platform. hipSOLVER is a dependency of a number of libraries for scientific computing including Tensorflow, JAX, and PETSc. This package is part of AMD's ROCm stack and will be maintained under the Debian AI team umbrella.
Bug#1023092: ITP: hipsparse -- portable interface for sparse linear algebra on the GPU
Package: wnpp Severity: wishlist Owner: Cordell Bloor X-Debbugs-Cc: debian-de...@lists.debian.org, c...@slerp.xyz, debian...@lists.debian.org * Package name: hipsparse Version : 5.3.0 Upstream Author : Advanced Micro Devices, Inc. * URL : https://github.com/ROCmSoftwarePlatform/hipSPARSE * License : Expat (MIT/X) Programming Lang: C, C++ Description : portable interface for sparse linear algebra on the GPU hipSPARSE is a wrapper library that provides a common interface to rocSPARSE and cuSPARSE. The hipSPARSE library is designed to help applications using cuSPARSE to port their code to the ROCm platform. hipSPARSE is a dependency of a number of libraries for scientific computing including PyTorch and Tensorflow. This package is part of AMD's ROCm stack and will be maintained under the Debian AI team umbrella.
Bug#1023091: ITP: hipfft -- portable interface for Fast Fourier Transforms on the GPU
Package: wnpp Severity: wishlist Owner: Cordell Bloor X-Debbugs-Cc: debian-de...@lists.debian.org, c...@slerp.xyz, debian...@lists.debian.org * Package name: hipfft Version : 5.3.0 Upstream Author : Advanced Micro Devices, Inc. * URL : https://github.com/ROCmSoftwarePlatform/hipFFT * License : Expat (MIT/X) Programming Lang: C, C++ Description : portable interface for Fast Fourier Transforms on the GPU hipFFT is a wrapper library that provides a common interface to rocFFT and cuFFT. The hipFFT library is designed to help applications using cuFFT to port their code to the ROCm platform. hipFFT is a dependency of a number of libraries for scientific computing including PyTorch and Tensorflow. This package is part of AMD's ROCm stack and will be maintained under the Debian AI team umbrella.
Bug#1023090: ITP: hipcub -- portable interface for GPU parallel primitives
Package: wnpp Severity: wishlist Owner: Cordell Bloor X-Debbugs-Cc: debian-de...@lists.debian.org, c...@slerp.xyz, debian...@lists.debian.org * Package name: hipcub Version : 5.3.0 Upstream Author : Advanced Micro Devices, Inc. * URL : https://github.com/ROCmSoftwarePlatform/hipCUB * License : BSD 3-Clause Programming Lang: HIP, CUDA Description : portable interface for GPU parallel primitives hipCUB is a wrapper library that provides a common interface to rocPRIM and CUB. The hipCUB library is designed to help applications using CUB to port their code to the ROCm platform. hipCUB is a dependency of a number of libraries for scientific computing including PyTorch and Tensorflow. This package is part of AMD's ROCm stack and will be maintained under the Debian AI team umbrella.
Bug#1023087: ITP: hipblas -- portable interface for GPU BLAS
Package: wnpp Severity: wishlist Owner: Cordell Bloor X-Debbugs-Cc: debian-de...@lists.debian.org, c...@slerp.xyz, debian...@lists.debian.org * Package name: hipblas Version : 5.3.0 Upstream Author : Advanced Micro Devices, Inc. * URL : https://github.com/ROCmSoftwarePlatform/hipBLAS * License : Expat (MIT/X) Programming Lang: C, C++ Description : portable interface for GPU BLAS hipBLAS is a wrapper library that provides a common interface to rocBLAS and cuBLAS. The hipBLAS library is designed to help applications using cuBLAS to port their code to the ROCm platform. hipBLAS is a dependency of a number of libraries for scientific computing including libCEED, Ginkgo, pika, DBCSR, and SuperLU. This package is part of AMD's ROCm stack and will be maintained under the Debian AI team umbrella.
Bug#1023081: ITP: rocsolver -- ROCm library for numerical linear algebra
Package: wnpp Severity: wishlist Owner: Cordell Bloor X-Debbugs-Cc: debian-de...@lists.debian.org, debian...@lists.debian.org * Package name: rocsolver Version : 5.3.0 Upstream Author : Advanced Micro Devices, Inc. * URL : https://github.com/ROCmSoftwarePlatform/rocSOLVER * License : BSD 2-Clause Programming Lang: C, C++, HIP Description : ROCm library for numerical linear algebra rocSOLVER is an implementation of LAPACK routines built on the AMD ROCm platform and optimized for discrete AMD GPUs. . rocSOLVER provides a C99 API that extends the rocBLAS interface. It includes LAPACK and LAPACK-like routines for triangular factorizations, orthogonal factorizations, linear-systems solvers, least-squares solvers, symmetric eigensolvers and singular value decomposition. rocSOLVER is a dependency of a number of libraries for scientific computing including Tensorflow, Tasmanian, and QMCPACK, as well as other components of the ROCm platform such as hipBLAS and hipSOLVER. This package is part of AMD's ROCm stack and will be maintained under the Debian AI team umbrella.
Bug#1023078: ITP: rocsparse -- ROCm library for sparse linear algebra
Package: wnpp Severity: wishlist Owner: Cordell Bloor X-Debbugs-Cc: debian-de...@lists.debian.org, debian...@lists.debian.org * Package name: rocsparse Version : 5.3.0 Upstream Author : Advanced Micro Devices, Inc. * URL : https://github.com/ROCmSoftwarePlatform/rocSPARSE * License : Expat (MIT/X) Programming Lang: C, C++, HIP Description : ROCm library for sparse linear algebra rocSPARSE is a library that implements BLAS operations for sparse data structures. It is built on the AMD ROCm platform and optimized for discrete AMD GPUs. . rocSPARSE provides a C99 API containing sparse level 1, sparse level 2 and sparse level 3 BLAS functions, as well as general matrix multiplication between dense and sparse matricies (GEMMI), sparse preconditioner functions, sparse matrix format conversion functions, and sparse matrix reordering functions. . It supports a variety of sparse matrix formats, including coordinate list in either array of structure form (COO) or structure of array form (COO SOA), compressed sparse row (CSR) or column (CSC), block compressed sparse row (BSR), general block compressed sparse row (GEBSR), and the ELLPACK-ITPACK format (ELL). rocSPARSE is a dependency of a number of libraries for scientific computing that can use AMD GPU hardware including PyTorch, Tensorflow, PETSc, Hypre, Tasmanian, and MFEM, as well as other ROCm libraries such as hipSPARSE and rocALUTION. This package is part of AMD's ROCm stack and will be maintained under the Debian AI team umbrella.
Bug#1022889: ITP: rccl -- ROCm Communication Collectives Library
Package: wnpp Severity: wishlist Owner: Cordell Bloor X-Debbugs-Cc: debian-de...@lists.debian.org, c...@slerp.xyz, debian...@lists.debian.org * Package name: rccl Version : 5.3.0 Upstream Author : Advanced Micro Devices, Inc. * URL : https://github.com/ROCmSoftwarePlatform/rccl * License : BSD 3-Clause Programming Lang: C, C++, HIP Description : ROCm Communication Collectives Library RCCL (pronounced "Rickle") is a library of collective communication routines built on the AMD ROCm platform and optimized for discrete AMD GPUs. It implements routines such as all-reduce, all-gather, reduce, broadcast, reduce-scatter, gather, scatter and all-to-all. The library is optimized for high bandwidth over channels including PCIe, xGMI, InfiniBand Verbs or TCP/IP sockets and supports an arbitrary number of GPUs. This library provides multi-GPU communication functionality that helps users depending on NVIDIA's NCCL library to port their code to run on AMD GPUs. RCCL is a dependency of both PyTorch and Tensorflow. This package is part of AMD's ROCm stack and will be maintained under the Debian AI team umbrella.
Bug#1022871: ITP: rocblas -- ROCm library for basic linear algebra
Package: wnpp Severity: wishlist Owner: Cordell Bloor X-Debbugs-Cc: debian-de...@lists.debian.org, c...@slerp.xyz, debian...@lists.debian.org * Package name: rocblas Version : 5.3.0 Upstream Author : Advanced Micro Devices, Inc. * URL : https://github.com/ROCmSoftwarePlatform/rocBLAS * License : Expat (MIT/X) Programming Lang: C, C++, HIP Description : ROCm library for basic linear algebra rocBLAS is a library for performing basic linear algebra routines on AMD GPUs. It provides implementations of BLAS operations such as dot product (DOT), general matrix multiplication (GEMM), triangular matrix multiplication (TRMM), and many others. It is built on the AMD ROCm platform and optimized for discrete AMD GPUs. The rocBLAS library is implemented in the HIP programming language and is supplemented by tuned assembly kernels. rocBLAS provides a C99 interface and supports operations on matrices of half-, single- or double-precision floating point numbers. The library provides variants for its routines that efficiently operate on batches of matrices and that support a variety of data layouts. rocBLAS is a key package required to leverage AMD GPU hardware for scientific computing and AI. Many applications in that domain benefit significantly from fast matrix multiplication. The rocBLAS library is a dependency of PyTorch, Tensorflow, Tasmanian, pika and others. It's also required by other components of the ROCm platform such as hipBLAS, rocSOLVER, hipSOLVER, rocALUTION, MIOpen, and MIGraphX.
Bug#1022869: ITP: rocfft -- ROCm library for computing Fast Fourier Transforms
The correct value for the URL field should be: https://github.com/ROCmSoftwarePlatform/rocFFT Apologies for the mistake. I didn't have my mail client set up correctly the first time I submitted and the original report was lost into the ether. It seems that I rushed through filling out the fields the second time around. Sincerely, Cory Bloor
Bug#1022869: ITP: rocfft -- ROCm library for computing Fast Fourier Transforms
Package: wnpp Severity: wishlist Owner: Cordell Bloor X-Debbugs-Cc: debian-de...@lists.debian.org, debian...@lists.debian.org * Package name: rocfft Version : 5.3.0 Upstream Author : Advanced Micro Devices, Inc. * URL : https://www.example.org/ * License : Expat (MIT/X) Programming Lang: C, C++, HIP Description : ROCm library for computing Fast Fourier Transforms rocFFT is a library for computing the discrete Fourier transform. It is built on the AMD ROCm platform and optimized for discrete AMD GPUs. rocFFT provides a C99 API and makes use of run-time compilation to create optimized execution plans for its computations. The rocFFT library supports single and double precision floating point formats for calculating one-, two- and three-dimensional transforms for real or complex data of arbitrary length. rocFFT is a dependency of scientific libraries such as PyTorch, Tensorflow and WarpX as well as other libraries on the ROCm platform such as hipFFT. This package is part of AMD's ROCm stack and will be maintained under the Debian AI team umbrella.
Bug#1021695: ITP: rocprim -- parallel primitives for GPU-accelerated code
Package: wnpp Severity: wishlist Owner: Cordell Bloor X-Debbugs-Cc: debian-de...@lists.debian.org, debian...@lists.debian.org * Package name : rocprim Version : 5.3.0 Upstream Author : Advanced Micro Devices, Inc. * URL : https://github.com/ROCmSoftwarePlatform/rocPRIM * License : Expat (MIT/X) Programming Lang: HIP Description : parallel primitives for GPU-accelerated code rocPRIM is a header-only library providing HIP parallel primitives for developing performant GPU-accelerated code on the AMD ROCm platform. This package is a dependency of several libraries for scientific computing on AMD GPUs, including rocthrust, rocsparse, rocalution, hipcub, pytorch and tensorflow. It fulfills a similar role in the AMD ROCm platform as CUB does in NVIDIA's CUDA platform. This package is part of AMD's ROCm stack and will be maintained under the Debian AI team umbrella.
Bug#1021863: ITP: rocthrust -- ROCm port of the Thrust parallel algorithm library
Package: wnpp Severity: wishlist Owner: Cordell Bloor X-Debbugs-Cc: debian-de...@lists.debian.org, c...@slerp.xyz, debian...@lists.debian.org * Package name: rocthrust Version : 5.3.0 Upstream Author : Advanced Micro Devices, Inc. * URL : https://github.com/ROCmSoftwarePlatform/rocThrust * License : Apache-2.0 Programming Lang: HIP Description : ROCm port of the Thrust parallel algorithm library rocThrust is a header-only library that provides common data structures and algorithms for high-performance GPU code. rocThrust is intended for use with AMD GPUs via the ROCm platform. It is implemented in the HIP language and is a source-compatible replacement for the CUDA Thrust library. This package is a dependency of a number of libraries for scientific computing on AMD GPUs, including rocalution and pytorch.
Bug#1021695: Correcting the rocprim ITP
retitle 1021695 ITP: rocprim -- parallel primitives for GPU-accelerated code thanks The source package name is rocprim and the binary package is librocprim-dev. My apologies for the error. I'm slowing getting the hang of the Debian tooling and processes, but I'm not quite there yet. I also made a mistake in stating its language. All the headers are *.hpp files. There's no C API. I'm also a bit unclear on whether any of the headers could be used with a normal C++ compiler and which headers require HIP extensions. In any case, rocPRIM would be the second HIP language library to be packaged for Debian after rocRAND. Sincerely, Cory Bloor
Bug#1021695: ITP: librocprim2-dev -- parallel primatives for GPU-accelerated code
Package: wnpp Severity: wishlist Owner: Cordell Bloor X-Debbugs-Cc: debian-de...@lists.debian.org, debian...@lists.debian.org * Package name: librocprim2-dev Version : 5.3.0 Upstream Author : Advanced Micro Devices, Inc. * URL : https://github.com/ROCmSoftwarePlatform/rocPRIM * License : Expat (MIT/X) Programming Lang: C, C++ Description : parallel primatives for GPU-accelerated code rocPRIM is a header-only library providing HIP parallel primitives for developing performant GPU-accelerated code on the AMD ROCm platform. This package is a dependency of several libraries for scientific computing on AMD GPUs, including rocthrust, rocsparse, rocalution, hipcub, pytorch and tensorflow. It fulfills a similar role in the AMD ROCm platform as CUB does in NVIDIA's CUDA platform. This package is part of AMD's ROCm stack and will be maintained under the Debian AI team umbrella.