1# Changes/Release Notes 2 3On this page we provide a summary of the main API changes, new features and examples 4for each release of libCEED. 5 6(main)= 7 8## Current `main` branch 9 10### Interface changes 11 12- Update {c:func}`CeedQFunctionGetFields` and {c:func}`CeedOperatorGetFields` to include number of fields. 13- Promote to the public API: QFunction and Operator field objects, `CeedQFunctionField` and `CeedOperatorField`, and associated getters, {c:func}`CeedQFunctionGetFields`; {c:func}`CeedQFunctionFieldGetName`; {c:func}`CeedQFunctionFieldGetSize`; {c:func}`CeedQFunctionFieldGetEvalMode`; {c:func}`CeedOperatorGetFields`; {c:func}`CeedOperatorFieldGetElemRestriction`; {c:func}`CeedOperatorFieldGetBasis`; and {c:func}`CeedOperatorFieldGetVector`. 14- Clarify and document conditions where `CeedQFunction` and `CeedOperator` become immutable and no further fields or suboperators can be added. 15- Add {c:func}`CeedOperatorLinearAssembleQFunctionBuildOrUpdate` to reduce object creation overhead in assembly of CeedOperator preconditioning ingredients. 16- Promote {c:func}`CeedOperatorCheckReady`to the public API to facilitate interactive interfaces. 17- Warning added when compiling OCCA backend to alert users that this backend is experimental. 18- `ceed-backend.h`, `ceed-hash.h`, and `ceed-khash.h` removed. Users should use `ceed/backend.h`, `ceed/hash.h`, and `ceed/khash.h`. 19- Added {c:func}`CeedQFunctionGetKernelName`; refactored {c:func}`CeedQFunctionGetSourcePath` to exclude function kernel name. 20 21### New features 22 23- `CeedScalar` can now be set as `float` or `double` at compile time. 24- Added JiT utilities in `ceed/jit-tools.h` to reduce duplicated code in GPU backends. 25- Added support for JiT of QFunctions with `#include "relative/path/local-file.h"` statements for additional local files. Note that files included with `""` are searched relative to the current file first, then by compiler paths (as with `<>` includes). To use this feature, one should adhere to relative paths only, not compiler flags like `-I`, which the JiT will not be aware of. 26- Remove need to guard library headers in QFunction source for code generation backends. 27- `CeedDebugEnv()` macro created to provide debugging outputs when Ceed context is not present. 28 29### Maintainability 30 31- Refactored preconditioner support internally to facilitate future development and improve GPU completeness/test coverage. 32- `Include-what-you-use` makefile target added as `make iwyu`. 33- Create backend constant `CEED_FIELD_MAX` to reduce magic numbers in codebase. 34 35(v0-9)= 36 37## v0.9 (Jul 6, 2021) 38 39### Interface changes 40 41- Minor modification in error handling macro to silence pedantic warnings when compiling with Clang, but no functional impact. 42 43### New features 44 45- Add {c:func}`CeedVectorAXPY` and {c:func}`CeedVectorPointwiseMult` as a convenience for stand-alone testing and internal use. 46- Add `CEED_QFUNCTION_HELPER` macro to properly annotate QFunction helper functions for code generation backends. 47- Add `CeedPragmaOptimizeOff` macro for code that is sensitive to floating point errors from fast math optimizations. 48- Rust support: split `libceed-sys` crate out of `libceed` and [publish both on crates.io](https://crates.io/crates/libceed). 49 50### Performance improvements 51 52### Examples 53 54- Solid mechanics mini-app updated to explore the performance impacts of various formulations in the initial and current configurations. 55- Fluid mechanics example adds GPU support and improves modularity. 56 57### Deprecated backends 58 59- The `/cpu/self/tmpl` and `/cpu/self/tmpl/sub` backends have been removed. These backends were intially added to test the backend inheritance mechanism, but this mechanism is now widely used and tested in multiple backends. 60 61(v0-8)= 62 63## v0.8 (Mar 31, 2021) 64 65### Interface changes 66 67- Error handling improved to include enumerated error codes for C interface return values. 68- Installed headers that will follow semantic versioning were moved to {code}`include/ceed` directory. These headers have been renamed from {code}`ceed-*.h` to {code}`ceed/*.h`. Placeholder headers with the old naming schema are currently provided, but these headers will be removed in the libCEED v0.9 release. 69 70### New features 71 72- Julia and Rust interfaces added, providing a nearly 1-1 correspondence with the C interface, plus some convenience features. 73- Static libraries can be built with `make STATIC=1` and the pkg-config file is installed accordingly. 74- Add {c:func}`CeedOperatorLinearAssembleSymbolic` and {c:func}`CeedOperatorLinearAssemble` to support full assembly of libCEED operators. 75 76### Performance improvements 77 78- New HIP MAGMA backends for hipMAGMA library users: `/gpu/hip/magma` and `/gpu/hip/magma/det`. 79- New HIP backends for improved tensor basis performance: `/gpu/hip/shared` and `/gpu/hip/gen`. 80 81### Examples 82 83- {ref}`example-petsc-elasticity` example updated with traction boundary conditions and improved Dirichlet boundary conditions. 84- {ref}`example-petsc-elasticity` example updated with Neo-Hookean hyperelasticity in current configuration as well as improved Neo-Hookean hyperelasticity exploring storage vs computation tradeoffs. 85- {ref}`example-petsc-navier-stokes` example updated with isentropic traveling vortex test case, an analytical solution to the Euler equations that is useful for testing boundary conditions, discretization stability, and order of accuracy. 86- {ref}`example-petsc-navier-stokes` example updated with support for performing convergence study and plotting order of convergence by polynomial degree. 87 88(v0-7)= 89 90## v0.7 (Sep 29, 2020) 91 92### Interface changes 93 94- Replace limited {code}`CeedInterlaceMode` with more flexible component stride {code}`compstride` in {code}`CeedElemRestriction` constructors. 95 As a result, the {code}`indices` parameter has been replaced with {code}`offsets` and the {code}`nnodes` parameter has been replaced with {code}`lsize`. 96 These changes improve support for mixed finite element methods. 97- Replace various uses of {code}`Ceed*Get*Status` with {code}`Ceed*Is*` in the backend API to match common nomenclature. 98- Replace {code}`CeedOperatorAssembleLinearDiagonal` with {c:func}`CeedOperatorLinearAssembleDiagonal` for clarity. 99- Linear Operators can be assembled as point-block diagonal matrices with {c:func}`CeedOperatorLinearAssemblePointBlockDiagonal`, provided in row-major form in a {code}`ncomp` by {code}`ncomp` block per node. 100- Diagonal assemble interface changed to accept a {ref}`CeedVector` instead of a pointer to a {ref}`CeedVector` to reduce memory movement when interfacing with calling code. 101- Added {c:func}`CeedOperatorLinearAssembleAddDiagonal` and {c:func}`CeedOperatorLinearAssembleAddPointBlockDiagonal` for improved future integration with codes such as MFEM that compose the action of {ref}`CeedOperator`s external to libCEED. 102- Added {c:func}`CeedVectorTakeAray` to sync and remove libCEED read/write access to an allocated array and pass ownership of the array to the caller. 103 This function is recommended over {c:func}`CeedVectorSyncArray` when the {code}`CeedVector` has an array owned by the caller that was set by {c:func}`CeedVectorSetArray`. 104- Added {code}`CeedQFunctionContext` object to manage user QFunction context data and reduce copies between device and host memory. 105- Added {c:func}`CeedOperatorMultigridLevelCreate`, {c:func}`CeedOperatorMultigridLevelCreateTensorH1`, and {c:func}`CeedOperatorMultigridLevelCreateH1` to facilitate creation of multigrid prolongation, restriction, and coarse grid operators using a common quadrature space. 106 107### New features 108 109- New HIP backend: `/gpu/hip/ref`. 110- CeedQFunction support for user `CUfunction`s in some backends 111 112### Performance improvements 113 114- OCCA backend rebuilt to facilitate future performance enhancements. 115- Petsc BPs suite improved to reduce noise due to multiple calls to {code}`mpiexec`. 116 117### Examples 118 119- {ref}`example-petsc-elasticity` example updated with strain energy computation and more flexible boundary conditions. 120 121### Deprecated backends 122 123- The `/gpu/cuda/reg` backend has been removed, with its core features moved into `/gpu/cuda/ref` and `/gpu/cuda/shared`. 124 125(v0-6)= 126 127## v0.6 (Mar 29, 2020) 128 129libCEED v0.6 contains numerous new features and examples, as well as expanded 130documentation in [this new website](https://libceed.readthedocs.io). 131 132### New features 133 134- New Python interface using [CFFI](https://cffi.readthedocs.io/) provides a nearly 135 1-1 correspondence with the C interface, plus some convenience features. For instance, 136 data stored in the {cpp:type}`CeedVector` structure are available without copy as 137 {py:class}`numpy.ndarray`. Short tutorials are provided in 138 [Binder](https://mybinder.org/v2/gh/CEED/libCEED/main?urlpath=lab/tree/examples/tutorials/). 139- Linear QFunctions can be assembled as block-diagonal matrices (per quadrature point, 140 {c:func}`CeedOperatorAssembleLinearQFunction`) or to evaluate the diagonal 141 ({c:func}`CeedOperatorAssembleLinearDiagonal`). These operations are useful for 142 preconditioning ingredients and are used in the libCEED's multigrid examples. 143- The inverse of separable operators can be obtained using 144 {c:func}`CeedOperatorCreateFDMElementInverse` and applied with 145 {c:func}`CeedOperatorApply`. This is a useful preconditioning ingredient, 146 especially for Laplacians and related operators. 147- New functions: {c:func}`CeedVectorNorm`, {c:func}`CeedOperatorApplyAdd`, 148 {c:func}`CeedQFunctionView`, {c:func}`CeedOperatorView`. 149- Make public accessors for various attributes to facilitate writing composable code. 150- New backend: `/cpu/self/memcheck/serial`. 151- QFunctions using variable-length array (VLA) pointer constructs can be used with CUDA 152 backends. (Single source is coming soon for OCCA backends.) 153- Fix some missing edge cases in CUDA backend. 154 155### Performance Improvements 156 157- MAGMA backend performance optimization and non-tensor bases. 158- No-copy optimization in {c:func}`CeedOperatorApply`. 159 160### Interface changes 161 162- Replace {code}`CeedElemRestrictionCreateIdentity` and 163 {code}`CeedElemRestrictionCreateBlocked` with more flexible 164 {c:func}`CeedElemRestrictionCreateStrided` and 165 {c:func}`CeedElemRestrictionCreateBlockedStrided`. 166- Add arguments to {c:func}`CeedQFunctionCreateIdentity`. 167- Replace ambiguous uses of {cpp:enum}`CeedTransposeMode` for L-vector identification 168 with {cpp:enum}`CeedInterlaceMode`. This is now an attribute of the 169 {cpp:type}`CeedElemRestriction` (see {c:func}`CeedElemRestrictionCreate`) and no 170 longer passed as `lmode` arguments to {c:func}`CeedOperatorSetField` and 171 {c:func}`CeedElemRestrictionApply`. 172 173### Examples 174 175libCEED-0.6 contains greatly expanded examples with {ref}`new documentation <Examples>`. 176Notable additions include: 177 178- Standalone {ref}`ex2-surface` ({file}`examples/ceed/ex2-surface`): compute the area of 179 a domain in 1, 2, and 3 dimensions by applying a Laplacian. 180 181- PETSc {ref}`example-petsc-area` ({file}`examples/petsc/area.c`): computes surface area 182 of domains (like the cube and sphere) by direct integration on a surface mesh; 183 demonstrates geometric dimension different from topological dimension. 184 185- PETSc {ref}`example-petsc-bps`: 186 187 - {file}`examples/petsc/bpsraw.c` (formerly `bps.c`): transparent CUDA support. 188 - {file}`examples/petsc/bps.c` (formerly `bpsdmplex.c`): performance improvements 189 and transparent CUDA support. 190 - {ref}`example-petsc-bps-sphere` ({file}`examples/petsc/bpssphere.c`): 191 generalizations of all CEED BPs to the surface of the sphere; demonstrates geometric 192 dimension different from topological dimension. 193 194- {ref}`example-petsc-multigrid` ({file}`examples/petsc/multigrid.c`): new p-multigrid 195 solver with algebraic multigrid coarse solve. 196 197- {ref}`example-petsc-navier-stokes` ({file}`examples/fluids/navierstokes.c`; formerly 198 `examples/navier-stokes`): unstructured grid support (using PETSc's `DMPlex`), 199 implicit time integration, SU/SUPG stabilization, free-slip boundary conditions, and 200 quasi-2D computational domain support. 201 202- {ref}`example-petsc-elasticity` ({file}`examples/solids/elasticity.c`): new solver for 203 linear elasticity, small-strain hyperelasticity, and globalized finite-strain 204 hyperelasticity using p-multigrid with algebraic multigrid coarse solve. 205 206(v0-5)= 207 208## v0.5 (Sep 18, 2019) 209 210For this release, several improvements were made. Two new CUDA backends were added to 211the family of backends, of which, the new `cuda-gen` backend achieves state-of-the-art 212performance using single-source {ref}`CeedQFunction`. From this release, users 213can define Q-Functions in a single source code independently of the targeted backend 214with the aid of a new macro `CEED QFUNCTION` to support JIT (Just-In-Time) and CPU 215compilation of the user provided {ref}`CeedQFunction` code. To allow a unified 216declaration, the {ref}`CeedQFunction` API has undergone a slight change: 217the `QFunctionField` parameter `ncomp` has been changed to `size`. This change 218requires setting the previous value of `ncomp` to `ncomp*dim` when adding a 219`QFunctionField` with eval mode `CEED EVAL GRAD`. 220 221Additionally, new CPU backends 222were included in this release, such as the `/cpu/self/opt/*` backends (which are 223written in pure C and use partial **E-vectors** to improve performance) and the 224`/cpu/self/ref/memcheck` backend (which relies upon the 225[Valgrind](http://valgrind.org/) Memcheck tool to help verify that user 226{ref}`CeedQFunction` have no undefined values). 227This release also included various performance improvements, bug fixes, new examples, 228and improved tests. Among these improvements, vectorized instructions for 229{ref}`CeedQFunction` code compiled for CPU were enhanced by using `CeedPragmaSIMD` 230instead of `CeedPragmaOMP`, implementation of a {ref}`CeedQFunction` gallery and 231identity Q-Functions were introduced, and the PETSc benchmark problems were expanded 232to include unstructured meshes handling were. For this expansion, the prior version of 233the PETSc BPs, which only included data associated with structured geometries, were 234renamed `bpsraw`, and the new version of the BPs, which can handle data associated 235with any unstructured geometry, were called `bps`. Additionally, other benchmark 236problems, namely BP2 and BP4 (the vector-valued versions of BP1 and BP3, respectively), 237and BP5 and BP6 (the collocated versions---for which the quadrature points are the same 238as the Gauss Lobatto nodes---of BP3 and BP4 respectively) were added to the PETSc 239examples. Furthermoew, another standalone libCEED example, called `ex2`, which 240computes the surface area of a given mesh was added to this release. 241 242Backends available in this release: 243 244| CEED resource (`-ceed`) | Backend | 245|--------------------------|-----------------------------------------------------| 246| `/cpu/self/ref/serial` | Serial reference implementation | 247| `/cpu/self/ref/blocked` | Blocked reference implementation | 248| `/cpu/self/ref/memcheck` | Memcheck backend, undefined value checks | 249| `/cpu/self/opt/serial` | Serial optimized C implementation | 250| `/cpu/self/opt/blocked` | Blocked optimized C implementation | 251| `/cpu/self/avx/serial` | Serial AVX implementation | 252| `/cpu/self/avx/blocked` | Blocked AVX implementation | 253| `/cpu/self/xsmm/serial` | Serial LIBXSMM implementation | 254| `/cpu/self/xsmm/blocked` | Blocked LIBXSMM implementation | 255| `/cpu/occa` | Serial OCCA kernels | 256| `/gpu/occa` | CUDA OCCA kernels | 257| `/omp/occa` | OpenMP OCCA kernels | 258| `/ocl/occa` | OpenCL OCCA kernels | 259| `/gpu/cuda/ref` | Reference pure CUDA kernels | 260| `/gpu/cuda/reg` | Pure CUDA kernels using one thread per element | 261| `/gpu/cuda/shared` | Optimized pure CUDA kernels using shared memory | 262| `/gpu/cuda/gen` | Optimized pure CUDA kernels using code generation | 263| `/gpu/magma` | CUDA MAGMA kernels | 264 265Examples available in this release: 266 267:::{list-table} 268:header-rows: 1 269:widths: auto 270* - User code 271 - Example 272* - `ceed` 273 - * ex1 (volume) 274 * ex2 (surface) 275* - `mfem` 276 - * BP1 (scalar mass operator) 277 * BP3 (scalar Laplace operator) 278* - `petsc` 279 - * BP1 (scalar mass operator) 280 * BP2 (vector mass operator) 281 * BP3 (scalar Laplace operator) 282 * BP4 (vector Laplace operator) 283 * BP5 (collocated scalar Laplace operator) 284 * BP6 (collocated vector Laplace operator) 285 * Navier-Stokes 286* - `nek5000` 287 - * BP1 (scalar mass operator) 288 * BP3 (scalar Laplace operator) 289::: 290 291(v0-4)= 292 293## v0.4 (Apr 1, 2019) 294 295libCEED v0.4 was made again publicly available in the second full CEED software 296distribution, release CEED 2.0. This release contained notable features, such as 297four new CPU backends, two new GPU backends, CPU backend optimizations, initial 298support for operator composition, performance benchmarking, and a Navier-Stokes demo. 299The new CPU backends in this release came in two families. The `/cpu/self/*/serial` 300backends process one element at a time and are intended for meshes with a smaller number 301of high order elements. The `/cpu/self/*/blocked` backends process blocked batches of 302eight interlaced elements and are intended for meshes with higher numbers of elements. 303The `/cpu/self/avx/*` backends rely upon AVX instructions to provide vectorized CPU 304performance. The `/cpu/self/xsmm/*` backends rely upon the 305[LIBXSMM](http://github.com/hfp/libxsmm) package to provide vectorized CPU 306performance. The `/gpu/cuda/*` backends provide GPU performance strictly using CUDA. 307The `/gpu/cuda/ref` backend is a reference CUDA backend, providing reasonable 308performance for most problem configurations. The `/gpu/cuda/reg` backend uses a simple 309parallelization approach, where each thread treats a finite element. Using just in time 310compilation, provided by nvrtc (NVidia Runtime Compiler), and runtime parameters, this 311backend unroll loops and map memory address to registers. The `/gpu/cuda/reg` backend 312achieve good peak performance for 1D, 2D, and low order 3D problems, but performance 313deteriorates very quickly when threads run out of registers. 314 315A new explicit time-stepping Navier-Stokes solver was added to the family of libCEED 316examples in the `examples/petsc` directory (see {ref}`example-petsc-navier-stokes`). 317This example solves the time-dependent Navier-Stokes equations of compressible gas 318dynamics in a static Eulerian three-dimensional frame, using structured high-order 319finite/spectral element spatial discretizations and explicit high-order time-stepping 320(available in PETSc). Moreover, the Navier-Stokes example was developed using PETSc, 321so that the pointwise physics (defined at quadrature points) is separated from the 322parallelization and meshing concerns. 323 324Backends available in this release: 325 326| CEED resource (`-ceed`) | Backend | 327|--------------------------|-----------------------------------------------------| 328| `/cpu/self/ref/serial` | Serial reference implementation | 329| `/cpu/self/ref/blocked` | Blocked reference implementation | 330| `/cpu/self/tmpl` | Backend template, defaults to `/cpu/self/blocked` | 331| `/cpu/self/avx/serial` | Serial AVX implementation | 332| `/cpu/self/avx/blocked` | Blocked AVX implementation | 333| `/cpu/self/xsmm/serial` | Serial LIBXSMM implementation | 334| `/cpu/self/xsmm/blocked` | Blocked LIBXSMM implementation | 335| `/cpu/occa` | Serial OCCA kernels | 336| `/gpu/occa` | CUDA OCCA kernels | 337| `/omp/occa` | OpenMP OCCA kernels | 338| `/ocl/occa` | OpenCL OCCA kernels | 339| `/gpu/cuda/ref` | Reference pure CUDA kernels | 340| `/gpu/cuda/reg` | Pure CUDA kernels using one thread per element | 341| `/gpu/magma` | CUDA MAGMA kernels | 342 343Examples available in this release: 344 345:::{list-table} 346:header-rows: 1 347:widths: auto 348* - User code 349 - Example 350* - `ceed` 351 - * ex1 (volume) 352* - `mfem` 353 - * BP1 (scalar mass operator) 354 * BP3 (scalar Laplace operator) 355* - `petsc` 356 - * BP1 (scalar mass operator) 357 * BP3 (scalar Laplace operator) 358 * Navier-Stokes 359* - `nek5000` 360 - * BP1 (scalar mass operator) 361 * BP3 (scalar Laplace operator) 362::: 363 364(v0-3)= 365 366## v0.3 (Sep 30, 2018) 367 368Notable features in this release include active/passive field interface, support for 369non-tensor bases, backend optimization, and improved Fortran interface. This release 370also focused on providing improved continuous integration, and many new tests with code 371coverage reports of about 90%. This release also provided a significant change to the 372public interface: a {ref}`CeedQFunction` can take any number of named input and output 373arguments while {ref}`CeedOperator` connects them to the actual data, which may be 374supplied explicitly to `CeedOperatorApply()` (active) or separately via 375`CeedOperatorSetField()` (passive). This interface change enables reusable libraries 376of CeedQFunctions and composition of block solvers constructed using 377{ref}`CeedOperator`. A concept of blocked restriction was added to this release and 378used in an optimized CPU backend. Although this is typically not visible to the user, 379it enables effective use of arbitrary-length SIMD while maintaining cache locality. 380This CPU backend also implements an algebraic factorization of tensor product gradients 381to perform fewer operations than standard application of interpolation and 382differentiation from nodes to quadrature points. This algebraic formulation 383automatically supports non-polynomial and non-interpolatory bases, thus is more general 384than the more common derivation in terms of Lagrange polynomials on the quadrature points. 385 386Backends available in this release: 387 388| CEED resource (`-ceed`) | Backend | 389|-------------------------|-----------------------------------------------------| 390| `/cpu/self/blocked` | Blocked reference implementation | 391| `/cpu/self/ref` | Serial reference implementation | 392| `/cpu/self/tmpl` | Backend template, defaults to `/cpu/self/blocked` | 393| `/cpu/occa` | Serial OCCA kernels | 394| `/gpu/occa` | CUDA OCCA kernels | 395| `/omp/occa` | OpenMP OCCA kernels | 396| `/ocl/occa` | OpenCL OCCA kernels | 397| `/gpu/magma` | CUDA MAGMA kernels | 398 399Examples available in this release: 400 401:::{list-table} 402:header-rows: 1 403:widths: auto 404* - User code 405 - Example 406* - `ceed` 407 - * ex1 (volume) 408* - `mfem` 409 - * BP1 (scalar mass operator) 410 * BP3 (scalar Laplace operator) 411* - `petsc` 412 - * BP1 (scalar mass operator) 413 * BP3 (scalar Laplace operator) 414* - `nek5000` 415 - * BP1 (scalar mass operator) 416 * BP3 (scalar Laplace operator) 417::: 418 419(v0-21)= 420 421## v0.21 (Sep 30, 2018) 422 423A MAGMA backend (which relies upon the 424[MAGMA](https://bitbucket.org/icl/magma) package) was integrated in libCEED for this 425release. This initial integration set up the framework of using MAGMA and provided the 426libCEED functionality through MAGMA kernels as one of libCEED’s computational backends. 427As any other backend, the MAGMA backend provides extended basic data structures for 428{ref}`CeedVector`, {ref}`CeedElemRestriction`, and {ref}`CeedOperator`, and implements 429the fundamental CEED building blocks to work with the new data structures. 430In general, the MAGMA-specific data structures keep the libCEED pointers to CPU data 431but also add corresponding device (e.g., GPU) pointers to the data. Coherency is handled 432internally, and thus seamlessly to the user, through the functions/methods that are 433provided to support them. 434 435Backends available in this release: 436 437| CEED resource (`-ceed`) | Backend | 438|-------------------------|---------------------------------| 439| `/cpu/self` | Serial reference implementation | 440| `/cpu/occa` | Serial OCCA kernels | 441| `/gpu/occa` | CUDA OCCA kernels | 442| `/omp/occa` | OpenMP OCCA kernels | 443| `/ocl/occa` | OpenCL OCCA kernels | 444| `/gpu/magma` | CUDA MAGMA kernels | 445 446Examples available in this release: 447 448:::{list-table} 449:header-rows: 1 450:widths: auto 451* - User code 452 - Example 453* - `ceed` 454 - * ex1 (volume) 455* - `mfem` 456 - * BP1 (scalar mass operator) 457 * BP3 (scalar Laplace operator) 458* - `petsc` 459 - * BP1 (scalar mass operator) 460* - `nek5000` 461 - * BP1 (scalar mass operator) 462::: 463 464(v0-2)= 465 466## v0.2 (Mar 30, 2018) 467 468libCEED was made publicly available the first full CEED software distribution, release 469CEED 1.0. The distribution was made available using the Spack package manager to provide 470a common, easy-to-use build environment, where the user can build the CEED distribution 471with all dependencies. This release included a new Fortran interface for the library. 472This release also contained major improvements in the OCCA backend (including a new 473`/ocl/occa` backend) and new examples. The standalone libCEED example was modified to 474compute the volume volume of a given mesh (in 1D, 2D, or 3D) and placed in an 475`examples/ceed` subfolder. A new `mfem` example to perform BP3 (with the application 476of the Laplace operator) was also added to this release. 477 478Backends available in this release: 479 480| CEED resource (`-ceed`) | Backend | 481|-------------------------|---------------------------------| 482| `/cpu/self` | Serial reference implementation | 483| `/cpu/occa` | Serial OCCA kernels | 484| `/gpu/occa` | CUDA OCCA kernels | 485| `/omp/occa` | OpenMP OCCA kernels | 486| `/ocl/occa` | OpenCL OCCA kernels | 487 488Examples available in this release: 489 490:::{list-table} 491:header-rows: 1 492:widths: auto 493* - User code 494 - Example 495* - `ceed` 496 - * ex1 (volume) 497* - `mfem` 498 - * BP1 (scalar mass operator) 499 * BP3 (scalar Laplace operator) 500* - `petsc` 501 - * BP1 (scalar mass operator) 502* - `nek5000` 503 - * BP1 (scalar mass operator) 504::: 505 506(v0-1)= 507 508## v0.1 (Jan 3, 2018) 509 510Initial low-level API of the CEED project. The low-level API provides a set of Finite 511Elements kernels and components for writing new low-level kernels. Examples include: 512vector and sparse linear algebra, element matrix assembly over a batch of elements, 513partial assembly and action for efficient high-order operators like mass, diffusion, 514advection, etc. The main goal of the low-level API is to establish the basis for the 515high-level API. Also, identifying such low-level kernels and providing a reference 516implementation for them serves as the basis for specialized backend implementations. 517This release contained several backends: `/cpu/self`, and backends which rely upon the 518[OCCA](http://github.com/libocca/occa) package, such as `/cpu/occa`, 519`/gpu/occa`, and `/omp/occa`. 520It also included several examples, in the `examples` folder: 521A standalone code that shows the usage of libCEED (with no external 522dependencies) to apply the Laplace operator, `ex1`; an `mfem` example to perform BP1 523(with the application of the mass operator); and a `petsc` example to perform BP1 524(with the application of the mass operator). 525 526Backends available in this release: 527 528| CEED resource (`-ceed`) | Backend | 529|-------------------------|---------------------------------| 530| `/cpu/self` | Serial reference implementation | 531| `/cpu/occa` | Serial OCCA kernels | 532| `/gpu/occa` | CUDA OCCA kernels | 533| `/omp/occa` | OpenMP OCCA kernels | 534 535Examples available in this release: 536 537| User code | Example | 538|-----------------------|-----------------------------------| 539| `ceed` | ex1 (scalar Laplace operator) | 540| `mfem` | BP1 (scalar mass operator) | 541| `petsc` | BP1 (scalar mass operator) | 542``` 543