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