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