xref: /libCEED/doc/sphinx/source/releasenotes.md (revision 7b63f5c6881a9a0bb827bb6972a33367d9223442)
1bcb2dfaeSJed Brown# Changes/Release Notes
2bcb2dfaeSJed Brown
3f374d6a3SJeremy L ThompsonOn this page we provide a summary of the main API changes, new features and examples for each release of libCEED.
4bcb2dfaeSJed Brown
5bcb2dfaeSJed Brown(main)=
6bcb2dfaeSJed Brown
7bcb2dfaeSJed Brown## Current `main` branch
8bcb2dfaeSJed Brown
97e7773b5SJeremy L Thompson### Interface changes
107e7773b5SJeremy L Thompson
11ea6b5821SJeremy L Thompson- Added {c:func}`CeedOperatorSetName` for more readable {c:func}`CeedOperatorView` output.
12ea6b5821SJeremy L Thompson
1344d7a66cSJeremy L Thompson### Bugfix
1444d7a66cSJeremy L Thompson
1544d7a66cSJeremy L Thompson- Fix storing of indices for `CeedElemRestriction` on the host with GPU backends.
16*7b63f5c6SJed Brown- Fix `CeedElemRestriction` sizing for {c:func}`CeedOperatorAssemblePointBlockDiagonal`.
1744d7a66cSJeremy L Thompson
18e0e35436SJeremy L Thompson### Examples
19e0e35436SJeremy L Thompson
20e0e35436SJeremy L Thompson- Added various performance enhancements for {ref}`example-petsc-navier-stokes`
21e0e35436SJeremy L Thompson
22f374d6a3SJeremy L Thompson(v0-10-1)=
23f374d6a3SJeremy L Thompson
24f374d6a3SJeremy L Thompson## v0.10.1 (Apr 11, 2022)
25f374d6a3SJeremy L Thompson
26f374d6a3SJeremy L Thompson### Interface changes
27f374d6a3SJeremy L Thompson
286e15d496SJeremy L Thompson- Added {c:func}`CeedQFunctionSetUserFlopsEstimate` and {c:func}`CeedOperatorGetFlopsEstimate` to facilitate estimating FLOPs in operator application.
296e15d496SJeremy L Thompson
305766aa57SJeremy L Thompson### Bugfix
315766aa57SJeremy L Thompson
325766aa57SJeremy L Thompson- Install JiT source files in install directory to fix GPU functionality for installed libCEED.
335766aa57SJeremy L Thompson
34667e613fSJeremy L Thompson(v0-10)=
35667e613fSJeremy L Thompson
363ed90579SJeremy L Thompson## v0.10 (Mar 21, 2022)
37667e613fSJeremy L Thompson
38667e613fSJeremy L Thompson### Interface changes
39667e613fSJeremy L Thompson
407e7773b5SJeremy L Thompson- Update {c:func}`CeedQFunctionGetFields` and {c:func}`CeedOperatorGetFields` to include number of fields.
41ce4822f6SJeremy L Thompson- 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`.
42f04ea552SJeremy L Thompson- Clarify and document conditions where `CeedQFunction` and `CeedOperator` become immutable and no further fields or suboperators can be added.
4370a7ffb3SJeremy L Thompson- Add {c:func}`CeedOperatorLinearAssembleQFunctionBuildOrUpdate` to reduce object creation overhead in assembly of CeedOperator preconditioning ingredients.
444db537f9SJeremy L Thompson- Promote {c:func}`CeedOperatorCheckReady`to the public API to facilitate interactive interfaces.
45dcc1e3ecSJeremy L Thompson- Warning added when compiling OCCA backend to alert users that this backend is experimental.
469a1d3511SJeremy L Thompson- `ceed-backend.h`, `ceed-hash.h`, and `ceed-khash.h` removed. Users should use `ceed/backend.h`, `ceed/hash.h`, and `ceed/khash.h`.
4743e1b16fSJeremy L Thompson- Added {c:func}`CeedQFunctionGetKernelName`; refactored {c:func}`CeedQFunctionGetSourcePath` to exclude function kernel name.
489c774eddSJeremy L Thompson- Clarify documentation for {c:func}`CeedVectorTakeArray`; this function will error if {c:func}`CeedVectorSetArray` with `copy_mode == CEED_USE_POINTER` was not previously called for the corresponding `CeedMemType`.
499c774eddSJeremy L Thompson- Added {c:func}`CeedVectorGetArrayWrite` that allows access to uninitalized arrays; require initalized data for {c:func}`CeedVectorGetArray`.
50c38440baSJed Brown- Added {c:func}`CeedQFunctionContextRegisterDouble` and {c:func}`CeedQFunctionContextRegisterInt32` with {c:func}`CeedQFunctionContextSetDouble` and {c:func}`CeedQFunctionContextSetInt32` to facilitate easy updating of {c:struct}`CeedQFunctionContext` data by user defined field names.
51cdf32b93SJeremy L Thompson- Added {c:func}`CeedQFunctionContextGetFieldDescriptions` to retreive user defined descriptions of fields that are registered with `CeedQFunctionContextRegister*`.
527a06ec9fSJeremy L Thompson- Renamed `CeedElemTopology` entries for clearer namespacing between libCEED enums.
53f4f98f9dSJeremy L Thompson- Added type `CeedSize` equivalent to `ptrdiff_t` for array sizes in {c:func}`CeedVectorCreate`, {c:func}`CeedVectorGetLength`, `CeedElemRestrictionCreate*`, {c:func}`CeedElemRestrictionGetLVectorSize`, and {c:func}`CeedOperatorLinearAssembleSymbolic`. This is a breaking change.
548b919e6bSJeremy L Thompson- Added {c:func}`CeedOperatorSetQFunctionUpdated` to facilitate QFunction data re-use between operators sharing the same quadrature space, such as in a multigrid hierarchy.
55c9366a6bSJeremy L Thompson- Added {c:func}`CeedOperatorGetActiveVectorLengths` to get shape of CeedOperator.
567e7773b5SJeremy L Thompson
57f479eb23SJeremy L Thompson### New features
58f479eb23SJeremy L Thompson
59f479eb23SJeremy L Thompson- `CeedScalar` can now be set as `float` or `double` at compile time.
6030601ac0SJeremy L Thompson- Added JiT utilities in `ceed/jit-tools.h` to reduce duplicated code in GPU backends.
61fb3c7d02SJeremy L Thompson- 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.
6223dfbf5bSJeremy L Thompson- Remove need to guard library headers in QFunction source for code generation backends.
633f21f6b1SJeremy L Thompson- `CeedDebugEnv()` macro created to provide debugging outputs when Ceed context is not present.
64f7e22acaSJeremy L Thompson- Added {c:func}`CeedStringAllocCopy` to reduce repeated code for copying strings internally.
653451974fSJeremy L Thompson- Added {c:func}`CeedPathConcatenate` to facilitate loading kernel source files with a path relative to the current file.
667a06ec9fSJeremy L Thompson- Added support for non-tensor H(div) elements, to include CPU backend implementations and {c:func}`CeedBasisCreateHdiv` convenience constructor.
67d34e270fSJeremy L Thompson- Added {c:func}`CeedQFunctionSetContextWritable` and read-only access to `CeedQFunctionContext` data as an optional feature to improve GPU performance. By default, calling the `CeedQFunctionUser` during {c:func}`CeedQFunctionApply` is assumed to write into the `CeedQFunctionContext` data, consistent with the previous behavior. Note that if a user asserts that their `CeedQFunctionUser` does not write into the `CeedQFunctionContext` data, they are responsible for the validity of this assertion.
6859ad764aSnbeams- Added support for element matrix assembly in GPU backends.
69f479eb23SJeremy L Thompson
70bcb2dfaeSJed Brown### Maintainability
71bcb2dfaeSJed Brown
72bcb2dfaeSJed Brown- Refactored preconditioner support internally to facilitate future development and improve GPU completeness/test coverage.
73db52d626SJeremy L Thompson- `Include-what-you-use` makefile target added as `make iwyu`.
74bf4cb664SJeremy L Thompson- Create backend constant `CEED_FIELD_MAX` to reduce magic numbers in codebase.
753451974fSJeremy L Thompson- Put GPU JiTed kernel source code into separate files.
76f9996dfdSJeremy L Thompson- Dropped legacy version support in PETSc based examples to better utilize PETSc DMPlex and Mat updates to support libCEED; current minimum PETSc version for the examples is v3.17.
77bcb2dfaeSJed Brown
78bcb2dfaeSJed Brown(v0-9)=
79bcb2dfaeSJed Brown
80bcb2dfaeSJed Brown## v0.9 (Jul 6, 2021)
81bcb2dfaeSJed Brown
82bcb2dfaeSJed Brown### Interface changes
83bcb2dfaeSJed Brown
84bcb2dfaeSJed Brown- Minor modification in error handling macro to silence pedantic warnings when compiling with Clang, but no functional impact.
85bcb2dfaeSJed Brown
86bcb2dfaeSJed Brown### New features
87bcb2dfaeSJed Brown
88bcb2dfaeSJed Brown- Add {c:func}`CeedVectorAXPY` and {c:func}`CeedVectorPointwiseMult` as a convenience for stand-alone testing and internal use.
89bcb2dfaeSJed Brown- Add `CEED_QFUNCTION_HELPER` macro to properly annotate QFunction helper functions for code generation backends.
90bcb2dfaeSJed Brown- Add `CeedPragmaOptimizeOff` macro for code that is sensitive to floating point errors from fast math optimizations.
91bcb2dfaeSJed Brown- Rust support: split `libceed-sys` crate out of `libceed` and [publish both on crates.io](https://crates.io/crates/libceed).
92bcb2dfaeSJed Brown
93bcb2dfaeSJed Brown### Performance improvements
94bcb2dfaeSJed Brown
95bcb2dfaeSJed Brown### Examples
96bcb2dfaeSJed Brown
97bcb2dfaeSJed Brown- Solid mechanics mini-app updated to explore the performance impacts of various formulations in the initial and current configurations.
98bcb2dfaeSJed Brown- Fluid mechanics example adds GPU support and improves modularity.
99bcb2dfaeSJed Brown
100bcb2dfaeSJed Brown### Deprecated backends
101bcb2dfaeSJed Brown
102bcb2dfaeSJed Brown- 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.
103bcb2dfaeSJed Brown
104bcb2dfaeSJed Brown(v0-8)=
105bcb2dfaeSJed Brown
106bcb2dfaeSJed Brown## v0.8 (Mar 31, 2021)
107bcb2dfaeSJed Brown
108bcb2dfaeSJed Brown### Interface changes
109bcb2dfaeSJed Brown
110bcb2dfaeSJed Brown- Error handling improved to include enumerated error codes for C interface return values.
111bcb2dfaeSJed Brown- 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.
112bcb2dfaeSJed Brown
113bcb2dfaeSJed Brown### New features
114bcb2dfaeSJed Brown
115bcb2dfaeSJed Brown- Julia and Rust interfaces added, providing a nearly 1-1 correspondence with the C interface, plus some convenience features.
116bcb2dfaeSJed Brown- Static libraries can be built with `make STATIC=1` and the pkg-config file is installed accordingly.
117bcb2dfaeSJed Brown- Add {c:func}`CeedOperatorLinearAssembleSymbolic` and {c:func}`CeedOperatorLinearAssemble` to support full assembly of libCEED operators.
118bcb2dfaeSJed Brown
119bcb2dfaeSJed Brown### Performance improvements
120bcb2dfaeSJed Brown
121bcb2dfaeSJed Brown- New HIP MAGMA backends for hipMAGMA library users: `/gpu/hip/magma` and `/gpu/hip/magma/det`.
122bcb2dfaeSJed Brown- New HIP backends for improved tensor basis performance: `/gpu/hip/shared` and `/gpu/hip/gen`.
123bcb2dfaeSJed Brown
124bcb2dfaeSJed Brown### Examples
125bcb2dfaeSJed Brown
126bcb2dfaeSJed Brown- {ref}`example-petsc-elasticity` example updated with traction boundary conditions and improved Dirichlet boundary conditions.
127bcb2dfaeSJed Brown- {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.
128bcb2dfaeSJed Brown- {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.
129bcb2dfaeSJed Brown- {ref}`example-petsc-navier-stokes` example updated with support for performing convergence study and plotting order of convergence by polynomial degree.
130bcb2dfaeSJed Brown
131bcb2dfaeSJed Brown(v0-7)=
132bcb2dfaeSJed Brown
133bcb2dfaeSJed Brown## v0.7 (Sep 29, 2020)
134bcb2dfaeSJed Brown
135bcb2dfaeSJed Brown### Interface changes
136bcb2dfaeSJed Brown
137bcb2dfaeSJed Brown- Replace limited {code}`CeedInterlaceMode` with more flexible component stride {code}`compstride` in {code}`CeedElemRestriction` constructors.
138bcb2dfaeSJed Brown  As a result, the {code}`indices` parameter has been replaced with {code}`offsets` and the {code}`nnodes` parameter has been replaced with {code}`lsize`.
139bcb2dfaeSJed Brown  These changes improve support for mixed finite element methods.
140bcb2dfaeSJed Brown- Replace various uses of {code}`Ceed*Get*Status` with {code}`Ceed*Is*` in the backend API to match common nomenclature.
141bcb2dfaeSJed Brown- Replace {code}`CeedOperatorAssembleLinearDiagonal` with {c:func}`CeedOperatorLinearAssembleDiagonal` for clarity.
142bcb2dfaeSJed Brown- 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.
143bcb2dfaeSJed Brown- 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.
144bcb2dfaeSJed Brown- 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.
145bcb2dfaeSJed Brown- 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.
146bcb2dfaeSJed Brown  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`.
147bcb2dfaeSJed Brown- Added {code}`CeedQFunctionContext` object to manage user QFunction context data and reduce copies between device and host memory.
148bcb2dfaeSJed Brown- 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.
149bcb2dfaeSJed Brown
150bcb2dfaeSJed Brown### New features
151bcb2dfaeSJed Brown
152bcb2dfaeSJed Brown- New HIP backend: `/gpu/hip/ref`.
153bcb2dfaeSJed Brown- CeedQFunction support for user `CUfunction`s in some backends
154bcb2dfaeSJed Brown
155bcb2dfaeSJed Brown### Performance improvements
156bcb2dfaeSJed Brown
157bcb2dfaeSJed Brown- OCCA backend rebuilt to facilitate future performance enhancements.
158bcb2dfaeSJed Brown- Petsc BPs suite improved to reduce noise due to multiple calls to {code}`mpiexec`.
159bcb2dfaeSJed Brown
160bcb2dfaeSJed Brown### Examples
161bcb2dfaeSJed Brown
162bcb2dfaeSJed Brown- {ref}`example-petsc-elasticity` example updated with strain energy computation and more flexible boundary conditions.
163bcb2dfaeSJed Brown
164bcb2dfaeSJed Brown### Deprecated backends
165bcb2dfaeSJed Brown
166bcb2dfaeSJed Brown- The `/gpu/cuda/reg` backend has been removed, with its core features moved into `/gpu/cuda/ref` and `/gpu/cuda/shared`.
167bcb2dfaeSJed Brown
168bcb2dfaeSJed Brown(v0-6)=
169bcb2dfaeSJed Brown
170bcb2dfaeSJed Brown## v0.6 (Mar 29, 2020)
171bcb2dfaeSJed Brown
172bcb2dfaeSJed BrownlibCEED v0.6 contains numerous new features and examples, as well as expanded
17313964f07SJed Browndocumentation in [this new website](https://libceed.org).
174bcb2dfaeSJed Brown
175bcb2dfaeSJed Brown### New features
176bcb2dfaeSJed Brown
177bcb2dfaeSJed Brown- New Python interface using [CFFI](https://cffi.readthedocs.io/) provides a nearly
178bcb2dfaeSJed Brown  1-1 correspondence with the C interface, plus some convenience features.  For instance,
179bcb2dfaeSJed Brown  data stored in the {cpp:type}`CeedVector` structure are available without copy as
180bcb2dfaeSJed Brown  {py:class}`numpy.ndarray`.  Short tutorials are provided in
181bcb2dfaeSJed Brown  [Binder](https://mybinder.org/v2/gh/CEED/libCEED/main?urlpath=lab/tree/examples/tutorials/).
182bcb2dfaeSJed Brown- Linear QFunctions can be assembled as block-diagonal matrices (per quadrature point,
183bcb2dfaeSJed Brown  {c:func}`CeedOperatorAssembleLinearQFunction`) or to evaluate the diagonal
184bcb2dfaeSJed Brown  ({c:func}`CeedOperatorAssembleLinearDiagonal`).  These operations are useful for
185bcb2dfaeSJed Brown  preconditioning ingredients and are used in the libCEED's multigrid examples.
186bcb2dfaeSJed Brown- The inverse of separable operators can be obtained using
187bcb2dfaeSJed Brown  {c:func}`CeedOperatorCreateFDMElementInverse` and applied with
188bcb2dfaeSJed Brown  {c:func}`CeedOperatorApply`.  This is a useful preconditioning ingredient,
189bcb2dfaeSJed Brown  especially for Laplacians and related operators.
190bcb2dfaeSJed Brown- New functions: {c:func}`CeedVectorNorm`, {c:func}`CeedOperatorApplyAdd`,
191bcb2dfaeSJed Brown  {c:func}`CeedQFunctionView`, {c:func}`CeedOperatorView`.
192bcb2dfaeSJed Brown- Make public accessors for various attributes to facilitate writing composable code.
193bcb2dfaeSJed Brown- New backend: `/cpu/self/memcheck/serial`.
194bcb2dfaeSJed Brown- QFunctions using variable-length array (VLA) pointer constructs can be used with CUDA
195bcb2dfaeSJed Brown  backends.  (Single source is coming soon for OCCA backends.)
196bcb2dfaeSJed Brown- Fix some missing edge cases in CUDA backend.
197bcb2dfaeSJed Brown
198bcb2dfaeSJed Brown### Performance Improvements
199bcb2dfaeSJed Brown
200bcb2dfaeSJed Brown- MAGMA backend performance optimization and non-tensor bases.
201bcb2dfaeSJed Brown- No-copy optimization in {c:func}`CeedOperatorApply`.
202bcb2dfaeSJed Brown
203bcb2dfaeSJed Brown### Interface changes
204bcb2dfaeSJed Brown
205bcb2dfaeSJed Brown- Replace {code}`CeedElemRestrictionCreateIdentity` and
206bcb2dfaeSJed Brown  {code}`CeedElemRestrictionCreateBlocked` with more flexible
207bcb2dfaeSJed Brown  {c:func}`CeedElemRestrictionCreateStrided` and
208bcb2dfaeSJed Brown  {c:func}`CeedElemRestrictionCreateBlockedStrided`.
209bcb2dfaeSJed Brown- Add arguments to {c:func}`CeedQFunctionCreateIdentity`.
210bcb2dfaeSJed Brown- Replace ambiguous uses of {cpp:enum}`CeedTransposeMode` for L-vector identification
211bcb2dfaeSJed Brown  with {cpp:enum}`CeedInterlaceMode`.  This is now an attribute of the
212bcb2dfaeSJed Brown  {cpp:type}`CeedElemRestriction` (see {c:func}`CeedElemRestrictionCreate`) and no
213bcb2dfaeSJed Brown  longer passed as `lmode` arguments to {c:func}`CeedOperatorSetField` and
214bcb2dfaeSJed Brown  {c:func}`CeedElemRestrictionApply`.
215bcb2dfaeSJed Brown
216bcb2dfaeSJed Brown### Examples
217bcb2dfaeSJed Brown
218bcb2dfaeSJed BrownlibCEED-0.6 contains greatly expanded examples with {ref}`new documentation <Examples>`.
219bcb2dfaeSJed BrownNotable additions include:
220bcb2dfaeSJed Brown
221bcb2dfaeSJed Brown- Standalone {ref}`ex2-surface` ({file}`examples/ceed/ex2-surface`): compute the area of
222bcb2dfaeSJed Brown  a domain in 1, 2, and 3 dimensions by applying a Laplacian.
223bcb2dfaeSJed Brown
224bcb2dfaeSJed Brown- PETSc {ref}`example-petsc-area` ({file}`examples/petsc/area.c`): computes surface area
225bcb2dfaeSJed Brown  of domains (like the cube and sphere) by direct integration on a surface mesh;
226bcb2dfaeSJed Brown  demonstrates geometric dimension different from topological dimension.
227bcb2dfaeSJed Brown
228bcb2dfaeSJed Brown- PETSc {ref}`example-petsc-bps`:
229bcb2dfaeSJed Brown
230bcb2dfaeSJed Brown  - {file}`examples/petsc/bpsraw.c` (formerly `bps.c`): transparent CUDA support.
231bcb2dfaeSJed Brown  - {file}`examples/petsc/bps.c` (formerly `bpsdmplex.c`): performance improvements
232bcb2dfaeSJed Brown    and transparent CUDA support.
233bcb2dfaeSJed Brown  - {ref}`example-petsc-bps-sphere` ({file}`examples/petsc/bpssphere.c`):
234bcb2dfaeSJed Brown    generalizations of all CEED BPs to the surface of the sphere; demonstrates geometric
235bcb2dfaeSJed Brown    dimension different from topological dimension.
236bcb2dfaeSJed Brown
237bcb2dfaeSJed Brown- {ref}`example-petsc-multigrid` ({file}`examples/petsc/multigrid.c`): new p-multigrid
238bcb2dfaeSJed Brown  solver with algebraic multigrid coarse solve.
239bcb2dfaeSJed Brown
240bcb2dfaeSJed Brown- {ref}`example-petsc-navier-stokes` ({file}`examples/fluids/navierstokes.c`; formerly
241bcb2dfaeSJed Brown  `examples/navier-stokes`): unstructured grid support (using PETSc's `DMPlex`),
242bcb2dfaeSJed Brown  implicit time integration, SU/SUPG stabilization, free-slip boundary conditions, and
243bcb2dfaeSJed Brown  quasi-2D computational domain support.
244bcb2dfaeSJed Brown
245bcb2dfaeSJed Brown- {ref}`example-petsc-elasticity` ({file}`examples/solids/elasticity.c`): new solver for
246bcb2dfaeSJed Brown  linear elasticity, small-strain hyperelasticity, and globalized finite-strain
247bcb2dfaeSJed Brown  hyperelasticity using p-multigrid with algebraic multigrid coarse solve.
248bcb2dfaeSJed Brown
249bcb2dfaeSJed Brown(v0-5)=
250bcb2dfaeSJed Brown
251bcb2dfaeSJed Brown## v0.5 (Sep 18, 2019)
252bcb2dfaeSJed Brown
253bcb2dfaeSJed BrownFor this release, several improvements were made. Two new CUDA backends were added to
254bcb2dfaeSJed Brownthe family of backends, of which, the new `cuda-gen` backend achieves state-of-the-art
255bcb2dfaeSJed Brownperformance using single-source {ref}`CeedQFunction`. From this release, users
256bcb2dfaeSJed Browncan define Q-Functions in a single source code independently of the targeted backend
257bcb2dfaeSJed Brownwith the aid of a new macro `CEED QFUNCTION` to support JIT (Just-In-Time) and CPU
258bcb2dfaeSJed Browncompilation of the user provided {ref}`CeedQFunction` code. To allow a unified
259bcb2dfaeSJed Browndeclaration, the {ref}`CeedQFunction` API has undergone a slight change:
260bcb2dfaeSJed Brownthe `QFunctionField` parameter `ncomp` has been changed to `size`. This change
261bcb2dfaeSJed Brownrequires setting the previous value of `ncomp` to `ncomp*dim` when adding a
262bcb2dfaeSJed Brown`QFunctionField` with eval mode `CEED EVAL GRAD`.
263bcb2dfaeSJed Brown
264bcb2dfaeSJed BrownAdditionally, new CPU backends
265bcb2dfaeSJed Brownwere included in this release, such as the `/cpu/self/opt/*` backends (which are
266bcb2dfaeSJed Brownwritten in pure C and use partial **E-vectors** to improve performance) and the
267bcb2dfaeSJed Brown`/cpu/self/ref/memcheck` backend (which relies upon the
268bcb2dfaeSJed Brown[Valgrind](http://valgrind.org/) Memcheck tool to help verify that user
269bcb2dfaeSJed Brown{ref}`CeedQFunction` have no undefined values).
270bcb2dfaeSJed BrownThis release also included various performance improvements, bug fixes, new examples,
271bcb2dfaeSJed Brownand improved tests. Among these improvements, vectorized instructions for
272bcb2dfaeSJed Brown{ref}`CeedQFunction` code compiled for CPU were enhanced by using `CeedPragmaSIMD`
273bcb2dfaeSJed Browninstead of `CeedPragmaOMP`, implementation of a {ref}`CeedQFunction` gallery and
274bcb2dfaeSJed Brownidentity Q-Functions were introduced, and the PETSc benchmark problems were expanded
275bcb2dfaeSJed Brownto include unstructured meshes handling were. For this expansion, the prior version of
276bcb2dfaeSJed Brownthe PETSc BPs, which only included data associated with structured geometries, were
277bcb2dfaeSJed Brownrenamed `bpsraw`, and the new version of the BPs, which can handle data associated
278bcb2dfaeSJed Brownwith any unstructured geometry, were called `bps`. Additionally, other benchmark
279bcb2dfaeSJed Brownproblems, namely BP2 and BP4 (the vector-valued versions of BP1 and BP3, respectively),
280bcb2dfaeSJed Brownand BP5 and BP6 (the collocated versions---for which the quadrature points are the same
281bcb2dfaeSJed Brownas the Gauss Lobatto nodes---of BP3 and BP4 respectively) were added to the PETSc
282bcb2dfaeSJed Brownexamples. Furthermoew, another standalone libCEED example, called `ex2`, which
283bcb2dfaeSJed Browncomputes the surface area of a given mesh was added to this release.
284bcb2dfaeSJed Brown
285bcb2dfaeSJed BrownBackends available in this release:
286bcb2dfaeSJed Brown
28768e843eeSJed Brown| CEED resource (`-ceed`)  | Backend                                             |
28868e843eeSJed Brown|--------------------------|-----------------------------------------------------|
28968e843eeSJed Brown| `/cpu/self/ref/serial`   | Serial reference implementation                     |
29068e843eeSJed Brown| `/cpu/self/ref/blocked`  | Blocked reference implementation                    |
29168e843eeSJed Brown| `/cpu/self/ref/memcheck` | Memcheck backend, undefined value checks            |
29268e843eeSJed Brown| `/cpu/self/opt/serial`   | Serial optimized C implementation                   |
29368e843eeSJed Brown| `/cpu/self/opt/blocked`  | Blocked optimized C implementation                  |
29468e843eeSJed Brown| `/cpu/self/avx/serial`   | Serial AVX implementation                           |
29568e843eeSJed Brown| `/cpu/self/avx/blocked`  | Blocked AVX implementation                          |
29668e843eeSJed Brown| `/cpu/self/xsmm/serial`  | Serial LIBXSMM implementation                       |
29768e843eeSJed Brown| `/cpu/self/xsmm/blocked` | Blocked LIBXSMM implementation                      |
29868e843eeSJed Brown| `/cpu/occa`              | Serial OCCA kernels                                 |
29968e843eeSJed Brown| `/gpu/occa`              | CUDA OCCA kernels                                   |
30068e843eeSJed Brown| `/omp/occa`              | OpenMP OCCA kernels                                 |
30168e843eeSJed Brown| `/ocl/occa`              | OpenCL OCCA kernels                                 |
30268e843eeSJed Brown| `/gpu/cuda/ref`          | Reference pure CUDA kernels                         |
30368e843eeSJed Brown| `/gpu/cuda/reg`          | Pure CUDA kernels using one thread per element      |
30468e843eeSJed Brown| `/gpu/cuda/shared`       | Optimized pure CUDA kernels using shared memory     |
30568e843eeSJed Brown| `/gpu/cuda/gen`          | Optimized pure CUDA kernels using code generation   |
30668e843eeSJed Brown| `/gpu/magma`             | CUDA MAGMA kernels                                  |
307bcb2dfaeSJed Brown
308bcb2dfaeSJed BrownExamples available in this release:
309bcb2dfaeSJed Brown
31068e843eeSJed Brown:::{list-table}
31168e843eeSJed Brown:header-rows: 1
31268e843eeSJed Brown:widths: auto
31368e843eeSJed Brown* - User code
31468e843eeSJed Brown  - Example
31568e843eeSJed Brown* - `ceed`
31668e843eeSJed Brown  - * ex1 (volume)
31768e843eeSJed Brown    * ex2 (surface)
31868e843eeSJed Brown* - `mfem`
31968e843eeSJed Brown  - * BP1 (scalar mass operator)
32068e843eeSJed Brown    * BP3 (scalar Laplace operator)
32168e843eeSJed Brown* - `petsc`
32268e843eeSJed Brown  - * BP1 (scalar mass operator)
32368e843eeSJed Brown    * BP2 (vector mass operator)
32468e843eeSJed Brown    * BP3 (scalar Laplace operator)
32568e843eeSJed Brown    * BP4 (vector Laplace operator)
32668e843eeSJed Brown    * BP5 (collocated scalar Laplace operator)
32768e843eeSJed Brown    * BP6 (collocated vector Laplace operator)
32868e843eeSJed Brown    * Navier-Stokes
32968e843eeSJed Brown* - `nek5000`
33068e843eeSJed Brown  - * BP1 (scalar mass operator)
33168e843eeSJed Brown    * BP3 (scalar Laplace operator)
33268e843eeSJed Brown:::
333bcb2dfaeSJed Brown
334bcb2dfaeSJed Brown(v0-4)=
335bcb2dfaeSJed Brown
336bcb2dfaeSJed Brown## v0.4 (Apr 1, 2019)
337bcb2dfaeSJed Brown
338bcb2dfaeSJed BrownlibCEED v0.4 was made again publicly available in the second full CEED software
339bcb2dfaeSJed Browndistribution, release CEED 2.0. This release contained notable features, such as
340bcb2dfaeSJed Brownfour new CPU backends, two new GPU backends, CPU backend optimizations, initial
341bcb2dfaeSJed Brownsupport for operator composition, performance benchmarking, and a Navier-Stokes demo.
342bcb2dfaeSJed BrownThe new CPU backends in this release came in two families. The `/cpu/self/*/serial`
343bcb2dfaeSJed Brownbackends process one element at a time and are intended for meshes with a smaller number
344bcb2dfaeSJed Brownof high order elements. The `/cpu/self/*/blocked` backends process blocked batches of
345bcb2dfaeSJed Browneight interlaced elements and are intended for meshes with higher numbers of elements.
346bcb2dfaeSJed BrownThe `/cpu/self/avx/*` backends rely upon AVX instructions to provide vectorized CPU
347bcb2dfaeSJed Brownperformance. The `/cpu/self/xsmm/*` backends rely upon the
348bcb2dfaeSJed Brown[LIBXSMM](http://github.com/hfp/libxsmm) package to provide vectorized CPU
349bcb2dfaeSJed Brownperformance. The `/gpu/cuda/*` backends provide GPU performance strictly using CUDA.
350bcb2dfaeSJed BrownThe `/gpu/cuda/ref` backend is a reference CUDA backend, providing reasonable
351bcb2dfaeSJed Brownperformance for most problem configurations. The `/gpu/cuda/reg` backend uses a simple
352bcb2dfaeSJed Brownparallelization approach, where each thread treats a finite element. Using just in time
353bcb2dfaeSJed Browncompilation, provided by nvrtc (NVidia Runtime Compiler), and runtime parameters, this
354bcb2dfaeSJed Brownbackend unroll loops and map memory address to registers. The `/gpu/cuda/reg` backend
355bcb2dfaeSJed Brownachieve good peak performance for 1D, 2D, and low order 3D problems, but performance
356bcb2dfaeSJed Browndeteriorates very quickly when threads run out of registers.
357bcb2dfaeSJed Brown
358bcb2dfaeSJed BrownA new explicit time-stepping Navier-Stokes solver was added to the family of libCEED
359bcb2dfaeSJed Brownexamples in the `examples/petsc` directory (see {ref}`example-petsc-navier-stokes`).
360bcb2dfaeSJed BrownThis example solves the time-dependent Navier-Stokes equations of compressible gas
361bcb2dfaeSJed Browndynamics in a static Eulerian three-dimensional frame, using structured high-order
362bcb2dfaeSJed Brownfinite/spectral element spatial discretizations and explicit high-order time-stepping
363bcb2dfaeSJed Brown(available in PETSc). Moreover, the Navier-Stokes example was developed using PETSc,
364bcb2dfaeSJed Brownso that the pointwise physics (defined at quadrature points) is separated from the
365bcb2dfaeSJed Brownparallelization and meshing concerns.
366bcb2dfaeSJed Brown
367bcb2dfaeSJed BrownBackends available in this release:
368bcb2dfaeSJed Brown
36968e843eeSJed Brown| CEED resource (`-ceed`)  | Backend                                             |
37068e843eeSJed Brown|--------------------------|-----------------------------------------------------|
37168e843eeSJed Brown| `/cpu/self/ref/serial`   | Serial reference implementation                     |
37268e843eeSJed Brown| `/cpu/self/ref/blocked`  | Blocked reference implementation                    |
37368e843eeSJed Brown| `/cpu/self/tmpl`         | Backend template, defaults to `/cpu/self/blocked`   |
37468e843eeSJed Brown| `/cpu/self/avx/serial`   | Serial AVX implementation                           |
37568e843eeSJed Brown| `/cpu/self/avx/blocked`  | Blocked AVX implementation                          |
37668e843eeSJed Brown| `/cpu/self/xsmm/serial`  | Serial LIBXSMM implementation                       |
37768e843eeSJed Brown| `/cpu/self/xsmm/blocked` | Blocked LIBXSMM implementation                      |
37868e843eeSJed Brown| `/cpu/occa`              | Serial OCCA kernels                                 |
37968e843eeSJed Brown| `/gpu/occa`              | CUDA OCCA kernels                                   |
38068e843eeSJed Brown| `/omp/occa`              | OpenMP OCCA kernels                                 |
38168e843eeSJed Brown| `/ocl/occa`              | OpenCL OCCA kernels                                 |
38268e843eeSJed Brown| `/gpu/cuda/ref`          | Reference pure CUDA kernels                         |
38368e843eeSJed Brown| `/gpu/cuda/reg`          | Pure CUDA kernels using one thread per element      |
38468e843eeSJed Brown| `/gpu/magma`             | CUDA MAGMA kernels                                  |
385bcb2dfaeSJed Brown
386bcb2dfaeSJed BrownExamples available in this release:
387bcb2dfaeSJed Brown
38868e843eeSJed Brown:::{list-table}
38968e843eeSJed Brown:header-rows: 1
39068e843eeSJed Brown:widths: auto
39168e843eeSJed Brown* - User code
39268e843eeSJed Brown  - Example
39368e843eeSJed Brown* - `ceed`
39468e843eeSJed Brown  - * ex1 (volume)
39568e843eeSJed Brown* - `mfem`
39668e843eeSJed Brown  - * BP1 (scalar mass operator)
39768e843eeSJed Brown    * BP3 (scalar Laplace operator)
39868e843eeSJed Brown* - `petsc`
39968e843eeSJed Brown  - * BP1 (scalar mass operator)
40068e843eeSJed Brown    * BP3 (scalar Laplace operator)
40168e843eeSJed Brown    * Navier-Stokes
40268e843eeSJed Brown* - `nek5000`
40368e843eeSJed Brown  - * BP1 (scalar mass operator)
40468e843eeSJed Brown    * BP3 (scalar Laplace operator)
40568e843eeSJed Brown:::
406bcb2dfaeSJed Brown
407bcb2dfaeSJed Brown(v0-3)=
408bcb2dfaeSJed Brown
409bcb2dfaeSJed Brown## v0.3 (Sep 30, 2018)
410bcb2dfaeSJed Brown
411bcb2dfaeSJed BrownNotable features in this release include active/passive field interface, support for
412bcb2dfaeSJed Brownnon-tensor bases, backend optimization, and improved Fortran interface. This release
413bcb2dfaeSJed Brownalso focused on providing improved continuous integration, and many new tests with code
414bcb2dfaeSJed Browncoverage reports of about 90%. This release also provided a significant change to the
415bcb2dfaeSJed Brownpublic interface: a {ref}`CeedQFunction` can take any number of named input and output
416bcb2dfaeSJed Brownarguments while {ref}`CeedOperator` connects them to the actual data, which may be
417bcb2dfaeSJed Brownsupplied explicitly to `CeedOperatorApply()` (active) or separately via
418bcb2dfaeSJed Brown`CeedOperatorSetField()` (passive). This interface change enables reusable libraries
419bcb2dfaeSJed Brownof CeedQFunctions and composition of block solvers constructed using
420bcb2dfaeSJed Brown{ref}`CeedOperator`. A concept of blocked restriction was added to this release and
421bcb2dfaeSJed Brownused in an optimized CPU backend. Although this is typically not visible to the user,
422bcb2dfaeSJed Brownit enables effective use of arbitrary-length SIMD while maintaining cache locality.
423bcb2dfaeSJed BrownThis CPU backend also implements an algebraic factorization of tensor product gradients
424bcb2dfaeSJed Brownto perform fewer operations than standard application of interpolation and
425bcb2dfaeSJed Browndifferentiation from nodes to quadrature points. This algebraic formulation
426bcb2dfaeSJed Brownautomatically supports non-polynomial and non-interpolatory bases, thus is more general
427bcb2dfaeSJed Brownthan the more common derivation in terms of Lagrange polynomials on the quadrature points.
428bcb2dfaeSJed Brown
429bcb2dfaeSJed BrownBackends available in this release:
430bcb2dfaeSJed Brown
43168e843eeSJed Brown| CEED resource (`-ceed`) | Backend                                             |
43268e843eeSJed Brown|-------------------------|-----------------------------------------------------|
43368e843eeSJed Brown| `/cpu/self/blocked`     | Blocked reference implementation                    |
43468e843eeSJed Brown| `/cpu/self/ref`         | Serial reference implementation                     |
43568e843eeSJed Brown| `/cpu/self/tmpl`        | Backend template, defaults to `/cpu/self/blocked`   |
43668e843eeSJed Brown| `/cpu/occa`             | Serial OCCA kernels                                 |
43768e843eeSJed Brown| `/gpu/occa`             | CUDA OCCA kernels                                   |
43868e843eeSJed Brown| `/omp/occa`             | OpenMP OCCA kernels                                 |
43968e843eeSJed Brown| `/ocl/occa`             | OpenCL OCCA kernels                                 |
44068e843eeSJed Brown| `/gpu/magma`            | CUDA MAGMA kernels                                  |
441bcb2dfaeSJed Brown
442bcb2dfaeSJed BrownExamples available in this release:
443bcb2dfaeSJed Brown
44468e843eeSJed Brown:::{list-table}
44568e843eeSJed Brown:header-rows: 1
44668e843eeSJed Brown:widths: auto
44768e843eeSJed Brown* - User code
44868e843eeSJed Brown  - Example
44968e843eeSJed Brown* - `ceed`
45068e843eeSJed Brown  - * ex1 (volume)
45168e843eeSJed Brown* - `mfem`
45268e843eeSJed Brown  - * BP1 (scalar mass operator)
45368e843eeSJed Brown    * BP3 (scalar Laplace operator)
45468e843eeSJed Brown* - `petsc`
45568e843eeSJed Brown  - * BP1 (scalar mass operator)
45668e843eeSJed Brown    * BP3 (scalar Laplace operator)
45768e843eeSJed Brown* - `nek5000`
45868e843eeSJed Brown  - * BP1 (scalar mass operator)
45968e843eeSJed Brown    * BP3 (scalar Laplace operator)
46068e843eeSJed Brown:::
461bcb2dfaeSJed Brown
462bcb2dfaeSJed Brown(v0-21)=
463bcb2dfaeSJed Brown
464bcb2dfaeSJed Brown## v0.21 (Sep 30, 2018)
465bcb2dfaeSJed Brown
466bcb2dfaeSJed BrownA MAGMA backend (which relies upon the
467bcb2dfaeSJed Brown[MAGMA](https://bitbucket.org/icl/magma) package) was integrated in libCEED for this
468bcb2dfaeSJed Brownrelease. This initial integration set up the framework of using MAGMA and provided the
469bcb2dfaeSJed BrownlibCEED functionality through MAGMA kernels as one of libCEED’s computational backends.
470bcb2dfaeSJed BrownAs any other backend, the MAGMA backend provides extended basic data structures for
471bcb2dfaeSJed Brown{ref}`CeedVector`, {ref}`CeedElemRestriction`, and {ref}`CeedOperator`, and implements
472bcb2dfaeSJed Brownthe fundamental CEED building blocks to work with the new data structures.
473bcb2dfaeSJed BrownIn general, the MAGMA-specific data structures keep the libCEED pointers to CPU data
474bcb2dfaeSJed Brownbut also add corresponding device (e.g., GPU) pointers to the data. Coherency is handled
475bcb2dfaeSJed Browninternally, and thus seamlessly to the user, through the functions/methods that are
476bcb2dfaeSJed Brownprovided to support them.
477bcb2dfaeSJed Brown
478bcb2dfaeSJed BrownBackends available in this release:
479bcb2dfaeSJed Brown
48068e843eeSJed Brown| CEED resource (`-ceed`) | Backend                         |
48168e843eeSJed Brown|-------------------------|---------------------------------|
48268e843eeSJed Brown| `/cpu/self`             | Serial reference implementation |
48368e843eeSJed Brown| `/cpu/occa`             | Serial OCCA kernels             |
48468e843eeSJed Brown| `/gpu/occa`             | CUDA OCCA kernels               |
48568e843eeSJed Brown| `/omp/occa`             | OpenMP OCCA kernels             |
48668e843eeSJed Brown| `/ocl/occa`             | OpenCL OCCA kernels             |
48768e843eeSJed Brown| `/gpu/magma`            | CUDA MAGMA kernels              |
488bcb2dfaeSJed Brown
489bcb2dfaeSJed BrownExamples available in this release:
490bcb2dfaeSJed Brown
49168e843eeSJed Brown:::{list-table}
49268e843eeSJed Brown:header-rows: 1
49368e843eeSJed Brown:widths: auto
49468e843eeSJed Brown* - User code
49568e843eeSJed Brown  - Example
49668e843eeSJed Brown* - `ceed`
49768e843eeSJed Brown  - * ex1 (volume)
49868e843eeSJed Brown* - `mfem`
49968e843eeSJed Brown  - * BP1 (scalar mass operator)
50068e843eeSJed Brown    * BP3 (scalar Laplace operator)
50168e843eeSJed Brown* - `petsc`
50268e843eeSJed Brown  - * BP1 (scalar mass operator)
50368e843eeSJed Brown* - `nek5000`
50468e843eeSJed Brown  - * BP1 (scalar mass operator)
50568e843eeSJed Brown:::
506bcb2dfaeSJed Brown
507bcb2dfaeSJed Brown(v0-2)=
508bcb2dfaeSJed Brown
509bcb2dfaeSJed Brown## v0.2 (Mar 30, 2018)
510bcb2dfaeSJed Brown
511bcb2dfaeSJed BrownlibCEED was made publicly available the first full CEED software distribution, release
512bcb2dfaeSJed BrownCEED 1.0. The distribution was made available using the Spack package manager to provide
513bcb2dfaeSJed Browna common, easy-to-use build environment, where the user can build the CEED distribution
514bcb2dfaeSJed Brownwith all dependencies. This release included a new Fortran interface for the library.
515bcb2dfaeSJed BrownThis release also contained major improvements in the OCCA backend (including a new
516bcb2dfaeSJed Brown`/ocl/occa` backend) and new examples. The standalone libCEED example was modified to
517bcb2dfaeSJed Browncompute the volume volume of a given mesh (in 1D, 2D, or 3D) and placed in an
518bcb2dfaeSJed Brown`examples/ceed` subfolder. A new `mfem` example to perform BP3 (with the application
519bcb2dfaeSJed Brownof the Laplace operator) was also added to this release.
520bcb2dfaeSJed Brown
521bcb2dfaeSJed BrownBackends available in this release:
522bcb2dfaeSJed Brown
52368e843eeSJed Brown| CEED resource (`-ceed`) | Backend                         |
52468e843eeSJed Brown|-------------------------|---------------------------------|
52568e843eeSJed Brown| `/cpu/self`             | Serial reference implementation |
52668e843eeSJed Brown| `/cpu/occa`             | Serial OCCA kernels             |
52768e843eeSJed Brown| `/gpu/occa`             | CUDA OCCA kernels               |
52868e843eeSJed Brown| `/omp/occa`             | OpenMP OCCA kernels             |
52968e843eeSJed Brown| `/ocl/occa`             | OpenCL OCCA kernels             |
530bcb2dfaeSJed Brown
531bcb2dfaeSJed BrownExamples available in this release:
532bcb2dfaeSJed Brown
53368e843eeSJed Brown:::{list-table}
53468e843eeSJed Brown:header-rows: 1
53568e843eeSJed Brown:widths: auto
53668e843eeSJed Brown* - User code
53768e843eeSJed Brown  - Example
53868e843eeSJed Brown* - `ceed`
53968e843eeSJed Brown  - * ex1 (volume)
54068e843eeSJed Brown* - `mfem`
54168e843eeSJed Brown  - * BP1 (scalar mass operator)
54268e843eeSJed Brown    * BP3 (scalar Laplace operator)
54368e843eeSJed Brown* - `petsc`
54468e843eeSJed Brown  - * BP1 (scalar mass operator)
54568e843eeSJed Brown* - `nek5000`
54668e843eeSJed Brown  - * BP1 (scalar mass operator)
54768e843eeSJed Brown:::
548bcb2dfaeSJed Brown
549bcb2dfaeSJed Brown(v0-1)=
550bcb2dfaeSJed Brown
551bcb2dfaeSJed Brown## v0.1 (Jan 3, 2018)
552bcb2dfaeSJed Brown
553bcb2dfaeSJed BrownInitial low-level API of the CEED project. The low-level API provides a set of Finite
554bcb2dfaeSJed BrownElements kernels and components for writing new low-level kernels. Examples include:
555bcb2dfaeSJed Brownvector and sparse linear algebra, element matrix assembly over a batch of elements,
556bcb2dfaeSJed Brownpartial assembly and action for efficient high-order operators like mass, diffusion,
557bcb2dfaeSJed Brownadvection, etc. The main goal of the low-level API is to establish the basis for the
558bcb2dfaeSJed Brownhigh-level API. Also, identifying such low-level kernels and providing a reference
559bcb2dfaeSJed Brownimplementation for them serves as the basis for specialized backend implementations.
560bcb2dfaeSJed BrownThis release contained several backends: `/cpu/self`, and backends which rely upon the
561bcb2dfaeSJed Brown[OCCA](http://github.com/libocca/occa) package, such as `/cpu/occa`,
562bcb2dfaeSJed Brown`/gpu/occa`, and `/omp/occa`.
563bcb2dfaeSJed BrownIt also included several examples, in the `examples` folder:
564bcb2dfaeSJed BrownA standalone code that shows the usage of libCEED (with no external
565bcb2dfaeSJed Browndependencies) to apply the Laplace operator, `ex1`; an `mfem` example to perform BP1
566bcb2dfaeSJed Brown(with the application of the mass operator); and a `petsc` example to perform BP1
567bcb2dfaeSJed Brown(with the application of the mass operator).
568bcb2dfaeSJed Brown
569bcb2dfaeSJed BrownBackends available in this release:
570bcb2dfaeSJed Brown
57168e843eeSJed Brown| CEED resource (`-ceed`) | Backend                         |
57268e843eeSJed Brown|-------------------------|---------------------------------|
57368e843eeSJed Brown| `/cpu/self`             | Serial reference implementation |
57468e843eeSJed Brown| `/cpu/occa`             | Serial OCCA kernels             |
57568e843eeSJed Brown| `/gpu/occa`             | CUDA OCCA kernels               |
57668e843eeSJed Brown| `/omp/occa`             | OpenMP OCCA kernels             |
577bcb2dfaeSJed Brown
578bcb2dfaeSJed BrownExamples available in this release:
579bcb2dfaeSJed Brown
580bcb2dfaeSJed Brown| User code             | Example                           |
58168e843eeSJed Brown|-----------------------|-----------------------------------|
58268e843eeSJed Brown| `ceed`                | ex1 (scalar Laplace operator)     |
58368e843eeSJed Brown| `mfem`                | BP1 (scalar mass operator)        |
58468e843eeSJed Brown| `petsc`               | BP1 (scalar mass operator)        |
585bcb2dfaeSJed Brown```
586