xref: /libCEED/doc/sphinx/source/releasenotes.md (revision 706efda32ea467a11ddceea40dd85d82235bf3f8)
1# Changes/Release Notes
2
3On this page we provide a summary of the main API changes, new features and examples
4for each release of libCEED.
5
6(main)=
7
8## Current `main` branch
9
10### Interface changes
11
12- Update {c:func}`CeedQFunctionGetFields` and {c:func}`CeedOperatorGetFields` to include number of fields.
13- Promote to the public API: QFunction and Operator field objects, `CeedQFunctionField` and `CeedOperatorField`, and associated getters, {c:func}`CeedQFunctionGetFields`; {c:func}`CeedQFunctionFieldGetName`; {c:func}`CeedQFunctionFieldGetSize`; {c:func}`CeedQFunctionFieldGetEvalMode`; {c:func}`CeedOperatorGetFields`; {c:func}`CeedOperatorFieldGetElemRestriction`; {c:func}`CeedOperatorFieldGetBasis`; and {c:func}`CeedOperatorFieldGetVector`.
14- Clarify and document conditions where `CeedQFunction` and `CeedOperator` become immutable and no further fields or suboperators can be added.
15- Add {c:func}`CeedOperatorLinearAssembleQFunctionBuildOrUpdate` to reduce object creation overhead in assembly of CeedOperator preconditioning ingredients.
16- Promote {c:func}`CeedOperatorCheckReady`to the public API to facilitate interactive interfaces.
17- Warning added when compiling OCCA backend to alert users that this backend is experimental.
18- `ceed-backend.h`, `ceed-hash.h`, and `ceed-khash.h` removed. Users should use `ceed/backend.h`, `ceed/hash.h`, and `ceed/khash.h`.
19- Added {c:func}`CeedQFunctionGetKernelName`; refactored {c:func}`CeedQFunctionGetSourcePath` to exclude function kernel name.
20- 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`.
21- Added {c:func}`CeedVectorGetArrayWrite` that allows access to uninitalized arrays; require initalized data for {c:func}`CeedVectorGetArray`.
22- 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.
23- Added {c:func}`CeedQFunctionContextGetFieldDescriptions` to retreive user defined descriptions of fields that are registered with `CeedQFunctionContextRegister*`.
24
25### New features
26
27- `CeedScalar` can now be set as `float` or `double` at compile time.
28- Added JiT utilities in `ceed/jit-tools.h` to reduce duplicated code in GPU backends.
29- 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.
30- Remove need to guard library headers in QFunction source for code generation backends.
31- `CeedDebugEnv()` macro created to provide debugging outputs when Ceed context is not present.
32- Added {c:func}`CeedStringAllocCopy` to reduce repeated code for copying strings internally.
33- Added {c:func}`CeedPathConcatenate` to facilitate loading kernel source files with a path relative to the current file.
34
35### Maintainability
36
37- Refactored preconditioner support internally to facilitate future development and improve GPU completeness/test coverage.
38- `Include-what-you-use` makefile target added as `make iwyu`.
39- Create backend constant `CEED_FIELD_MAX` to reduce magic numbers in codebase.
40- Put GPU JiTed kernel source code into separate files.
41
42(v0-9)=
43
44## v0.9 (Jul 6, 2021)
45
46### Interface changes
47
48- Minor modification in error handling macro to silence pedantic warnings when compiling with Clang, but no functional impact.
49
50### New features
51
52- Add {c:func}`CeedVectorAXPY` and {c:func}`CeedVectorPointwiseMult` as a convenience for stand-alone testing and internal use.
53- Add `CEED_QFUNCTION_HELPER` macro to properly annotate QFunction helper functions for code generation backends.
54- Add `CeedPragmaOptimizeOff` macro for code that is sensitive to floating point errors from fast math optimizations.
55- Rust support: split `libceed-sys` crate out of `libceed` and [publish both on crates.io](https://crates.io/crates/libceed).
56
57### Performance improvements
58
59### Examples
60
61- Solid mechanics mini-app updated to explore the performance impacts of various formulations in the initial and current configurations.
62- Fluid mechanics example adds GPU support and improves modularity.
63
64### Deprecated backends
65
66- 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.
67
68(v0-8)=
69
70## v0.8 (Mar 31, 2021)
71
72### Interface changes
73
74- Error handling improved to include enumerated error codes for C interface return values.
75- 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.
76
77### New features
78
79- Julia and Rust interfaces added, providing a nearly 1-1 correspondence with the C interface, plus some convenience features.
80- Static libraries can be built with `make STATIC=1` and the pkg-config file is installed accordingly.
81- Add {c:func}`CeedOperatorLinearAssembleSymbolic` and {c:func}`CeedOperatorLinearAssemble` to support full assembly of libCEED operators.
82
83### Performance improvements
84
85- New HIP MAGMA backends for hipMAGMA library users: `/gpu/hip/magma` and `/gpu/hip/magma/det`.
86- New HIP backends for improved tensor basis performance: `/gpu/hip/shared` and `/gpu/hip/gen`.
87
88### Examples
89
90- {ref}`example-petsc-elasticity` example updated with traction boundary conditions and improved Dirichlet boundary conditions.
91- {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.
92- {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.
93- {ref}`example-petsc-navier-stokes` example updated with support for performing convergence study and plotting order of convergence by polynomial degree.
94
95(v0-7)=
96
97## v0.7 (Sep 29, 2020)
98
99### Interface changes
100
101- Replace limited {code}`CeedInterlaceMode` with more flexible component stride {code}`compstride` in {code}`CeedElemRestriction` constructors.
102  As a result, the {code}`indices` parameter has been replaced with {code}`offsets` and the {code}`nnodes` parameter has been replaced with {code}`lsize`.
103  These changes improve support for mixed finite element methods.
104- Replace various uses of {code}`Ceed*Get*Status` with {code}`Ceed*Is*` in the backend API to match common nomenclature.
105- Replace {code}`CeedOperatorAssembleLinearDiagonal` with {c:func}`CeedOperatorLinearAssembleDiagonal` for clarity.
106- 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.
107- 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.
108- 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.
109- 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.
110  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`.
111- Added {code}`CeedQFunctionContext` object to manage user QFunction context data and reduce copies between device and host memory.
112- 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.
113
114### New features
115
116- New HIP backend: `/gpu/hip/ref`.
117- CeedQFunction support for user `CUfunction`s in some backends
118
119### Performance improvements
120
121- OCCA backend rebuilt to facilitate future performance enhancements.
122- Petsc BPs suite improved to reduce noise due to multiple calls to {code}`mpiexec`.
123
124### Examples
125
126- {ref}`example-petsc-elasticity` example updated with strain energy computation and more flexible boundary conditions.
127
128### Deprecated backends
129
130- The `/gpu/cuda/reg` backend has been removed, with its core features moved into `/gpu/cuda/ref` and `/gpu/cuda/shared`.
131
132(v0-6)=
133
134## v0.6 (Mar 29, 2020)
135
136libCEED v0.6 contains numerous new features and examples, as well as expanded
137documentation in [this new website](https://libceed.readthedocs.io).
138
139### New features
140
141- New Python interface using [CFFI](https://cffi.readthedocs.io/) provides a nearly
142  1-1 correspondence with the C interface, plus some convenience features.  For instance,
143  data stored in the {cpp:type}`CeedVector` structure are available without copy as
144  {py:class}`numpy.ndarray`.  Short tutorials are provided in
145  [Binder](https://mybinder.org/v2/gh/CEED/libCEED/main?urlpath=lab/tree/examples/tutorials/).
146- Linear QFunctions can be assembled as block-diagonal matrices (per quadrature point,
147  {c:func}`CeedOperatorAssembleLinearQFunction`) or to evaluate the diagonal
148  ({c:func}`CeedOperatorAssembleLinearDiagonal`).  These operations are useful for
149  preconditioning ingredients and are used in the libCEED's multigrid examples.
150- The inverse of separable operators can be obtained using
151  {c:func}`CeedOperatorCreateFDMElementInverse` and applied with
152  {c:func}`CeedOperatorApply`.  This is a useful preconditioning ingredient,
153  especially for Laplacians and related operators.
154- New functions: {c:func}`CeedVectorNorm`, {c:func}`CeedOperatorApplyAdd`,
155  {c:func}`CeedQFunctionView`, {c:func}`CeedOperatorView`.
156- Make public accessors for various attributes to facilitate writing composable code.
157- New backend: `/cpu/self/memcheck/serial`.
158- QFunctions using variable-length array (VLA) pointer constructs can be used with CUDA
159  backends.  (Single source is coming soon for OCCA backends.)
160- Fix some missing edge cases in CUDA backend.
161
162### Performance Improvements
163
164- MAGMA backend performance optimization and non-tensor bases.
165- No-copy optimization in {c:func}`CeedOperatorApply`.
166
167### Interface changes
168
169- Replace {code}`CeedElemRestrictionCreateIdentity` and
170  {code}`CeedElemRestrictionCreateBlocked` with more flexible
171  {c:func}`CeedElemRestrictionCreateStrided` and
172  {c:func}`CeedElemRestrictionCreateBlockedStrided`.
173- Add arguments to {c:func}`CeedQFunctionCreateIdentity`.
174- Replace ambiguous uses of {cpp:enum}`CeedTransposeMode` for L-vector identification
175  with {cpp:enum}`CeedInterlaceMode`.  This is now an attribute of the
176  {cpp:type}`CeedElemRestriction` (see {c:func}`CeedElemRestrictionCreate`) and no
177  longer passed as `lmode` arguments to {c:func}`CeedOperatorSetField` and
178  {c:func}`CeedElemRestrictionApply`.
179
180### Examples
181
182libCEED-0.6 contains greatly expanded examples with {ref}`new documentation <Examples>`.
183Notable additions include:
184
185- Standalone {ref}`ex2-surface` ({file}`examples/ceed/ex2-surface`): compute the area of
186  a domain in 1, 2, and 3 dimensions by applying a Laplacian.
187
188- PETSc {ref}`example-petsc-area` ({file}`examples/petsc/area.c`): computes surface area
189  of domains (like the cube and sphere) by direct integration on a surface mesh;
190  demonstrates geometric dimension different from topological dimension.
191
192- PETSc {ref}`example-petsc-bps`:
193
194  - {file}`examples/petsc/bpsraw.c` (formerly `bps.c`): transparent CUDA support.
195  - {file}`examples/petsc/bps.c` (formerly `bpsdmplex.c`): performance improvements
196    and transparent CUDA support.
197  - {ref}`example-petsc-bps-sphere` ({file}`examples/petsc/bpssphere.c`):
198    generalizations of all CEED BPs to the surface of the sphere; demonstrates geometric
199    dimension different from topological dimension.
200
201- {ref}`example-petsc-multigrid` ({file}`examples/petsc/multigrid.c`): new p-multigrid
202  solver with algebraic multigrid coarse solve.
203
204- {ref}`example-petsc-navier-stokes` ({file}`examples/fluids/navierstokes.c`; formerly
205  `examples/navier-stokes`): unstructured grid support (using PETSc's `DMPlex`),
206  implicit time integration, SU/SUPG stabilization, free-slip boundary conditions, and
207  quasi-2D computational domain support.
208
209- {ref}`example-petsc-elasticity` ({file}`examples/solids/elasticity.c`): new solver for
210  linear elasticity, small-strain hyperelasticity, and globalized finite-strain
211  hyperelasticity using p-multigrid with algebraic multigrid coarse solve.
212
213(v0-5)=
214
215## v0.5 (Sep 18, 2019)
216
217For this release, several improvements were made. Two new CUDA backends were added to
218the family of backends, of which, the new `cuda-gen` backend achieves state-of-the-art
219performance using single-source {ref}`CeedQFunction`. From this release, users
220can define Q-Functions in a single source code independently of the targeted backend
221with the aid of a new macro `CEED QFUNCTION` to support JIT (Just-In-Time) and CPU
222compilation of the user provided {ref}`CeedQFunction` code. To allow a unified
223declaration, the {ref}`CeedQFunction` API has undergone a slight change:
224the `QFunctionField` parameter `ncomp` has been changed to `size`. This change
225requires setting the previous value of `ncomp` to `ncomp*dim` when adding a
226`QFunctionField` with eval mode `CEED EVAL GRAD`.
227
228Additionally, new CPU backends
229were included in this release, such as the `/cpu/self/opt/*` backends (which are
230written in pure C and use partial **E-vectors** to improve performance) and the
231`/cpu/self/ref/memcheck` backend (which relies upon the
232[Valgrind](http://valgrind.org/) Memcheck tool to help verify that user
233{ref}`CeedQFunction` have no undefined values).
234This release also included various performance improvements, bug fixes, new examples,
235and improved tests. Among these improvements, vectorized instructions for
236{ref}`CeedQFunction` code compiled for CPU were enhanced by using `CeedPragmaSIMD`
237instead of `CeedPragmaOMP`, implementation of a {ref}`CeedQFunction` gallery and
238identity Q-Functions were introduced, and the PETSc benchmark problems were expanded
239to include unstructured meshes handling were. For this expansion, the prior version of
240the PETSc BPs, which only included data associated with structured geometries, were
241renamed `bpsraw`, and the new version of the BPs, which can handle data associated
242with any unstructured geometry, were called `bps`. Additionally, other benchmark
243problems, namely BP2 and BP4 (the vector-valued versions of BP1 and BP3, respectively),
244and BP5 and BP6 (the collocated versions---for which the quadrature points are the same
245as the Gauss Lobatto nodes---of BP3 and BP4 respectively) were added to the PETSc
246examples. Furthermoew, another standalone libCEED example, called `ex2`, which
247computes the surface area of a given mesh was added to this release.
248
249Backends available in this release:
250
251| CEED resource (`-ceed`)  | Backend                                             |
252|--------------------------|-----------------------------------------------------|
253| `/cpu/self/ref/serial`   | Serial reference implementation                     |
254| `/cpu/self/ref/blocked`  | Blocked reference implementation                    |
255| `/cpu/self/ref/memcheck` | Memcheck backend, undefined value checks            |
256| `/cpu/self/opt/serial`   | Serial optimized C implementation                   |
257| `/cpu/self/opt/blocked`  | Blocked optimized C implementation                  |
258| `/cpu/self/avx/serial`   | Serial AVX implementation                           |
259| `/cpu/self/avx/blocked`  | Blocked AVX implementation                          |
260| `/cpu/self/xsmm/serial`  | Serial LIBXSMM implementation                       |
261| `/cpu/self/xsmm/blocked` | Blocked LIBXSMM implementation                      |
262| `/cpu/occa`              | Serial OCCA kernels                                 |
263| `/gpu/occa`              | CUDA OCCA kernels                                   |
264| `/omp/occa`              | OpenMP OCCA kernels                                 |
265| `/ocl/occa`              | OpenCL OCCA kernels                                 |
266| `/gpu/cuda/ref`          | Reference pure CUDA kernels                         |
267| `/gpu/cuda/reg`          | Pure CUDA kernels using one thread per element      |
268| `/gpu/cuda/shared`       | Optimized pure CUDA kernels using shared memory     |
269| `/gpu/cuda/gen`          | Optimized pure CUDA kernels using code generation   |
270| `/gpu/magma`             | CUDA MAGMA kernels                                  |
271
272Examples available in this release:
273
274:::{list-table}
275:header-rows: 1
276:widths: auto
277* - User code
278  - Example
279* - `ceed`
280  - * ex1 (volume)
281    * ex2 (surface)
282* - `mfem`
283  - * BP1 (scalar mass operator)
284    * BP3 (scalar Laplace operator)
285* - `petsc`
286  - * BP1 (scalar mass operator)
287    * BP2 (vector mass operator)
288    * BP3 (scalar Laplace operator)
289    * BP4 (vector Laplace operator)
290    * BP5 (collocated scalar Laplace operator)
291    * BP6 (collocated vector Laplace operator)
292    * Navier-Stokes
293* - `nek5000`
294  - * BP1 (scalar mass operator)
295    * BP3 (scalar Laplace operator)
296:::
297
298(v0-4)=
299
300## v0.4 (Apr 1, 2019)
301
302libCEED v0.4 was made again publicly available in the second full CEED software
303distribution, release CEED 2.0. This release contained notable features, such as
304four new CPU backends, two new GPU backends, CPU backend optimizations, initial
305support for operator composition, performance benchmarking, and a Navier-Stokes demo.
306The new CPU backends in this release came in two families. The `/cpu/self/*/serial`
307backends process one element at a time and are intended for meshes with a smaller number
308of high order elements. The `/cpu/self/*/blocked` backends process blocked batches of
309eight interlaced elements and are intended for meshes with higher numbers of elements.
310The `/cpu/self/avx/*` backends rely upon AVX instructions to provide vectorized CPU
311performance. The `/cpu/self/xsmm/*` backends rely upon the
312[LIBXSMM](http://github.com/hfp/libxsmm) package to provide vectorized CPU
313performance. The `/gpu/cuda/*` backends provide GPU performance strictly using CUDA.
314The `/gpu/cuda/ref` backend is a reference CUDA backend, providing reasonable
315performance for most problem configurations. The `/gpu/cuda/reg` backend uses a simple
316parallelization approach, where each thread treats a finite element. Using just in time
317compilation, provided by nvrtc (NVidia Runtime Compiler), and runtime parameters, this
318backend unroll loops and map memory address to registers. The `/gpu/cuda/reg` backend
319achieve good peak performance for 1D, 2D, and low order 3D problems, but performance
320deteriorates very quickly when threads run out of registers.
321
322A new explicit time-stepping Navier-Stokes solver was added to the family of libCEED
323examples in the `examples/petsc` directory (see {ref}`example-petsc-navier-stokes`).
324This example solves the time-dependent Navier-Stokes equations of compressible gas
325dynamics in a static Eulerian three-dimensional frame, using structured high-order
326finite/spectral element spatial discretizations and explicit high-order time-stepping
327(available in PETSc). Moreover, the Navier-Stokes example was developed using PETSc,
328so that the pointwise physics (defined at quadrature points) is separated from the
329parallelization and meshing concerns.
330
331Backends available in this release:
332
333| CEED resource (`-ceed`)  | Backend                                             |
334|--------------------------|-----------------------------------------------------|
335| `/cpu/self/ref/serial`   | Serial reference implementation                     |
336| `/cpu/self/ref/blocked`  | Blocked reference implementation                    |
337| `/cpu/self/tmpl`         | Backend template, defaults to `/cpu/self/blocked`   |
338| `/cpu/self/avx/serial`   | Serial AVX implementation                           |
339| `/cpu/self/avx/blocked`  | Blocked AVX implementation                          |
340| `/cpu/self/xsmm/serial`  | Serial LIBXSMM implementation                       |
341| `/cpu/self/xsmm/blocked` | Blocked LIBXSMM implementation                      |
342| `/cpu/occa`              | Serial OCCA kernels                                 |
343| `/gpu/occa`              | CUDA OCCA kernels                                   |
344| `/omp/occa`              | OpenMP OCCA kernels                                 |
345| `/ocl/occa`              | OpenCL OCCA kernels                                 |
346| `/gpu/cuda/ref`          | Reference pure CUDA kernels                         |
347| `/gpu/cuda/reg`          | Pure CUDA kernels using one thread per element      |
348| `/gpu/magma`             | CUDA MAGMA kernels                                  |
349
350Examples available in this release:
351
352:::{list-table}
353:header-rows: 1
354:widths: auto
355* - User code
356  - Example
357* - `ceed`
358  - * ex1 (volume)
359* - `mfem`
360  - * BP1 (scalar mass operator)
361    * BP3 (scalar Laplace operator)
362* - `petsc`
363  - * BP1 (scalar mass operator)
364    * BP3 (scalar Laplace operator)
365    * Navier-Stokes
366* - `nek5000`
367  - * BP1 (scalar mass operator)
368    * BP3 (scalar Laplace operator)
369:::
370
371(v0-3)=
372
373## v0.3 (Sep 30, 2018)
374
375Notable features in this release include active/passive field interface, support for
376non-tensor bases, backend optimization, and improved Fortran interface. This release
377also focused on providing improved continuous integration, and many new tests with code
378coverage reports of about 90%. This release also provided a significant change to the
379public interface: a {ref}`CeedQFunction` can take any number of named input and output
380arguments while {ref}`CeedOperator` connects them to the actual data, which may be
381supplied explicitly to `CeedOperatorApply()` (active) or separately via
382`CeedOperatorSetField()` (passive). This interface change enables reusable libraries
383of CeedQFunctions and composition of block solvers constructed using
384{ref}`CeedOperator`. A concept of blocked restriction was added to this release and
385used in an optimized CPU backend. Although this is typically not visible to the user,
386it enables effective use of arbitrary-length SIMD while maintaining cache locality.
387This CPU backend also implements an algebraic factorization of tensor product gradients
388to perform fewer operations than standard application of interpolation and
389differentiation from nodes to quadrature points. This algebraic formulation
390automatically supports non-polynomial and non-interpolatory bases, thus is more general
391than the more common derivation in terms of Lagrange polynomials on the quadrature points.
392
393Backends available in this release:
394
395| CEED resource (`-ceed`) | Backend                                             |
396|-------------------------|-----------------------------------------------------|
397| `/cpu/self/blocked`     | Blocked reference implementation                    |
398| `/cpu/self/ref`         | Serial reference implementation                     |
399| `/cpu/self/tmpl`        | Backend template, defaults to `/cpu/self/blocked`   |
400| `/cpu/occa`             | Serial OCCA kernels                                 |
401| `/gpu/occa`             | CUDA OCCA kernels                                   |
402| `/omp/occa`             | OpenMP OCCA kernels                                 |
403| `/ocl/occa`             | OpenCL OCCA kernels                                 |
404| `/gpu/magma`            | CUDA MAGMA kernels                                  |
405
406Examples available in this release:
407
408:::{list-table}
409:header-rows: 1
410:widths: auto
411* - User code
412  - Example
413* - `ceed`
414  - * ex1 (volume)
415* - `mfem`
416  - * BP1 (scalar mass operator)
417    * BP3 (scalar Laplace operator)
418* - `petsc`
419  - * BP1 (scalar mass operator)
420    * BP3 (scalar Laplace operator)
421* - `nek5000`
422  - * BP1 (scalar mass operator)
423    * BP3 (scalar Laplace operator)
424:::
425
426(v0-21)=
427
428## v0.21 (Sep 30, 2018)
429
430A MAGMA backend (which relies upon the
431[MAGMA](https://bitbucket.org/icl/magma) package) was integrated in libCEED for this
432release. This initial integration set up the framework of using MAGMA and provided the
433libCEED functionality through MAGMA kernels as one of libCEED’s computational backends.
434As any other backend, the MAGMA backend provides extended basic data structures for
435{ref}`CeedVector`, {ref}`CeedElemRestriction`, and {ref}`CeedOperator`, and implements
436the fundamental CEED building blocks to work with the new data structures.
437In general, the MAGMA-specific data structures keep the libCEED pointers to CPU data
438but also add corresponding device (e.g., GPU) pointers to the data. Coherency is handled
439internally, and thus seamlessly to the user, through the functions/methods that are
440provided to support them.
441
442Backends available in this release:
443
444| CEED resource (`-ceed`) | Backend                         |
445|-------------------------|---------------------------------|
446| `/cpu/self`             | Serial reference implementation |
447| `/cpu/occa`             | Serial OCCA kernels             |
448| `/gpu/occa`             | CUDA OCCA kernels               |
449| `/omp/occa`             | OpenMP OCCA kernels             |
450| `/ocl/occa`             | OpenCL OCCA kernels             |
451| `/gpu/magma`            | CUDA MAGMA kernels              |
452
453Examples available in this release:
454
455:::{list-table}
456:header-rows: 1
457:widths: auto
458* - User code
459  - Example
460* - `ceed`
461  - * ex1 (volume)
462* - `mfem`
463  - * BP1 (scalar mass operator)
464    * BP3 (scalar Laplace operator)
465* - `petsc`
466  - * BP1 (scalar mass operator)
467* - `nek5000`
468  - * BP1 (scalar mass operator)
469:::
470
471(v0-2)=
472
473## v0.2 (Mar 30, 2018)
474
475libCEED was made publicly available the first full CEED software distribution, release
476CEED 1.0. The distribution was made available using the Spack package manager to provide
477a common, easy-to-use build environment, where the user can build the CEED distribution
478with all dependencies. This release included a new Fortran interface for the library.
479This release also contained major improvements in the OCCA backend (including a new
480`/ocl/occa` backend) and new examples. The standalone libCEED example was modified to
481compute the volume volume of a given mesh (in 1D, 2D, or 3D) and placed in an
482`examples/ceed` subfolder. A new `mfem` example to perform BP3 (with the application
483of the Laplace operator) was also added to this release.
484
485Backends available in this release:
486
487| CEED resource (`-ceed`) | Backend                         |
488|-------------------------|---------------------------------|
489| `/cpu/self`             | Serial reference implementation |
490| `/cpu/occa`             | Serial OCCA kernels             |
491| `/gpu/occa`             | CUDA OCCA kernels               |
492| `/omp/occa`             | OpenMP OCCA kernels             |
493| `/ocl/occa`             | OpenCL OCCA kernels             |
494
495Examples available in this release:
496
497:::{list-table}
498:header-rows: 1
499:widths: auto
500* - User code
501  - Example
502* - `ceed`
503  - * ex1 (volume)
504* - `mfem`
505  - * BP1 (scalar mass operator)
506    * BP3 (scalar Laplace operator)
507* - `petsc`
508  - * BP1 (scalar mass operator)
509* - `nek5000`
510  - * BP1 (scalar mass operator)
511:::
512
513(v0-1)=
514
515## v0.1 (Jan 3, 2018)
516
517Initial low-level API of the CEED project. The low-level API provides a set of Finite
518Elements kernels and components for writing new low-level kernels. Examples include:
519vector and sparse linear algebra, element matrix assembly over a batch of elements,
520partial assembly and action for efficient high-order operators like mass, diffusion,
521advection, etc. The main goal of the low-level API is to establish the basis for the
522high-level API. Also, identifying such low-level kernels and providing a reference
523implementation for them serves as the basis for specialized backend implementations.
524This release contained several backends: `/cpu/self`, and backends which rely upon the
525[OCCA](http://github.com/libocca/occa) package, such as `/cpu/occa`,
526`/gpu/occa`, and `/omp/occa`.
527It also included several examples, in the `examples` folder:
528A standalone code that shows the usage of libCEED (with no external
529dependencies) to apply the Laplace operator, `ex1`; an `mfem` example to perform BP1
530(with the application of the mass operator); and a `petsc` example to perform BP1
531(with the application of the mass operator).
532
533Backends available in this release:
534
535| CEED resource (`-ceed`) | Backend                         |
536|-------------------------|---------------------------------|
537| `/cpu/self`             | Serial reference implementation |
538| `/cpu/occa`             | Serial OCCA kernels             |
539| `/gpu/occa`             | CUDA OCCA kernels               |
540| `/omp/occa`             | OpenMP OCCA kernels             |
541
542Examples available in this release:
543
544| User code             | Example                           |
545|-----------------------|-----------------------------------|
546| `ceed`                | ex1 (scalar Laplace operator)     |
547| `mfem`                | BP1 (scalar mass operator)        |
548| `petsc`               | BP1 (scalar mass operator)        |
549```
550