Checks: Remove mentions of codee tools

The descriptions of checks should be independent of the tooling that
helps detect them.

Checks/PWR050/README.md

@@ -9,8 +9,7 @@ be sped up using [multithreading](../../Glossary/Multithreading.md).
 ### Actions
 
 Implement a version of the forall loop using an Application Program Interface
-(API) that enables multithreading on the CPU. Codee assists the programmer by
-providing source code rewriting capabilities using OpenMP compiler directives.
+(API) that enables multithreading on the CPU, such as OpenMP.
 
 ### Relevance
 

Checks/PWR051/README.md

@@ -9,9 +9,7 @@ pattern can be sped up using [multithreading](../../Glossary/Multithreading.md).
 ### Actions
 
 Implement a version of the scalar reduction loop using an Application Program
-Interface (API) that enables multithreading on the CPU. Codee assists the
-programmer by providing source code rewriting capabilities using OpenMP compiler
-directives.
+Interface (API) that enables multithreading on the CPU, such as OpenMP.
 
 ### Relevance
 

Checks/PWR052/README.md

@@ -9,9 +9,7 @@ pattern can be sped up using [multithreading](../../Glossary/Multithreading.md).
 ### Actions
 
 Implement a version of the sparse reduction loop using an Application Program
-Interface (API) that enables multithreading on the CPU. Codee assists the
-programmer by providing source code rewriting capabilities using OpenMP compiler
-directives.
+Interface (API) that enables multithreading on the CPU, such as OpenMP.
 
 ### Relevance
 

Checks/PWR053/README.md

@@ -9,9 +9,8 @@ can be sped up using [vectorization](../../Glossary/Vectorization.md).
 ### Actions
 
 Implement a version of the forall loop using an Application Program Interface
-(API) that enables vectorization. Codee assists the programmer by providing
-source code rewriting capabilities using compiler directives provided by OpenMP
-standard, GNU compiler, LLVM compiler and Intel compiler.
+(API) that enables vectorization, such as OpenMP, or using compiler-specific
+directives.
 
 ### Relevance
 

Checks/PWR054/README.md

@@ -9,9 +9,8 @@ pattern can be sped up using [vectorization](../../Glossary/Vectorization.md).
 ### Actions
 
 Implement a version of the scalar reduction loop using an Application Program
-Interface (API) that enables vectorization. Codee assists the programmer by
-providing source code rewriting capabilities using compiler directives provided
-by OpenMP standard, GNU compiler, LLVM compiler and Intel compiler.
+Interface (API) that enables vectorization, or using compiler-specific
+directives.
 
 ### Relevance
 

Checks/PWR055/README.md

@@ -9,9 +9,7 @@ can be sped up by [offloading it to an accelerator](../../Glossary/Offloading.md
 ### Actions
 
 Implement a version of the forall loop using an Application Program Interface
-(API) that enables offloading to accelerators. Codee assists the programmer by
-providing source code rewriting capabilities using OpenMP and OpenACC compiler
-directives.
+(API) that enables offloading to accelerators, such as OpenMP or OpenACC.
 
 ### Relevance
 

Checks/PWR056/README.md

@@ -10,10 +10,8 @@ pattern can be sped up by
 ### Actions
 
 Implement a version of the scalar reduction loop using an Application Program
-Interface (API) that enables offloading to accelerators. Codee assists the
-programmer by providing source code rewriting capabilities using
-[OpenMP](https://en.wikipedia.org/wiki/OpenMP) and
-[OpenACC](https://en.wikipedia.org/wiki/OpenACC) compiler directives.
+Interface (API) that enables offloading to accelerators, such as OpenMP or
+OpenACC.
 
 ### Relevance
 

Checks/PWR057/README.md

@@ -5,16 +5,13 @@
 A loop containing the
 [sparse reduction](../../Glossary/Patterns-for-performance-optimization/Sparse-reduction.md)
 pattern can be sped up by
-[offloading it to an accelerator](../../Glossary/Offloading.md). Codee can do this
-automatically, no source code modification is needed by the developer.
+[offloading it to an accelerator](../../Glossary/Offloading.md).
 
 ### Actions
 
 Implement a version of the sparse reduction loop using an Application Program
-Interface (API) that enables offloading to accelerators. Codee assists the
-programmer by providing source code rewriting capabilities using
-[OpenMP](https://en.wikipedia.org/wiki/OpenMP) and
-[OpenACC](https://en.wikipedia.org/wiki/OpenACC) compiler directives.
+Interface (API) that enables offloading to accelerators, such as OpenMP or
+OpenACC.
 
 ### Relevance
 

Checks/RMK001/README.md

@@ -8,11 +8,7 @@ performance of the nested loop.
 ### Actions
 
 Parallelize the outer loop using multi-threading and the inner loop using
-vectorization. You can do so for OpenMP by invoking:
-
-```bash
-pwdirectives --multi omp-for+simd foo.c:5 -o foo-hybrid.c
-```
+vectorization.
 
 ### Relevance
 

Checks/RMK002/README.md

@@ -8,11 +8,7 @@ parallel performance of the nested loop.
 ### Actions
 
 Parallelize the nested loops using offloading for the outer loop and SIMD
-vectorization for the inner loop. You can do so for OpenMP by invoking:
-
-```bash
-pwdirectives --offload omp-teams+simd foo.c:5 -o foo-hybrid.c
-```
+vectorization for the inner loop.
 
 ### Relevance
 

Checks/RMK014/README.md

@@ -7,10 +7,9 @@ the loop body.
 
 ### Actions
 
-Use `pwdirectives` to vectorize the loop.
-
-Alternatively, you can perform [loop fission](../../Glossary/Loop-fission.md) to
-isolate the unpredictable memory accesses to a separate loop.
+Perform [loop fission](../../Glossary/Loop-fission.md) to isolate the
+unpredictable memory accesses to a separate loop. If not possible, consider
+evaluating the performance of the alternative scenarios described below.
 
 ### Relevance
 

Glossary/Offloading-data-transfers.md

@@ -26,9 +26,3 @@ layouts to reduce data transfers and memory access times.
 
 Alternatively, the OpenACC offload directives also enable the implementation of
 efficient data transfers in GPU programming.
-
-Overall, the Codee tools support GPU programming for C/C++/Fortran using OpenMP
-and OpenACC. It can automatically insert OpenMP or OpenACC directives in the
-source code to offload to accelerator devices such as GPUs, helping novice
-programmers to write GPU code at the expert level. Codee provides a systematic,
-more predictable approach that leverages parallel programming best practices.