Back to the main page of this manual | Input reference of CP2K version 8.2 |
CP2K_INPUT /
GLOBAL /
DBCSR
AVG_ELEMENTS_IMAGES {Integer} |
|
Average number of elements (dense limit) for each image, which also corresponds to the average number of elements exchanged between MPI processes during the operations. A negative or zero value means unlimited. | |
This keyword cannot be repeated and it expects precisely one integer. | |
Default value:
-1 |
COMM_THREAD_LOAD {Integer} |
|
If a communications thread is used, specify how much multiplication workload (%) the thread should perform in addition to communication tasks. A negative value leaves the decision up to DBCSR. | |
This keyword cannot be repeated and it expects precisely one integer. | |
Default value:
-1 |
MAX_ELEMENTS_PER_BLOCK {Integer} |
|
Default block size for turning dense matrices in blocked ones | |
This keyword cannot be repeated and it expects precisely one integer. | |
Default value:
32 |
MM_DRIVER {Keyword} |
|
Select which backend to use preferably for matrix block multiplications on the host. | |
This keyword cannot be repeated and it expects precisely one keyword. | |
Default value:
AUTO |
|
List of valid keywords:
|
|
This keyword cites the following reference: [Heinecke2016] |
MM_STACK_SIZE {Integer} |
|
Size of multiplication parameter stack. A negative value leaves the decision up to DBCSR. | |
This keyword cannot be repeated and it expects precisely one integer. | |
Default value:
-1 |
MULTREC_LIMIT {Integer} |
|
Recursion limit of cache oblivious multrec algorithm. | |
This keyword cannot be repeated and it expects precisely one integer. | |
Default value:
512 |
NUM_LAYERS_3D {Integer} |
|
Number of layers for the 3D multplication algorithm. | |
This keyword cannot be repeated and it expects precisely one integer. | |
Default value:
1 |
NUM_MULT_IMAGES {Integer} |
|
Multiplicative factor for number of virtual images. | |
This keyword cannot be repeated and it expects precisely one integer. | |
Default value:
1 |
N_SIZE_MNK_STACKS {Integer} |
|
Number of stacks to use for distinct atomic sizes (e.g., 2 for a system of mostly waters). | |
This keyword cannot be repeated and it expects precisely one integer. | |
Default value:
3 |
USE_COMM_THREAD {Logical} |
|
During multiplication, use a thread to periodically poll MPI to progress outstanding message completions. This is beneficial on systems without a DMA-capable network adapter e.g. Cray XE6. | |
This keyword cannot be repeated and it expects precisely one logical. | |
Default value:
.TRUE. |
USE_MEMPOOLS_CPU {Logical} |
|
Enable memory pools on the CPU. | |
This keyword cannot be repeated and it expects precisely one logical. | |
Default value:
.FALSE. |
USE_MPI_ALLOCATOR {Logical} |
|
Use MPI allocator to allocate buffers used in MPI communications. | |
This keyword cannot be repeated and it expects precisely one logical. | |
Default value:
.FALSE. |
USE_MPI_RMA {Logical} |
|
Use RMA for MPI communications for each image, which also corresponds to the number of elements exchanged between MPI processes during the operations. | |
This keyword cannot be repeated and it expects precisely one logical. | |
Default value:
.FALSE. |
Back to the main page of this manual or the CP2K home page | (Last update: 14.12.2021) |