NBSEBLOCKV - Revision history

Tal: Tal moved page Construction:NBSEBLOCKV to NBSEBLOCKV

2024-12-18T09:52:44Z

Tal moved page Construction:NBSEBLOCKV to NBSEBLOCKV

← Older revision	Revision as of 09:52, 18 December 2024
(No difference)

Tal at 10:57, 14 June 2024

2024-06-14T10:57:55Z

← Older revision		Revision as of 10:57, 14 June 2024
Line 27:		Line 27:
	== Related tags and sections ==		== Related tags and sections ==
	{{TAG\|BSE}}, {{TAG\|NBSEBLOCKO}}, [[BSE calculations]], [[TDDFT calculations]]		{{TAG\|BSE}}, {{TAG\|NBSEBLOCKO}}, [[BSE calculations]], [[TDDFT calculations]]

	----		----
	[[Category:INCAR tag]][[Category:Bethe-Salpeter_equations]][[Category:Many-body perturbation theory]]		[[Category:INCAR tag]][[Category:Bethe-Salpeter_equations]][[Category:Many-body perturbation theory]]

Tal at 10:57, 14 June 2024

2024-06-14T10:57:45Z

← Older revision		Revision as of 10:57, 14 June 2024
Line 24:		Line 24:

	We recommend using parallelization over bands only if the number of MPI ranks in the calculation exceeds <math>\text{total ranks}</math>.		We recommend using parallelization over bands only if the number of MPI ranks in the calculation exceeds <math>\text{total ranks}</math>.
			{{NB\|mind\|The {{TAG\|NBSEBLOCKV}} and {{TAG\|NBSEBLOCKO}} tags are available as of VASP.6.5.0}}
	== Related tags and sections ==		== Related tags and sections ==
	{{TAG\|BSE}}, {{TAG\|NBSEBLOCKO}}, [[BSE calculations]], [[TDDFT calculations]]		{{TAG\|BSE}}, {{TAG\|NBSEBLOCKO}}, [[BSE calculations]], [[TDDFT calculations]]

Tal: /* Related tags and sections */

2024-06-12T13:01:35Z

Related tags and sections

← Older revision		Revision as of 13:01, 12 June 2024
Line 26:		Line 26:

	== Related tags and sections ==		== Related tags and sections ==
	{{TAG\|BSE}}, {{TAG\|NBSEBLOCKO}}, [[BSE calculations]]		{{TAG\|BSE}}, {{TAG\|NBSEBLOCKO}}, [[BSE calculations]], [[TDDFT calculations]]

	----		----
	[[Category:INCAR tag]][[Category:Bethe-Salpeter_equations]][[Category:Many-body perturbation theory]]		[[Category:INCAR tag]][[Category:Bethe-Salpeter_equations]][[Category:Many-body perturbation theory]]

Tal at 12:57, 12 June 2024

2024-06-12T12:57:19Z

← Older revision		Revision as of 12:57, 12 June 2024
Line 23:		Line 23:
	{{NB\|mind\|Parallelization over bands with {{TAG\|NBSEBLOCKV}} does not work with the old BSE driver, i.e., {{TAG\|IBSE}}{{=}}0}}		{{NB\|mind\|Parallelization over bands with {{TAG\|NBSEBLOCKV}} does not work with the old BSE driver, i.e., {{TAG\|IBSE}}{{=}}0}}

	We recommend using parallelization over bands only if the number of MPI ranks in the calculation exceeds <math>\text{total ranks}</math>~~. Furthermore, to minimize load imbalance, the blocking factor should divide the corresponding number of bands~~.		We recommend using parallelization over bands only if the number of MPI ranks in the calculation exceeds <math>\text{total ranks}</math>.

	== Related tags and sections ==		== Related tags and sections ==

Tal at 12:54, 12 June 2024

2024-06-12T12:54:59Z

← Older revision		Revision as of 12:54, 12 June 2024
Line 22:		Line 22:
	If neither <math>\mathrm{NBSEBLOCKV}</math> nor <math>\mathrm{NBSEBLOCKO}</math> is specified, no paralliziation over bands is used and <math>\mathrm{NBLKO=1}</math> and <math>\mathrm{NBLKV=1}</math>.		If neither <math>\mathrm{NBSEBLOCKV}</math> nor <math>\mathrm{NBSEBLOCKO}</math> is specified, no paralliziation over bands is used and <math>\mathrm{NBLKO=1}</math> and <math>\mathrm{NBLKV=1}</math>.
	{{NB\|mind\|Parallelization over bands with {{TAG\|NBSEBLOCKV}} does not work with the old BSE driver, i.e., {{TAG\|IBSE}}{{=}}0}}		{{NB\|mind\|Parallelization over bands with {{TAG\|NBSEBLOCKV}} does not work with the old BSE driver, i.e., {{TAG\|IBSE}}{{=}}0}}

			We recommend using parallelization over bands only if the number of MPI ranks in the calculation exceeds <math>\text{total ranks}</math>. Furthermore, to minimize load imbalance, the blocking factor should divide the corresponding number of bands.

	== Related tags and sections ==		== Related tags and sections ==

Tal at 12:39, 12 June 2024

2024-06-12T12:39:15Z

← Older revision		Revision as of 12:39, 12 June 2024
Line 4:		Line 4:
	Description: {{TAG\|NBSEBLOCKV}} specifies the blocking factor for the unoccupied states when setting up the BSE Hamiltonian.		Description: {{TAG\|NBSEBLOCKV}} specifies the blocking factor for the unoccupied states when setting up the BSE Hamiltonian.
	----		----
	By default, the construction of the BSE Hamiltonian in VASP is parallelized over k-points, such that each MPI rank can compute a pair of k-points. This way the BSE Hamiltonian setup can be parallelized ~~up to~~ <math>\~~mathrm~~{~~NKPTS(NKPTS+1)/2~~}~~</math> MPI ranks or <math>~~\mathrm{NKPTS~~2~~(NKPTS2+1)/2}</math> for spin-polarized case~~, where~~ <math>\~~mathrm~~{~~NKPTS}</math> is the~~ total ~~number of k-points in the full Brillouin zone. However, if a large number of MPI~~ ranks is used in a calculation with too few k-point, this leads to load imbalance, where some of the MPI rank will have no data to compute. In such cases, it is recommended to use parallelization over bands. If the paralliziation over bands is used, all occupied (unoccupied) bands are divided into a <math>\mathrm{NBLKO (NBLKV)}~~</math> number of blocks, where <math>~~\mathrm{~~NBLKO=NBANDSO/NBSEBLOCKO}</math> and <math>~~\~~mathrm{NBLKV=NBANDSV/NBSEBLOCKV}</math>. This bands blocking allows VASP to parallelize the setup of the matrix up to <math>~~\mathrm{NBLKONBLKVNKPTS(NBLKONBLKV*NKPTS+1)/2}</math> ~~ranks.~~		By default, the construction of the BSE Hamiltonian in VASP is parallelized over '''k'''-points, such that each MPI rank can compute a pair of '''k'''-points. This way the BSE Hamiltonian setup can be parallelized with
			::<math>\text{total ranks}=\mathrm{NKPTS\times(NKPTS+1)/2}</math>
			or for spin-polarized case
			::<math>\text{total ranks}=\mathrm{NKPTS\times 2\times(NKPTS\times 2+1)/2}</math>,

	If no <math>\mathrm{NBSEBLOCKV}</math> or <math>\mathrm{NBSEBLOCKO}</math> is specified, no paralliziation over bands is used and <math>\mathrm{NBLKO=1 (NBLKV=1)}</math>.		where <math>\mathrm{NKPTS}</math> is the total number of '''k'''-points in the full Brillouin zone. However, if a large number of MPI ranks is used in a calculation with too few '''k'''-point, this leads to load imbalance, where some of the MPI ranks will have no data to compute. In such cases, it is recommended to use parallelization over bands. If the parallelization over bands is used, all occupied (unoccupied) bands are divided into
			::<math>\mathrm{NBLKO=NBANDSO/NBSEBLOCKO}</math>
			and
			::<math>\mathrm{NBLKV=NBANDSV/NBSEBLOCKV}</math>
			blocks, respectively.

			Such a band blocking allows VASP to parallelize the setup of the matrix with
			::<math>\text{total ranks}=\mathrm{NBLKO\times NBLKV\times NKPTS\times (NBLKO\times NBLKV\times NKPTS+1)/2}</math>
			or for spin-polarized case
			::<math>\text{total ranks}=\mathrm{NBLKO\times NBLKV\times NKPTS\times 2\times (NBLKO\times NBLKV\times NKPTS\times 2+1)/2}</math>

			If neither <math>\mathrm{NBSEBLOCKV}</math> nor <math>\mathrm{NBSEBLOCKO}</math> is specified, no paralliziation over bands is used and <math>\mathrm{NBLKO=1}</math> and <math>\mathrm{NBLKV=1}</math>.
	{{NB\|mind\|Parallelization over bands with {{TAG\|NBSEBLOCKV}} does not work with the old BSE driver, i.e., {{TAG\|IBSE}}{{=}}0}}		{{NB\|mind\|Parallelization over bands with {{TAG\|NBSEBLOCKV}} does not work with the old BSE driver, i.e., {{TAG\|IBSE}}{{=}}0}}

Tal at 12:17, 11 June 2024

2024-06-11T12:17:10Z

← Older revision		Revision as of 12:17, 11 June 2024
Line 13:		Line 13:

	----		----
	[[Category:INCAR tag]][[Category:Bethe-Salpeter_equations]][[Category:Many-body perturbation theory~~]][[Category:GW~~]]		[[Category:INCAR tag]][[Category:Bethe-Salpeter_equations]][[Category:Many-body perturbation theory]]

Tal at 12:16, 11 June 2024

2024-06-11T12:16:57Z

← Older revision		Revision as of 12:16, 11 June 2024
Line 10:		Line 10:

	== Related tags and sections ==		== Related tags and sections ==
	{{TAG\|BSE}}, {{TAG\|~~NBSEBLOCKV~~}}, [[BSE calculations]]		{{TAG\|BSE}}, {{TAG\|NBSEBLOCKO}}, [[BSE calculations]]

	----		----
	[[Category:INCAR tag]][[Category:Bethe-Salpeter_equations]][[Category:Many-body perturbation theory]][[Category:GW]]		[[Category:INCAR tag]][[Category:Bethe-Salpeter_equations]][[Category:Many-body perturbation theory]][[Category:GW]]

Tal at 12:16, 11 June 2024

2024-06-11T12:16:29Z

← Older revision		Revision as of 12:16, 11 June 2024
Line 4:		Line 4:
	Description: {{TAG\|NBSEBLOCKV}} specifies the blocking factor for the unoccupied states when setting up the BSE Hamiltonian.		Description: {{TAG\|NBSEBLOCKV}} specifies the blocking factor for the unoccupied states when setting up the BSE Hamiltonian.
	----		----
	By default, the construction of the BSE Hamiltonian in VASP is parallelized over k-points, such that each MPI rank can compute a pair of k-points. This way the BSE Hamiltonian setup can be parallelized up to <math>\mathrm{NKPTS(NKPTS+1)/2}</math> MPI ranks or <math>\mathrm{NKPTS2(NKPTS2+1)/2}</math> for spin-polarized case, where <math>\mathrm{NKPTS}</math> is the total number of k-points in the full Brillouin zone. However, if a large number of MPI ranks is used in a calculation with too few k-point, this leads to load imbalance, where some of the MPI rank will have no data to compute. In such cases, it is recommended to use parallelization over bands. If the paralliziation over bands is used, all occupied (unoccupied) bands are divided into a <math>\mathrm{NBLKO (NBLKV)}</math> number of blocks, where <math>\mathrm{NBLKO=NBANDSO/NBSEBLOCKO}</math> and <math>\mathrm{NBLKV=NBANDSV/NBSEBLOCKV}</~~mathrm}~~. This bands blocking allows VASP to parallelize the setup of the matrix up to <math>\mathrm{NBLKONBLKVNKPTS(NBLKONBLKV*NKPTS+1)/2}</math> ranks.		By default, the construction of the BSE Hamiltonian in VASP is parallelized over k-points, such that each MPI rank can compute a pair of k-points. This way the BSE Hamiltonian setup can be parallelized up to <math>\mathrm{NKPTS(NKPTS+1)/2}</math> MPI ranks or <math>\mathrm{NKPTS2(NKPTS2+1)/2}</math> for spin-polarized case, where <math>\mathrm{NKPTS}</math> is the total number of k-points in the full Brillouin zone. However, if a large number of MPI ranks is used in a calculation with too few k-point, this leads to load imbalance, where some of the MPI rank will have no data to compute. In such cases, it is recommended to use parallelization over bands. If the paralliziation over bands is used, all occupied (unoccupied) bands are divided into a <math>\mathrm{NBLKO (NBLKV)}</math> number of blocks, where <math>\mathrm{NBLKO=NBANDSO/NBSEBLOCKO}</math> and <math>\mathrm{NBLKV=NBANDSV/NBSEBLOCKV}</math>. This bands blocking allows VASP to parallelize the setup of the matrix up to <math>\mathrm{NBLKONBLKVNKPTS(NBLKONBLKV*NKPTS+1)/2}</math> ranks.

	If no <math>\mathrm{NBSEBLOCKV}</math> or <math>\mathrm{NBSEBLOCKO}</math> is specified, no paralliziation over bands is used and <math>\mathrm{NBLKO=1 (NBLKV=1)}</math>.		If no <math>\mathrm{NBSEBLOCKV}</math> or <math>\mathrm{NBSEBLOCKO}</math> is specified, no paralliziation over bands is used and <math>\mathrm{NBLKO=1 (NBLKV=1)}</math>.