古いデータ
をテンプレートにして作成
[
トップ
] [
新規
|
一覧
|
単語検索
|
最終更新
|
ヘルプ
|
ログイン
]
開始行:
[[実験メモ]]
間違ったプログラムでとったデータを待避させておくページ
**行列積の演算比率 [#mf2df921]
device0:GTX680,device1:C2075 デバイス毎に演算する領域を0...
||1024|2048|4096|
|0/16|0.557949 sec|3.624982 sec|37.670954 sec|
|1/16|0.527421 sec|3.427084 sec|35.371037 sec|
|2/16|0.502123 sec|3.215250 sec|33.157107 sec|
|3/16|0.479628 sec|3.038906 sec|31.029226 sec|
|4/16|0.459781 sec|2.882433 sec|28.985048 sec|
|5/16|0.440090 sec|2.735505 sec|32.648426 sec|
|6/16|0.422981 sec|2.593523 sec|30.382086 sec|
|7/16|0.407589 sec|2.471455 sec|28.209738 sec|
|8/16|0.392270 sec|2.372276 sec|26.146735 sec|
|9/16|0.377942 sec|2.247158 sec|24.146682 sec|
|10/16|0.362653 sec|2.147961 sec|27.723224 sec|
|11/16|0.348761 sec|2.044283 sec|25.497689 sec|
|12/16|0.334292 sec|2.060612 sec|23.373813 sec|
|13/16|0.325189 sec|2.247655 sec|21.338587 sec|
|14/16|0.330456 sec|2.407149 sec|25.111639 sec|
|15/16|0.350757 sec|2.598573 sec|22.884979 sec|
|16/16|0.375198 sec|2.702994 sec|23.795871 sec|
演算する順番を入れ替えたヤツ
||1024|2048|4096|
|0/16|0.384743 sec|2.747596 sec|23.536639 sec|
|1/16|0.350861 sec|2.665804 sec|22.079087 sec|
|2/16|0.332244 sec|2.578075 sec|20.719701 sec|
|3/16|0.313784 sec|2.406388 sec|19.314298 sec|
|4/16|0.293451 sec|2.190142 sec|17.860523 sec|
|5/16|0.274033 sec|1.954465 sec|18.662917 sec|
|6/16|0.255426 sec|1.900703 sec|19.516629 sec|
|7/16|0.265986 sec|1.675310 sec|20.373492 sec|
|8/16|0.298371 sec|1.868513 sec|21.238329 sec|
|9/16|0.331015 sec|2.084559 sec|22.088926 sec|
|10/16|0.363965 sec|2.299673 sec|23.652402 sec|
|11/16|0.396812 sec|2.525626 sec|25.989311 sec|
|12/16|0.429121 sec|2.739838 sec|28.325305 sec|
|13/16|0.461695 sec|2.960269 sec|30.658135 sec|
|14/16|0.494613 sec|3.176896 sec|32.990394 sec|
|15/16|0.526124 sec|3.401501 sec|35.329504 sec|
|16/16|0.558489 sec|3.627642 sec|37.667827 sec|
**動的振り分け [#laff77ea]
|分割数||1024||2048||4096|
|16|7/16|0.271802 sec|6/16|1.928171 sec|3/16|18.806624 sec|
|32|20/32|0.374901 sec|18/32|2.167672 sec|10/32|16.188250...
↑バグがあったので信用ならないデータ
2048
||16||32||64||128||256||512||1024|
|6|1.85s|14|1.69s|28|1.73s|56|1.72s|106|1.79s|204|1.76s|3...
** [#o5d0ffaa]
|N|分割数|実行時間|
|512|13/16|0.075 sec|
|1024|13/16|0.326 sec|
|1536|12/16|0.920 sec|
|2048|12/16|2.03 sec|
|2560|11/16|4.14 sec|
|3072|13/16|8.01 sec|
|3584|11/16|14.15 sec|
|4096|13/16|21.3 sec|
| | | |
|8192|14/16|183.7 sec|
細かく
|N|分割数|実行時間|
|1020|14/16|0.27 sec|
|1021|14/16|0.28 sec|
|1022|14/16|0.27 sec|
|1023|14/16|0.28 sec|
|1024|13/16|0.32 sec|
|1025|13/16|0.35 sec|
|1026|13/16|0.35 sec|
|1027|13/16|0.35 sec|
|1028|13/16|0.34 sec|
|1029|13/16|0.35 sec|
| | | |
|2045|12/16|1.98 sec|
|2046|12/16|1.95 sec|
|2047|11/16|1.98 sec|
|2048|11/16|2.03 sec|
|2049|11/16|2.43 sec|
|2050|11/16|2.4 sec|
終了行:
[[実験メモ]]
間違ったプログラムでとったデータを待避させておくページ
**行列積の演算比率 [#mf2df921]
device0:GTX680,device1:C2075 デバイス毎に演算する領域を0...
||1024|2048|4096|
|0/16|0.557949 sec|3.624982 sec|37.670954 sec|
|1/16|0.527421 sec|3.427084 sec|35.371037 sec|
|2/16|0.502123 sec|3.215250 sec|33.157107 sec|
|3/16|0.479628 sec|3.038906 sec|31.029226 sec|
|4/16|0.459781 sec|2.882433 sec|28.985048 sec|
|5/16|0.440090 sec|2.735505 sec|32.648426 sec|
|6/16|0.422981 sec|2.593523 sec|30.382086 sec|
|7/16|0.407589 sec|2.471455 sec|28.209738 sec|
|8/16|0.392270 sec|2.372276 sec|26.146735 sec|
|9/16|0.377942 sec|2.247158 sec|24.146682 sec|
|10/16|0.362653 sec|2.147961 sec|27.723224 sec|
|11/16|0.348761 sec|2.044283 sec|25.497689 sec|
|12/16|0.334292 sec|2.060612 sec|23.373813 sec|
|13/16|0.325189 sec|2.247655 sec|21.338587 sec|
|14/16|0.330456 sec|2.407149 sec|25.111639 sec|
|15/16|0.350757 sec|2.598573 sec|22.884979 sec|
|16/16|0.375198 sec|2.702994 sec|23.795871 sec|
演算する順番を入れ替えたヤツ
||1024|2048|4096|
|0/16|0.384743 sec|2.747596 sec|23.536639 sec|
|1/16|0.350861 sec|2.665804 sec|22.079087 sec|
|2/16|0.332244 sec|2.578075 sec|20.719701 sec|
|3/16|0.313784 sec|2.406388 sec|19.314298 sec|
|4/16|0.293451 sec|2.190142 sec|17.860523 sec|
|5/16|0.274033 sec|1.954465 sec|18.662917 sec|
|6/16|0.255426 sec|1.900703 sec|19.516629 sec|
|7/16|0.265986 sec|1.675310 sec|20.373492 sec|
|8/16|0.298371 sec|1.868513 sec|21.238329 sec|
|9/16|0.331015 sec|2.084559 sec|22.088926 sec|
|10/16|0.363965 sec|2.299673 sec|23.652402 sec|
|11/16|0.396812 sec|2.525626 sec|25.989311 sec|
|12/16|0.429121 sec|2.739838 sec|28.325305 sec|
|13/16|0.461695 sec|2.960269 sec|30.658135 sec|
|14/16|0.494613 sec|3.176896 sec|32.990394 sec|
|15/16|0.526124 sec|3.401501 sec|35.329504 sec|
|16/16|0.558489 sec|3.627642 sec|37.667827 sec|
**動的振り分け [#laff77ea]
|分割数||1024||2048||4096|
|16|7/16|0.271802 sec|6/16|1.928171 sec|3/16|18.806624 sec|
|32|20/32|0.374901 sec|18/32|2.167672 sec|10/32|16.188250...
↑バグがあったので信用ならないデータ
2048
||16||32||64||128||256||512||1024|
|6|1.85s|14|1.69s|28|1.73s|56|1.72s|106|1.79s|204|1.76s|3...
** [#o5d0ffaa]
|N|分割数|実行時間|
|512|13/16|0.075 sec|
|1024|13/16|0.326 sec|
|1536|12/16|0.920 sec|
|2048|12/16|2.03 sec|
|2560|11/16|4.14 sec|
|3072|13/16|8.01 sec|
|3584|11/16|14.15 sec|
|4096|13/16|21.3 sec|
| | | |
|8192|14/16|183.7 sec|
細かく
|N|分割数|実行時間|
|1020|14/16|0.27 sec|
|1021|14/16|0.28 sec|
|1022|14/16|0.27 sec|
|1023|14/16|0.28 sec|
|1024|13/16|0.32 sec|
|1025|13/16|0.35 sec|
|1026|13/16|0.35 sec|
|1027|13/16|0.35 sec|
|1028|13/16|0.34 sec|
|1029|13/16|0.35 sec|
| | | |
|2045|12/16|1.98 sec|
|2046|12/16|1.95 sec|
|2047|11/16|1.98 sec|
|2048|11/16|2.03 sec|
|2049|11/16|2.43 sec|
|2050|11/16|2.4 sec|
ページ名: