I came across the following article https://blogs.saphana.com/2013/04/07/best-practices-for-sap-hana-data-loads/ and was wondering why it is a good practice not to use large integer and decimal types as data types for columns in HANA, the reason being they cannot use SIMD instructions optimally. I am not able to link SIMD instructions with large integer / decimal data types. Could somebody please illustrate why it is so ..TIA
我遇到了以下文章https://blogs.saphana.com/2013/04/07/best-practices-for-sap-hana-data-loads/,并想知道为什么不使用大整数是一个好习惯和小数类型作为HANA中列的数据类型,原因是它们无法以最佳方式使用SIMD指令。我无法将SIMD指令与大整数/十进制数据类型链接。有人可以说明为什么会这样.. .. ..
1 个解决方案
#1
2
I believe you refer to this part of the blog post:
我相信你参考博文的这一部分:
[...] Same with large integer and decimal fields – they are more expensive because they can’t use such efficient SIMD instructions. [...]
[...]与大整数和十进制字段相同 - 它们更昂贵,因为它们无法使用这种有效的SIMD指令。 [...]
This statement is inaccurate and you can safely ignore it.
此声明不准确,您可以放心地忽略它。
#1
2
I believe you refer to this part of the blog post:
我相信你参考博文的这一部分:
[...] Same with large integer and decimal fields – they are more expensive because they can’t use such efficient SIMD instructions. [...]
[...]与大整数和十进制字段相同 - 它们更昂贵,因为它们无法使用这种有效的SIMD指令。 [...]
This statement is inaccurate and you can safely ignore it.
此声明不准确,您可以放心地忽略它。