用于向量和矩阵计算的高性能数学库
在哪里可以找到矢量和矩阵计算的快速库？ 我需要高性能。
更新：我需要它来进行游戏计算。
Where Can i find fast library for vector and Matrix Calculations?
I need high performance.
Update:I need it for games calculations.
如果你对这篇内容有疑问，欢迎到本站社区发帖提问 参与讨论，获取更多帮助，或者扫码二维码加入 Web 技术交流群。
绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱，如果其他用户或者作者回复了您的评论，将不能在第一时间通知您！
发布评论
评论（5）
选择 Blaze。另请检查并行核心设置中的基准测试。它当然是最快的。
Go for Blaze. Also check out the bench marking in parallel cores setup. Its certainly the fastest.
编辑：抱歉，我的水平不够高，无法在任何地方添加评论，但只是注意到您提到它是用于游戏的，除非您在没有任何硬件加速的情况下用软件编写引擎，否则您将获得的好处与代码的其余部分相比，矩阵操作（假设它们用于 TCL）大约为零。
我无法将您链接到任何特定的库，但是当您指定高性能时，如果没有人为您提供一些库，那么一些建议可能会有所帮助：
Edit: Sorry, I'm not high enough to add comments anywhere but just noticed you mentioned it is for games, unless you are writing your engine in software without any hardware acceleration, then the benefits you will get from the matrix manipulations (assuming that they are for TCL) are about nil in comparison to the rest of your code.
I can't link you to any particular library, but as you are specifying high performance a couple of suggestions that may help if nobody is forthcoming with a few libraries for you:
Eigen 应该非常好：
Eigen
Eigen is supposed to be very good:
Eigen
如果您有 x86 系统，则可以使用 Intel Math Kernel Library。
http://software.intel.com/en-us/articles/intel- mkl/
在我的旧四盒上进行 linpack 期间，我能够拉出 50 GFlops。
另一个选择可能是 cuda
If you have an x86 system, you can use the Intel Math Kernel Library.
http://software.intel.com/en-us/articles/intel-mkl/
I was able to pull out 50 GFlops during linpack on my old quad box.
Another option may be cuda
您可能想查看几何工具的数学库。如果您不创建自己的引擎，您可能需要查看诸如 OGRE 之类的完整包。
You might want to have a look at the math library of Geometric Tools. And if you're not creating your own engine, you might want to look at something like OGRE for the complete package.