naeioi / PBF-CUDA

Position Based Fluids CUDA implementation
129 stars 21 forks source link

关于渲染速度 #1

Closed cryer closed 5 years ago

cryer commented 5 years ago

作者您好,我想问一下,我在知乎上看到视频中iter为4渲染速度也蛮快的,我笔记本1050Ti 设置iter为1都比你那个慢很多,可以问下您的电脑配置吗?或者说我这里是不是还有除了配置以外的问题,还有什么改进方法?因为我还没有研究代码,只把最大粒子数量缩小到5万,还是一样的慢。

naeioi commented 5 years ago

您好,谢谢你的关注。1050Ti理应能应付30w粒子的实时渲染。您的编译模式调整到Release了吗?

cryer commented 5 years ago

哈哈,忘了一直在debug模式,后来试了release,发现nanogui的windows.cpp 36行一直报内存错误,读取位置 0xFFFFFFFFFFFFFFFF 时发生访问冲突,解决了一会还是这个错误,就把release的优化选项关了,就可以运行了,优化关了速度还是比我原来快很多很多。。我也没想到差距会这么大。。谢谢大佬了

naeioi commented 5 years ago

不客气。

DanSon3 commented 4 years ago

我想问下作者,怎么编译的时候一直提示GLuint这些变量未声明

Sycamore-Ma commented 1 year ago

Hello, I was trying to do some paper reproduction based on your project recently, but some configurations needed to be fixed about NanoGUI when I tried to compile your project. Does your NanoGUI dependency work correctly in your local environment? Have you made any other external references globally?