Tiled Hacker news on React Router

CUDA Shared Memory Swizzling

53 points - last Thursday at 2:13 PM

johndough
today at 9:42 AM
I was wondering why the author was using braced initialization like
```
    size_t i{0};
```
instead of the more common
```
    size_t i = 0;
```
Apparently, braced initialization does not allow narrowing conversion, so you'd get a compiler error for e.g. casting double to float
```
    size_t i{0.0};
```
and a warning for
```
    double d = 0.0;
    size_t i{d};
```
which might silently overflow size_t otherwise, so this is a bit safer.
In C++, you can get the same effect without the unusual syntax by passing -Wfloat-conversion to gcc/clang, but not sure how to do that with CUDA:
KeplerBoy
today at 8:49 AM
Lei Mao's blog is such an amazing resource for GPU performance engineering. I am stunned by the sheer amount of insight he puts out on his blog.