Search:

Type: Posts; User: multifitter

Search: Search took 0.01 seconds.

  1. Yes, i know that and therefore i investigated the...

    Yes, i know that and therefore i investigated the register spill with KernelAnalyzer / CodeAnalyst. Both show me no register pressure if i use float8 or float16.

    The loop inside the Kernel runs 2...
  2. slow addition of vector-components using float16 ?!

    Hello,

    i have tried to optimize my kernel using "float8" and "float16" instead of "float4".
    System is XP32, OpenCL 1.2 on an AMD Athlon X2 250 and a Radeon 6750. (testing on a AMD A6 3450M APU...
Results 1 to 2 of 2