For coding Presto on the CPU, we didn’t want just a “port” of the code from Nvidia CUDA architecture to x86/CPU architecture. We wanted to squeeze the CPUs to the max and see a real performance comparison between GPUs and CPUs. For this, we decided to adapt Intel’s Embree library. What does this mean? with every GPU and CPU core running Presto, it means fast, very fast, rendering!
Detailed results in our Benchmark page